Learn about NexusCon!
Happy Thursday!
Welcome to the first Nexus deep dive exclusively for Nexus Pro members. It’s an honor to have you here. And let’s kick this off right: use the comments to introduce yourself and let us know what you thought of the episode and deep dive.
This deep dive is a follow up to my recent conversation with Alex Grace, VP of Business Development at KGS Buildings. I thoroughly enjoyed this conversation and want to share my takeaways and the full transcript with you below.
In case you missed it in your inbox, you can find the audio or video here:
Nexus site | Apple Podcasts | Spotify | YouTube | Add to other podcast apps
Enjoy!
Disclaimer: James is a researcher at the National Renewable Energy Laboratory (NREL). All opinions expressed via Nexus emails, podcasts, or the website belong solely to James. No resources from NREL are used to support Nexus. NREL does not endorse or support any aspect of Nexus.
Here’s an outline of today’s deep dive:
After these two conversations with KGS Buildings in back to back podcasts, it felt like we stumbled upon a potential new interview strategy: talk to different people from the same company and go deep into different facets of their business. In this case, Nick provided what sets KGS apart and where the analytics industry is headed. Alex provided the biz dev lens and a view of analytics-based digital transformation for all types of service providers.
My #1 takeaway from Alex’s wisdom happened when he got on his soapbox about commissioning providers. Here’s his powerful quote:
And here's a different perspective to take. I firmly believe if you're a commissioning firm listening to this right now, five years from now, your best customers are all going to have fault detection and diagnostics. All of them. Do you want to be driving that and be a part of that solution or not? I really think that's what it comes down to. And I know that's a strong statement, but I'm going to make it here, talking with James Dice. I think it's really true.
I obviously agree, but this has made me think…
From a building owner’s perspective, there needs to be a separate strategy from the service provider’s strategy. The owner needs to have a digitization approach that selects the best platform and then drives each service provider to provide a better and more digitized service using that single platform. I’ve seen what happens when multiple service providers serving a single building owner bring their preferred platform, and they’re not necessarily bringing all the other service providers into the solution with them. Chaos can ensue.
What do you think?
Alex Grace: [00:29:35] So just to talk about services for a second, broadly speaking, the old approach of: I have a checklist; I'm going to roll a truck to a customer site; I'm going to go down that PM list; and I grease bearings this frequently, and I change the belts that frequently, and filters on that timeframe. Or I'm time-and-materials, break-fix type focused, and I show up on site every once in a while, maybe one day a month, one day a quarter and my guy tells me what his biggest headaches are, maybe it's written on a piece of paper. And then I go and try and take care of those things. That model is fundamentally transforming and changing, and I really see that there's, there is a carrot and there's a stick, right?
So there's the carrot being the opportunity. The opportunity is for business leaders running the mechanical service business or control service business to realize that I have a major opportunity to differentiate the way we're doing things and to show customer value and to elevate the value of my services to a different level in the organization, that I'm not just thought of as someone that fixes things when there's a major problem, but I'm also thought of someone that helps me strategically understand the operation of my facilities, the risk factors that I'm facing, and where I can drive energy reduction, performance and long term value when it comes to capital planning and asset management as well. So there's a lot, there, but that transformation is occurring.
So, you know, that's the opportunity. And what I mean by that, okay, when you're up for renewal on that service agreement-, you have a $20,000 service agreement with the customer. And now you're up for renewal, and there's a new finance guy, and he says, well, what are we getting for this $20,000? Well, here's my checklist. You know, I checked a ton of boxes. We did all these things for you that was really important based on ASHRAE guideline, blah, blah, blah, blah. Okay. Does that guy know what you're talking about? Does he understand the value of that $20,000 agreement?
Now let's talk about a digital approach. So how does this transform? How does this change? I no longer only roll trucks reactively or based on a checklist. I now, when that guy goes onsite, when my technician appears, he knows exactly what the top priorities are in this building based on energy, what it's costing my customer, based on comfort, which is arguably the biggest cost, your employees themselves. Okay, during the normal COVID times when there are people in those buildings. And then thirdly, the mechanical severity of the problem. Can I fix something that otherwise is going to wake someone up at two in the morning and they're going to have to come into the building and have a massive headache on their hands? Right?
So now the technician shows up with these priorities. He knows what's in scope, based on his current agreement. He knows what's out of scope that he can now propose a work on. So obviously the massive business value there from the vendor side of being able to have a consistent ability to propose ROI-justified work, and from the customer's side, massively more valuable because they know that that technician is focused on the biggest risk factors for them, the biggest energy factors, and the biggest comfort factors for them.
And now let's flash forward to that same conversation a year into your agreement. You're up for renewal with that finance person. You now have a dynamic dashboard or a static report as you wish saying, what did that $20,000 buy you? Here's exactly what we saved you last year as a result of doing predictive maintenance and continuous commissioning. Here's exactly the impact we had on your facility in terms of quantified maintenance value and quantified comfort value.
Let's spend less time identifying and more time fixing. And you know to be clear, obviously we're not replacing all the PMs, right? I mean, you're still gonna need to grease bearings. You're still gonna need to change filters. You're still gonna need to look at certain things. Absolutely. But there's a certain percent of that agreement and that labor that can be shifted to higher value activities. And that's the percent we target.
And we actually go to a pretty deep level on that. So we'll go through an exercise with partners. We'll look at their task list. How do you do maintenance today? And what are your typical agreements? And are you standardized? So that's the first step, right?
You have organizations that made acquisitions, and this branch does something this way, and this branch does something completely differently. You have others that have consolidated their approach to services and say, we have a three tier program. We have good, better, best, or you know, a T and M block hours, a PM, and full service.
So whatever that is, if you've already done that standardization, you're already a leg up. If you haven't, incorporating FDD into your plan could help you do that standardization. So we really go to this level of understanding: what are your service tiers that you have today? We go through a task by task basis. We assign hours to that with our partner. So yeah, if I don't need to check that thing because analytics is looking for it every single day because I have diagnostics running, that's going to save me 15 minutes per activity, cause I don't have to do that thing. And then you add up all those hours and you've come out with, Hey, it looks like I could actually offset a third of the hours here.
And you know, operations leaders can be a little bit threatened by that. I just want to make that comment. If they don't understand the big picture, which is that their guys are going to get way more work out of this than they ever had before. It's just that that work is going to be higher value.
The other thing is what can you do remotely, you know, really figuring out what can you do remotely versus what needs to be on site. And I think now in the age of, you know, coronavirus that's probably even more important than ever, but well after coronavirus has left us, that's still going to be really important for a business model perspective.
James Dice: [00:37:07] Great. Yeah, I agree. You mentioned the carrot and the stick or the carrot or the stick. So we've kind of painted a picture of those that are going after the carrot. What about the people that are getting hit with the stick? What are the organizations that aren't quite keeping up with this digital transformation? What are you seeing happening to them?
Alex Grace: [00:37:28] Sure. I think that the organizations that are able to pick their head out of the noise and the day-to-day, I've got a million things on my to do list, and able to look into the future are seeing that there is a threat here. So there's the opportunity to really have better client outcomes, better client relationships, and derive additional pull-through revenue through their service organizations and provide just a better service for their customers.
And there's a stick of, if we don't do this, in five years is our organization going to be relevant? And that's a bold statement to make. But I really think that's how leaders of these companies are thinking and need to be thinking if they're not. Because imagine for a second that you're going to renew that service agreement, and you're doing things the exact same way you've always done them, and you're doing T & M or block hours, and they started getting calls from people that are offering the same value service from a similar price point with way more. With an analytics driven approach and reporting on that value to the customer at the end of the year . Where the customer value is just that much more explicit.
And then one thing I do want to circle back on, and the carrot that we're really focused on is I think a lot of companies are still thinking about FDD or a tool like Clockworks as another tool in the toolbox. So, you know, we talk about shifting from reactive to proactive maintenance. We also talk about shifting from reactive to proactive business model, right? So I want to bring that up, which is, you know what I mean by that is if you're an organization that is thinking about FDD just because one of your more sophisticated customers is asking about it and asking for your help, you might want to take a step back and say, do we want to continue to be reactive to customers asking us about this, or do we want to have a strategy?
So we are not, frankly, not as interested in partners that want to resell Clockworks as another tool. You know, okay, the customer asks about it, you know, here's something I can maybe talk to them about. We're much more interested in business model transformation, because that's what this industry, that's where things are going. And that requires a whole different range of discussions about understanding the partner organization, how they operationally deliver maintenance today, and the structures I was talking about: good, better, best; block hours, PMs, full service. And it also requires understanding how they go to market. You know, who's selling those service contracts? How are they incentivized? You know, are they thinking about building up a remote operation center, a network operation center? So it's the business side. It's the sales side. It's the operational delivery.
And for that reason, sometimes people are surprised, like they'll reach out to us with an RFP and say, Hey, we're evaluating five FDD vendors and they're sometimes surprised about how much we ask about their organizations and their businesses. And the reason is that we need to qualify our partners as well as partners qualify us, because we go deep. We go deep in understanding them and consulting on that business model transformation because we think that's where the most value is on both sides.
James Dice: [00:40:27] Fascinating. So let's go through some of these business models.
So I have a list of about five different potential partners and channels of yours. So starting with these service contractors, what does the service contract of the digital world look like? So how are they packaging the technology in with their services into this new package?
Alex Grace: [00:40:49] Yeah, absolutely. So part of that equation I was saying before where you determine are there costs that can be offset? That's part of the equation. So let's say for example, you have, this is just hypothetical because it does depend and you need to dive into the details here to really be able to stay this, but as a hypothetical situation, you have a $20,000 service agreement. Let's say you're able to remove $5,000 in labor. And by remove, I mean offset. So you're gonna replace that with some diagnostic monitoring and you're gonna shift some of those hours to remote operation center approach.
And it doesn't mean you're starting with, you know, NASA space command. It means that you have a guy in the office, who is one of your most experienced technicians or an energy engineer, someone who really understands data and buildings and systems, and that person, by the way, can now be 10 times more valuable because you don't have them on customer sites one-to-one. You have them one-to-many, and they're now able to guide, with the diagnostic results, less experienced technicians. And there's a lot of operational value there as well. So you take that best guy, you put them in the office and you say, okay, we're gonna take an analytics-driven approach.
So you've got that $20,000 agreement, let's say $5,000 gets offset by a combination of hours internally from your experienced guy in the remote operation center and the cost of Clockworks to look at things. And then there's going to be some delta there that you're going to have to upsell. But the upsell is now way more value add, right? You're able to now talk about prioritization, about scoring that building in terms of metrics.
So every day, what are your top priorities relevant to energy, comfort, and maintenance, and how do you drive actions and quantify the results of the fix? You're now able to do that and show examples of that. And we're even seeing some organizations that are actually subsidizing that transformation. Meaning, are you willing to actually eat a little cost because you know that you're going to get way more project pull through work? And because you see it as a strategic priority to have this transformation occur for all the reasons we talked about related to the carrot and the stick.
So some combination there, and it's gonna look different for different orientations. You absolutely don't need to be eating it, but some are getting aggressive and saying, you know, this is important. We're just going to do this because we know it's gonna produce fruit and we know it's going to make us more profitable and expand our service business over time.
James Dice: [01:00:28] The one thing I was thinking about though, that I don't know that the service providers are thinking about is that when you have this common single source of truth that you were talking about earlier in the context of institutional versus shared knowledge, so that shared knowledge from my perspective, what I think is valuable for all of those service providers is that when you have, say, your fan motor is not working. Whatever the problem is that comes out of Clockworks or any FDD, that service provider is going to know the model number, the size of the motor, everything to do with that in that information model, before they leave the shop, right? So they're already knowing, like you said earlier, what their exact task list is when they get to the building. But the point would be that they already are ready to perform the exact task. And I think there's this, I don't know if that message is getting out there as far as the process efficiencies.
So it's not just about labor savings, it's also about everything to do with the process of performing these ongoing services.
Alex Grace: [01:01:40] So I think you're really touching on a key thing, which is the asset management side. So what are the datasets and the data silos that need to be combined, right? So you now have this fault detection information and a history of fault detection. So how many times has that valve leaked on that air handler, and how many excess kBtus has it cost you in the last three years, every time you've fixed it?
You also have the asset service history. So what are the lists of fixes and PMs and reactive break fix, and what are the labor and material costs associated with that? So being able to combine the datasets from an asset management tool or a CMS tool with fault detection, we see as really critical, and we are having that conversation with some service organizations that are larger, that are really thinking about the value of that data.
I mean, everyone's talking now from a business perspective, the values of data, data is the new oil, et cetera, et cetera. Right? And those that are thinking about like, what is the strategy there? So you have all this asset history, you have all this fault detection history. What are you providing?
So specifically to what you're saying, the process efficiency, absolutely. Not just showing on site with a prioritized list, but knowing what parts to bring. I mean, you can't underestimate that if you're rolling a truck, particularly if you're in a situation where you have to travel some distance. You know, if you have to send a guy an hour away or two hours away, even more critical that you know what the problem is and that you couldn't have fixed it remotely.
So imagine you fix the things that you can fix remotely, cause you have remote access and you're providing, for example, control service. And the things that you can't fix remotely, you know the exact sensor you need to bring or the exact actuator that needs to be replaced or whenever the tools are that are gonna allow you to pinpoint the fix that's needed.
James Dice: [00:05:12] How about from an energy efficiency standpoint? So when I think about an energy efficiency project, you're doing some sort of study, right? And then you're installing some sort of ECMs, energy conservation measures, and then you're realizing savings after that. But what I've found is with building owners that have a hard time wrapping their heads around the Software as a Service fee, because it doesn't produce savings in and of itself usually, and so how are you guys helping to justify those ongoing fees? Because as we know, they're vital for maintaining the savings that come out of energy efficiency projects, but they're not necessarily required to implement energy efficiency projects themselves.
Alex Grace: [00:05:56] Right. Yeah. There's a couple interesting pieces there that you touched on for sure. So, some changes that we're seeing... I mean, you're absolutely right. There's a fundamental market education piece there, which is about energy drift, which those in the energy auditing, commissioning world know well, right? That just because you implemented a beautiful reset schedule, it doesn't mean it's still being followed a day later, much less six months later, much less two years later, right?
So probably how we've been from an M&V perspective, calculating savings and thinking that low cost, no cost measures actually just persist, is probably not right. That's one point.
And then secondly, you know, I think that on the maintenance side of the house, this really isn't as much of a concern, but on the energy project side of the house, it is more. And really the obvious point is valves don't stop leaking; dampers don't stop sticking; sensors don't stop failing. So everyone, again, on the O&M and the maintenance world knows that very well, but from an energy projects perspective, you may have the idea that you go and implement something, you squeeze the savings out, and you walk away. And it's just simply not the case.
So you know, it's, it's why our angle is really... Our clients and our partners are looking to drive predictive maintenance outcomes. They're looking to implement continuous commissioning processes for all the reasons that we know: because it improves operational efficiency, because their people can be focused on what matters most rather than just reacting and going down a checklist. They're zeroed in on the issues that are costing them the most amount of money, that are having the greatest maintenance impact or having the biggest impact to building occupants, the building comfort. So you know, that's why they're doing it. And then the natural byproduct of that process is energy saved.
And I like to talk about it and frame it in that way: that continuous energy savings is there, of course it is, but for a lot of our users, it's not the primary driver, although it may be very important to get that initial budget approval. And we have a client that will give a speech on stage and say, you know, we didn't care about energy. We're completely focused on maintenance optimization. We're having trouble retaining staff. You know, people are retiring. There's a major labor shortage in the industry, as we all know. We need to make sure that our people can really be laser-focused on having the biggest impact and getting more from our existing teams.
And then you'll say, oh, but yeah, we also saved eight hundred thousand dollars. So it's like the energy savings is there, whether you're focused on it or not.
Here’s a summary of their program:
James Dice: [00:43:13] Fascinating. Okay, let's move on from the service contractors and go into commissioning firms and talk about the inherent one-time commissioning effort or one-time retrocommissioning effort business model versus the ongoing monitoring based commissioning model that analytics offers and presents as an opportunity. So can you talk a little bit about the carrot and stick for those guys and how the ones that are on top of digital transformation are transitioning their business models?
Alex Grace: [00:43:47] Yeah, sure thing. And this is a cool one cause I know you've got a lot of background here. Well, you have a lot of background in a lot of areas, but this'll be a fun one to talk about.
So the commissioning world has honestly been fascinating to me. And it's fascinating because I've always sort of wondered, why don't we have more commissioning firms that are using Clockworks and providing an ongoing service? And I saw, actually one proactive commissioning company that we're talking to right now turned me on to this report from the Building Commissioning Association, the BCA, that was really confirming for me. And it showed a study, and it was survey of their members, and the result of that survey was that over 90% of monitoring based commissioning projects are not going past 12 months. So they're ending after one year.
James Dice: [00:44:34] That is a staggering statistic for me to think about.
Alex Grace: [00:44:39] Yeah. Yeah. I thought that was incredibly fascinating. So I think the problems that I'm seeing the commissioning world are really business model based. They're obviously not technically based. I mean commissioning firms are incredibly well positioned in terms of skill sets of their people to drive an ongoing commissioning process with fault detection and diagnostics at its core. I mean incredibly well positioned.
And don't get me wrong, we're seeing plenty of companies that are using, I would say, toolkits in their commissioning toolset. So, okay. Rather than, you know, downloading data from the control system. First of all, there's still a lot of companies doing this: downloading data from the control system, spending three weeks messing around with it in Excel back in the office to come up with a report. Like, if you're still doing that, obviously that needs to change. I mean, that's just crazy.
But , and then there are companies that are changing that by going to, I'd say, a toolkit approach. So they're using an FDD tool where they can write their custom algorithms, because it helps them do their commissioning process. But they're still fundamentally project-based businesses, meaning: they do a project; they end it, whether it's instead of three months, now it's 12 months, it's still very limited; and then they move on. And the mentality is, well, the facilities teams are gonna pick it up from there. There is a big gap in the big market opportunity, I think, for commissioning firms to stay involved with their clients longer and to be adding more value on a continuous basis as it relates to operations and maintenance, by running FDD longer and continuing that process. I think it's not happening because it's not how project teams are based, and it's not how companies are structured or incentivized. So I think they're missing an opportunity.
And here's a different perspective to take. I firmly believe if you're a commissioning firm listening to this right now, you're working at a commissioning firm, five years from now, your best customers are all going to have fault detection and diagnostics. All of them. Do you want to be driving that and be a part of that solution or not? I really think that's what it comes down to. And I know that's a strong statement, but I'm going to make it here, talking with James Dice. I think it's really true.
You know, so if you're in this mentality of, well, we just don't know why customers aren't renewing and we're doing it for 12 months. Then you're missing something, because we have clients doing this for 10 years. And fundamentally, when organizations we work with start using Clockworks, they don't stop. Because how do you go back to reactive maintenance after you have a tool that tells you every single day where your priorities are?
And there's stuff that comes out of Clockworks that is more complicated, right? There's the leaking valves and the stuck dampers and the broken sensors, but there's also the: we could be staging our cooling towers more efficiently than we are, I need someone to rewrite the sequence; or our loops are all under loaded, Clockworks is telling me that the delta T in our chilled water loops is performing consistently bad, and it gives me a range of recommendations on how to fix that, but I still need an engineer now figure out what is the solution? What is the exact sequence I need to rewrite here to improve?
So I think sometimes people miss that, that there's a lot of engineering work that comes out of FDD if the FDD is actually doing the diagnostic piece and not just fault detection. So yeah, there's an opportunity there and there's the stick, which is all your clients are going to have this and how is your work going to change if they all do and you weren't a part of driving that solution?
James Dice: [00:48:09] Right. And I came from this world, so the way I see it is these companies have existed for a while. You know, commissioning and retrocommissioning have been, well-ingrained processes for a little bit. You know, there's still a lot of construction projects that don't use commissioning, which is a whole different conversation. There are still a lot of buildings that haven't been retrocommissioned, which is a totally different conversation. But in general, those two practices are pretty well -accepted as projects that are worth people's time, worth people's money.
But those firms that provide those services have a business model that is just like a construction project. It's a one-time event, right? So how are you seeing the business models for those companies transform to accommodate monitoring essentially?
Alex Grace: [00:48:56] Yeah, that's a great point. Okay, so practically speaking, you're doing commissioning, you're doing post-occupancy commissioning. Maybe you have FDD in there already to some degree. Um, this is where I think Clockworks does differentiate from some of the other tools because we do provide this O&M focus and we serve a lot of different stakeholder groups. But let's just say more generally, it doesn't have to be specific to Clockworks, you've got it running, and rather than going away to the next project, you now propose or you've proposed from the beginning that, how about we have a one day a month? So this agent that knows your system, has been testing things, has gone through the commissioning process, we're going to keep him on one day a month. We're going to check in on the information that the fault detection and diagnostics has produced. We're going to make sure you're not missing anything, and we're going to make sure any system retuning that's done continues to happen.
And I think, you know, at a minimum that's gonna make a lot of sense to the customer post-occupancy in terms of identifying things during the warranty period and holding vendors accountable. I'm talking about new construction now or major retrofit, but through that process, you know, you've got to make sure you're getting to the O&M folks, first of all, that you're not just seeing yourself as like, Oh, I'm just giving them an O&M manual and walking away, that you're really providing value of helping them see their issues, and that, you know, just because it's a new building certainly those of us who have done commissioning know that the problems don't stop. Right? So even after you've tuned things up, things are going to keep coming back.
Alex Grace: [00:48:56] If you're working with a portfolio and you're, for example, retrocommissioning a handful of buildings a year, moving your way through the portfolio, maybe you take a different approach.
If you look at the cost of that deep dive retrocommissioning every single year only touching a handful of the portfolio, versus what if I monitored everything every year, and then that same man hours I was spending doing my deep dive functional performance testing, I've shifted to letting analytics drive where to focus. So rather than testing a bunch of systems to figure out where the problems are, how about diagnostics tells me exactly where the problems are, and I spend those man hours helping engineer solutions across the portfolio? Because the cost benefit on that is dramatic. And it may not be even more costly from a customer perspective to do that.
Or the other thing to keep in mind is you don't have to run FDD on everything. You might decide, my customer has a 50 building portfolio. I'm going to monitor every chiller plant, every boiler plant, and every air handling unit across the portfolio. And then focus my retrocommissioning efforts really targeted on the buildings that are outliers based on that data analytics. And it's not just, you know, EUI cost per square foot. It's a much deeper dataset where I'm now seeing, maybe I'm looking at kW per tons, maybe I'm looking at kW per CFMs on ventilation system efficiency, and maybe I'm able to go even deeper and see exactly where sequences could be improved, where they're not being followed, where the resets that I engineered five years ago are no longer in place cause they've been over-ridden, and et cetera, et cetera, right.
I just want to mention on the positive side, we have, we do have commissioning firms that are working with Clockworks. Uh, shout out to WSP in Boston, for example. But you know, organizations that are really differentiating what they're doing and they're going into the customer's sites.
So from a sales perspective, you're a commissioning firm, the power of going into a customer site in a competitive situation-, obviously, if the customer is totally race-to-the-bottom and focused, then yeah, this is maybe not what I'm talking about. But if you have a more proactive, forward-thinking customer, and you're able to walk into that meeting as we've seen some of our partners do and say, we do commissioning differently. Here's how we do it. You know, that's very powerful. And that we're going to have an ongoing O&M focus if you want that, you know . I think there's a big opportunity there to really differentiate your service in the market.
The other thing I want to just quickly mention is macroeconomic trends. So because I think that's been a factor lately also. Because commissioning firms are busy, like they're getting a ton of work, right? No one's, no one has guys sitting around at home starving for time.
So when construction is booming, when the economy is doing really well, and you've got new construction and commissioning projects left and right, why take the time to really take a step back and look at some of these things we're talking about? You know, I get that.
Well, the current reality of where we are with the current economic situation related to coronavirus, but I'll say more broadly, just economic cycles, macroeconomic cycles in general, there's going to be periods of construction slow down. There's going to be periods where we were taking a step back. Now is a phenomenal time to take a step back and think about that and where things are going. And I think that we're seeing that because I'm seeing inbound messages lately and I think a lot more actually than we had been before. And I think that's part of it is people just have a little more time to pick their head above the water from all the work they'd had and say, okay, where are things going? Where is the industry moving? How can I position our company to be where we need to be, you know, now and five years now?
James Dice: [00:54:44] That's fascinating. On this last, so we're kind of continuing on these different types of business models. So commissioning firms are often, sometimes part of, you mentioned WSP, part of larger MEP design firms. So how do all of these lessons apply to those types of businesses and what are you seeing for them right now?
Alex Grace: [00:55:04] Yeah, that's really cool. I think a couple things. One, every design engineer I've ever talked to is dying to know how their designs actually are performing in the field, because you never get that feedback as a design engineer, right? I mean, you have a model, you put the best, all of your knowledge into this design, but you don't really know how the operation is going.
You know, is it being operated the way it was designed or not? So that feedback is obviously very powerful , and we've seen that everyone wants that. You know, there's a, there's a message there when it comes to new construction from an owner perspective also, around you're spending millions, hundreds of millions of dollars, whatever the case might be, on this building and on these designs and on the install of these systems. For pennies on the dollar, do you want to ensure that you're getting what you paid for? Right? So , yeah, I think that is an interesting area. If commissioning firms that often are tied to larger MEP firms, if they're able to create that feedback loop to the design, that's incredibly powerful. And a tool like Clockworks, you know, would certainly allow you to do that, but that is something interesting to think about.
The other thing I'll mention is just around M&V specs. So we've seen M&V specs come out that, you know, I want a submeter every last thing in order to just see how my designs are performing or at least like compare that to the energy model. Don't get me wrong, that's fantastic. We love sub-metering. Put all the data acquisition out there that you can. But at the end of the day, you're still limited by, okay, performance is different than model. Now what? Like where's the problem? Right.
So I think, I think FDD needs to be thought of more in M&V than it is today. And there's a big opportunity there where it's to say, look, that's great you have a lots of sub-metering. That's great you're able to compare performance against your energy model. But can you figure out where the problem is specifically? Can you pinpoint it? And FDD can tell you exactly what's going on. Well, you know, these two things were reversed and your actuation is totally backwards in terms of how you're economizing or whatever.
James Dice: [00:10:33] Um, so one thing I'm doing right now is creating a standard RFP for anyone who's procuring analytics software. And one of the challenges with that is creating some way to compare software like this FDD software in an apples-to-apples sort of way. So what are your thoughts on how that can best be done, and is it even possible at this point?
Alex Grace: [00:11:15] Yeah, it's a great question. It's definitely a challenge. I think it is possible. I think it's really about thinking beyond the initial... I have analytics up and running, or I have some basic fault detection up and running, but what is this gonna look like at scale, and what is this going to look like three years, five years from now? Meaning, what is the maintenance on the fault detection and diagnostics?
So I think something that's often not maybe given as much attention to as it needs to, is that buildings are dynamic. They're changing all the time, and the FDD needs to evolve with those operational changes. So you set up a bunch of static rules that were customized to a building, and then you change your sequence or you replace and air handler a year later. Who maintains that code set?
So, you know, are you getting costs associated with just standing up some initial FDD and hoping it works today? Well, that FDD may be producing a lot more false positives a year from now if you're not thinking about how that code set evolves and needs to be maintained. That's one point.
It is a challenge. You know, I think we love seeing people test multiple solutions for that reason. It just, it helps us because sometimes things can be abstract until you've had your hands in it, you know? So we've had clients-
James Dice: [00:12:29] Sorry, what was that? Test multiple solutions?
Alex Grace: [00:12:31] Yeah. So if you're looking at multiple FDD products, you know, you have a couple of buildings. Or you know, the most fun cases we've had is where someone actually has a historian that they're able to serve data to multiple FDD vendors and see what happens. I mean, that's a great way, but there's very few people that can do that. And you know, I don't recommend spending a year trying to figure out how you abstract your data in order to do that. You know, you want to get started and that's really the bottom line, right? Get started. You're going to learn a lot, you gotta move forward. And I think that requires leadership to really have that fail fast mentality. You know, I talk a lot about the University of Iowa because they've really done that well, you know, from the beginning, they had this mentality that we're going to test a solution, we're not going to be tied to it. We're going to learn a lot. And, and having that, making sure that your organization is a learning organization is, I think fundamental here. Yeah, so apples to apples is tricky, I'd say the majority of our biggest, and best clients have already done other types of fault detection before because they just get the complexity. They get that it's not as simple as just tagging everything and expecting things to work, that you have to really understand the operation, and that false positive is the Achilles heel of this industry, you know? So the ability to effectively prioritize and avoid false positives is critical.
And I do think that there's ways to write requirements that ensure that you're getting, for example, energy calculations accurate. And that rules are aware of other rules. And I know when you had Nick on, you guys talked about mass customization, so to have a custom solution that's going to degrade the second that you have it up and running, but you have something that is constantly evolving and code sets that are shared and parameterized in this mass customization approach. And another way, I like to say that just that the rules are aware of each other. So if you're implementing a bunch of rules, you know, I think that's key.
And I love the conversation you sparked on LinkedIn, James, around like, where's the IP? Is the IP in the rules themselves? And I just think that's a really important point from my perspective, that, or from the KGS standpoint, that, you know, identifying that a valve is leaking or a damper or stuck is not that challenging. There's no IP in the math associated with finding a leaky valve. The IP is how do you structure those code sets in umpteen different ways and hundreds of different scenarios with different types of mechanical configurations and sequence of operations, and how exactly is that economizer controlled that's controlling the damper? And what about dehumidification and all these things that you run into as you know from your background and your experience, right?
We hear a lot of conversation in the industry around fault detection focused on tagging. Tagging's important. Normalization's important. You know, Nick spoke more eloquently than I can on all the different things he's involved with related to the ASHRAE BACnet Committee and establishing industry standards that we want to be really heavily involved with as we think it's incredibly important to push things forward.
That being said, you know, for us, tagging is a starting point. It's not an ending point. I think a lot of people talk about it as an endpoint, and frankly, if you have very simple rules, it is an endpoint, right? I know what the data is, and therefore I'm running a full linear logic to say, you know, is something off here? Has it deviated?
You know, we sort of see that type of quote unquote fault detection as really alarming or advanced alarming potentially, but to get to the level of diagnostics, you have to understand a lot more than just, what are the points? You have to understand where equipment is. You have to understand where thermal elements are exactly in an air stream.
So not just understanding the points, but what is the order of those points in that air stream in an air handler? And then what are the sequences of operations that may or may not be reflected in the modes or the set points or the parameters that are visible from the points. They may be just written in a document somewhere, right?
So, and then having code sets that can dynamically determine based on the best available information. So that's a key thing that's a little bit harder to understand and requires some explanation, but it means that you want to have a situation where your rules don't just break because it's missing a variable.
And the way Clockworks diagnostics work is that it always looks for the best available information, meaning, okay, if this point's here, I'll use it. If this point is not here, can I use this other point or can I approximate with these two points or reference a inputed piece of metadata from a written sequence of operation to enter it?
So a really simple example would be, do I have a flow station? Yes, I have a CFM point. I'm going to use that in my calculation. No, there's no flow station. Okay. Reference fan speed and reference the rate of flow of the unit to approximate flow at any given point in time. Really simple example, but there's hundreds and hundreds of those.
Alex Grace: [00:18:45] And Don I hope he doesn't mind me stealing this quote from him where he says, let me tell you an example of institutional knowledge. He gives example of a technician who would walk around with pieces of wood. He knew exactly the right size piece of wood to jam in the damper to be able to fix a problem. And he goes, that's institutional knowledge. He's like, we don't need institutional knowledge. What we need is shared knowledge. And I just think that's a really powerful illustration of what we're talking about.
And so what is shared knowledge? It means that you have a single source of truth. It means that you have, in this case, a diagnostics platform, where the data doesn't lie. It's accurate. It's tuned to your operation and tuned to your sequences, and that a mechanical technician, a controls technician, an energy engineer, a sustainability professional, a design engineer on the capital planning side, they all have access to that same information based on what they need to do their job. And I think overall, as an industry, that's a really useful concept.
The reality is institutional knowledge is retiring. That's the reality that we're all facing. Right? And the next five years is more and more of that in 10 years more and more of that. So it's not really an either or, it's how do you replace institutional knowledge with shared knowledge and what are the organizational benefits associated with that? That I think is really powerful.
The other thing that's powerful, I think, from their story is just having many different stakeholders involved very early in the process. You know, I think something we may have talked about before, James, that oftentimes I think FDD gets siloed within a particular use case, and it's something Nick touched on a little bit as well. Right? You know, is it within commissioning? Is it within energy manager who's doing energy projects, or is it that shared knowledge tool that is spanning these different use cases? And it's also fine to get started with one use case, right? You don't want to boil the ocean. Totally fine to get started with one and then be aware of how you expand.
But for those that are higher up in an organization and thinking about this from a leadership perspective, mapping out your stakeholders, which is not a mystery. It doesn't require a whole, you know, consulting engagement. It's your controls technicians, it's your maintenance technicians, it's your commissioning agents or third party commissioning folks, and those on the design or capital planning team side. And how do you bring them together to derive value from a platform like Clockworks?
James Dice: [00:21:58] Got it. Okay, cool. Yeah, thanks for sharing that. It sounds like a really cool project. So a couple of last questions around sales. I'm just wondering from the perspective of you, you know, spreading the messages of analytics over the last however many years, many years, right? Um, what's it been like watching these buzzwords come into our industry? So, I'll start with IoT. I mean, when I think about Clockworks, you guys have been putting things on the internet for long before IoT was hot. So what's that been like for you and crafting a message around that?
Alex Grace: [00:22:34] Yeah. That is interesting. I mean, we're a pretty conservative company, so our marketing goes about as far as, you know, us doing podcasts with you, James. Right? So, so from that perspective, you know, we've always been a little bit skeptical of the hype cycle, and we try to not get too deep into it, right? We don't just shift our messaging immediately that, okay, now everything's IoT, or now everything's AI or whatever, right?
Um, but there's a lot to dive into there. So, you know, I think early on when IoT was coming out... So, okay. The whole conversation on IT OT convergence is totally fascinating. It's a 10 year old or 15 year old discussion and it's still totally relevant. Right? So I think early on when you saw the, the big IT players coming out, you know, I saw a presentation from Cisco, this is a number of years ago now, but to me it sounded like they'd never heard of a building automation system. In the way that they were talking about IoT, you know, it was like, we're gonna have these sensors, and they're going to talk to the centralized location, and you're going to be able to control things. I was like, um, excuse me. Right?
So that being said, obviously there's a lot of great IoT use cases that are out there. There's a lot of really exciting stuff. I mean, I know you talked about InfiSense, and we arere watching LoRaWAN, and we're watching that whole world. And there's a lot of really cool use cases that are out there, don't get me wrong. But I think my organization... I guess my reference point a lot of time with the types of clients that we work with, which those that have big facilities portfolios. They have large facilities teams. They have a lot of infrastructure. Before you start chasing IoT and new buzzwords, let's start using the data you have. Like you have already made millions and millions of dollars of investments in sensing technology, and that data is incredibly underutilized. So there's a massive opportunity to say, look, let's start with what you have. And then absolutely as there are more opportunities to bring in IoT, that's great.
But now I'm also talking from a little bit of an HVAC-centric perspective. You know, there's a lot of really cool use cases out there in janitorial or in other types of areas where, you know, having sensing that wasn't there before, it can be understandably a game changer.
Alex Grace: [00:25:03] I think there's a lot of, frankly, market education that is needed around that. I mean, we still sometimes see RFPs, for example, come out and say like, you know, do you have AI? And it's, you know, there's a lot there. So let me start. So, machine learning, we see a lot of potential in certain techniques and we're using it today for the main area of improving onboarding.
So we use natural language processing and machine learning to help onboard buildings more effectively. We have an incredible dataset. We have about 260,000 mechanical assets, meaning an air handler, a chiller, a boiler, a VAV, a pump in Clockworks today. And all of the data associated with those 260,000 equipment from about 380 million square feet, has all been totally normalized to a standard information model.
So based on that information that we have, when we ingest data from a new building, we have tools that will take that points list and help figure out what those points are based on everything we've seen before. And that is not just, have I seen that exact point name before? And match it to a database. It's obviously more complicated than that. And, our chief scientist can speak a lot more to the actual techniques we're using there, but what's exciting about it is it will continue to evolve. The more data you collect, the smarter you get, and we have a really significant dataset that will only continue to grow. So we see a lot of potential there for machine learning.
On the diagnostic side , it gets interesting. Right? So for those that are a little less familiar, you know, fundamentally machine learning, you're going to train a model on a certain dataset, and then you're going to identify deviations from that model, right?
We have yet to see examples of machine learning for diagnostics that get us closer to an accurate result than what we are already doing, which the academic term will be an expert system, which is a form of AI, an early form of AI, and hierarchal rule-based FDD, which is a fancy way of saying the rules are aware of each other, and getting to the root cause of the problem as much as possible.
So in other words, if you find a deviation-, if the air handler was in normal operation, and you trained a model on that normal operation, and now the air handler is in something different, it's deviated from that, okay. But can you say exactly what's wrong as a result of that deviation in a way that is more accurate than saying, I can see the temperatures on either side of this coil and the valve position is closed and therefore that valve is leaking? Or we're simultaneous heating and cooling, and I'm referencing my dehumidification sequence, and I can see that, yes, we should be dehumidifying, but we're dehumidifying too much. Like that degree of specificity and then adding the loads, looking at the excess kBtus, turning that into dollars, adding a comfort ranking or maintenance range to it, which is what we do today with our expert system. We're not seeing anything yet within machine learning that lets us do it better than that.
James Dice: [01:03:18] Totally. Yeah. And so I can't help myself. I also have to ask about, so from the perspective of the owner, a lot of what I hear from owners in this regard is, is it going to integrate with my work order system? And I hear two perspectives, it seems like from the marketplace in this regard. One is, Oh yeah, we do it and it's happening all the time and it's an easy integration. And then on the other hand, I hear, I heard this in a meeting last week: No one's doing it. It's too difficult to do. It's too hard to write integrations for every CMMS or computerized maintenance management system. So what are you guys seeing for your clients and what have you guys built up as far as integrations with work orders from faults?
Alex Grace: [01:04:00] Yeah, that's super interesting. I'd say the truth is somewhere in the middle, from my perspective. So I'll say that off the bat, we are doing it successfully. We have work order integrations in place with multiple customers. And it's also true that just because you have Maximo, everyone's implementation of Maximo is different.
What fields specifically you're filling out and how those fields need to be mapped to combine an output from diagnostics from Clockworks with that system needs to be defined. So it's both-and. So fundamentally, if your work order system has an API, we can talk to it. If you don't, please upgrade to a system that does, you need to be thinking about that, right?
So, for example, if you have like a legacy platform, a lot of people are in process of upgrading from, for example, FAMIS local to FAMIS online or whatever the equivalent is , and there's a ton of tools out there. So we can do the integration, absolutely. There is a certain element of custom software development associated, because again, you have to define those fields and make sure that you're mapping that process appropriately. But we also have, once you do it once for a Maximo or AiM or FAMIS, that is replicable for us. But then it's the fields and the details that still need to be defined.
So basically what we're seeing is it's not something you do out of the gate because there are costs. You may not do it for the first building that you do, but if you've reached a certain degree of scale, everyone we know both in terms of our enterprise accounts and partner organizations are thinking about this, that they need to do it. Some of them already have. There's great process efficiency gains there that are useful. And also having those combined datasets, you know, as you pointed out around asset management can be powerful and is important to think about.
What did you think about these highlights? Let us know in the comments.
Note: transcript was created using an imperfect machine learning tool and lightly edited by a human (so you can get the gist). Please forgive errors!
James Dice: [00:00:00] Hello friends. Welcome to Nexus, the smart buildings technology podcast for smart humans. I'm your host, James Dice. If we haven't met before, I write a weekly newsletter on this same topic. It's also called Nexus. Each week I share what I've learned, my opinions, and what I'm excited about in the quickly evolving world of intelligent buildings. Readers have called Nexus the best way to stay up to date on the future of this industry without all the marketing fluff. You can check it out and subscribe at nexus.substack.com or click the link in the show notes. Since starting the Nexus newsletter, many of you have reached out to me wanting to talk shop, and we have. After a few weeks of those wonderful conversations, I realized I needed to record and share them with our growing community.
So here we are, the Nexus podcast is born. This is our chance to explore and learn with the brightest in our industry together.
One more quick note before we get to this week's episode. I'm a researcher at the National Renewable Energy Laboratory, otherwise known as NREL. All opinions expressed on this podcast belong solely to me or the guest. No resources from NREL are used to support Nexus. NREL does not endorse or support any aspect of Nexus.
Let's get to it then. Episode four is a conversation with Alex Grace, Vice President of Business Development at KGS Buildings. This episode is complementary to last week's episode with Nick Gajewski, Alex's CEO, and I think you'll like it just as much. These guys are like the dynamic duo of fault detection and diagnostics.
I picked Alex's brain on a range of topics such as: keys to selling software as a service in the buildings industry; how the leading mechanical controls and commissioning service providers are digitizing their offerings and business models; and where today's buzzwords like IoT, machine learning, and AI fit in with more traditional analytics like FDD; and much, much more.
You can find Alex online, on LinkedIn and at KGSBuildings.com both of these links can be found in the show notes on nexus.substack.com. Without further ado, please enjoy Nexus podcast episode four with Alex. Grace.
Hello, Alex, welcome to the podcast.
Alex Grace: [00:02:05] Hi James, great to be with you.
James Dice: [00:02:07] Please introduce yourself and your role, at KGS.
Alex Grace: [00:02:12] Sure thing. My name is Alex Grace. I'm the Vice President of Business Development at KGS Buildings, and been with the company about seven and a half years now.
James Dice: [00:02:21] Great. And for those who don't know, what is KGS Buildings, and what is Clockworks?
Alex Grace: [00:02:27] Sure, yeah. So KGS Buildings is a fault detection and diagnostics and performance monitoring company, and we are totally focused on FDD. So that is all that we do, and Clockworks is our product. And that is the platform to be able to consistently diagnose building issues and help service providers and end users figure out what's going on in their buildings.
James Dice: [00:02:50] Great. And, just for everyone who's listening to this and hasn't listened to the episode from last week, with Nick Gajewski, we went deep into Clockworks, and I recommend checking out that conversation. So let's jump into the business development side of things at KGS. First question, how do you think selling software as a service is different for buildings than it is in other industries?
Alex Grace: [00:03:16] Sure, yeah. That's interesting. I think in some ways the facilities world, you know, was a little late to the game of SaaS, right? Um, you know, Software as a Service disrupted a lot of industries first, I would say. So, you know, if we think about when's the last time you went and actually loaded a CD in your personal life onto your harddrive to install some software or installed anything locally? I mean, everything we do now is, is through a browser, right? And delivered in a SaaS format. I mean, there's no, it's, it's all about Microsoft 365, right? Or Adobe, or you name the product in any other realm, you know, it's pretty much all Software as a Service.
Um, so I think we were a little late to that, but I think at this point, there's a lot of comfort with that model, especially because, again, in the rest of our lives, it's sort of a foregone conclusion.
James Dice: [00:04:06] Got it. What about from the budgeting standpoint? So are building owners getting used to paying for things in this ongoing way versus... I think about a building owner. They're buying a construction project. They're buying a retrofit project. And those are two separate budgets, right? From operations to capital. So how do they think about that from the standpoint of buying software these days?
Alex Grace: [00:04:29] Yeah, that's a good point. It's definitely still a challenge sometimes. I mean, different organizations deal with it differently, I'd say different sectors, but for sure, you know, we're in the category of an operations budget. That being said, we do have plenty of clients that will, especially with new construction, sort of bury the cost of monitoring for a couple of years in a new construction budget as a way to fund the initial piece, particularly as they're thinking about their commissioning budget or an M&V budget, or, you know, you can end up in different categories depending on what makes sense.
Um, but for the most part, you know, being able to build that into O&M , is key, and you're right, can be a challenge sometimes for people to realize that there's operational tools that need to be in that category.
James Dice: [00:05:12] Totally. And how about from an energy efficiency standpoint? So when I think about an energy efficiency project, you're doing some sort of study, right? And then you're installing some sort of ECMs, energy conservation measures, and then you're realizing savings after that. But what I've found is with building owners that have a hard time wrapping their heads around the Software as a Service fee, because it doesn't produce savings in and of itself usually, and so how are you guys helping to justify those ongoing fees? Because as we know, they're vital for maintaining the savings that come out of energy efficiency projects, but they're not necessarily required to implement energy efficiency projects themselves.
Alex Grace: [00:05:56] Right. Yeah. There's a couple interesting pieces there that you touched on for sure. So, some changes that we're seeing... I mean, you're absolutely right. There's a fundamental market education piece there, which is about energy drift, which those in the energy auditing, commissioning world know well, right? That just because you implemented a beautiful reset schedule, it doesn't mean it's still being followed a day later, much less six months later, much less two years later, right?
So probably how we've been from an M&V perspective, calculating savings and thinking that low cost, no cost measures actually just persist, is probably not right. That's one point.
And then secondly, you know, I think that on the maintenance side of the house, this really isn't as much of a concern, but on the energy project side of the house, it is more. And really the obvious point is valves don't stop leaking; dampers don't stop sticking; sensors don't stop failing. So everyone, again, on the O&M and the maintenance world knows that very well, but from an energy projects perspective, you may have the idea that you go and implement something, you squeeze the savings out, and you walk away. And it's just simply not the case.
So you know, it's, it's why our angle is really... Our clients and our partners are looking to drive predictive maintenance outcomes. They're looking to implement continuous commissioning processes for all the reasons that we know: because it improves operational efficiency, because their people can be focused on what matters most rather than just reacting and going down a checklist. They're zeroed in on the issues that are costing them the most amount of money, that are having the greatest maintenance impact or having the biggest impact to building occupants, the building comfort. So you know, that's why they're doing it. And then the natural byproduct of that process is energy saved.
And I like to talk about it and frame it in that way: that continuous energy savings is there, of course it is, but for a lot of our users, it's not the primary driver, although it may be very important to get that initial budget approval. And we have a client that will give a speech on stage and say, you know, we didn't care about energy. We're completely focused on maintenance optimization. We're having trouble retaining staff. You know, people are retiring. There's a major labor shortage in the industry, as we all know. We need to make sure that our people can really be laser-focused on having the biggest impact and getting more from our existing teams.
And then you'll say, oh, but yeah, we also saved eight hundred thousand dollars. So it's like the energy savings is there, whether you're focused on it or not.
James Dice: [00:08:25] Right. That's fascinating. And I think guys like you can really help the industry, and that's why I wanted to ask you this. We can really help the industry explain these new services and processes and technology a lot better because I think they're misunderstood in many ways.
Alex Grace: [00:08:41] Yeah. And James, one thing that came to mind there, just like as an interesting question that I think used to come up more, but it still comes up sometimes is this idea that I think maybe comes more from the commissioning approach where you're like, Hey, let's put fault detection on a building, and then let's move it, right? We do it for a year, and then let's move it to another building. And that's, that's a little bit from that same mentality, right, of sort of missing the fact that that mechanical degradation doesn't stop.
James Dice: [00:09:07] Yeah. And I've seen monitoring based commissioning incentive programs structured, where it's literally incentivizing firms to do exactly what you just described. I know the one in Chicago with ComEd is exactly like that. It's providing a one year incentive for something that's supposed to be an ongoing part of operations from year zero to infinity, right?
Alex Grace: [00:09:31] Totally. Yeah, and in some ways, I think that those programs are actually hurting the industry.
James Dice: [00:09:38] I agree.
Alex Grace: [00:09:38] And what you see consistently is where those programs came out of, are utilities that already incentivized a retrocommissioning program. They were used to doing that. They had the process down, and then they said, okay, now we're going to change that RCx incentive to an MBCx incentive, and we extend it from maybe two months to a year, but that's it. So like those that already had that model, because they already had a list of vendors that were approved, it was sort of an easy way to bring in the monitoring based conditioning idea. Whereas there are, I think, really shining examples that've taken a different approach, like NYSERDA's Real Time Energy Management is, I think, the most forward-thinking for a number of reasons. Mainly because they're incentivizing for five years.
So that's a whole different animal. But yeah, I do think that comes out of... It used to be pretty consistently that where those programs have appeared have been where they had a history of funding retrocommissioning, and then they just kind of barely shifted the model.
James Dice: [00:10:33] Yeah. And that just seems to be, to me, misunderstanding the opportunity, and kind of leads, like you said, to hurt rather than help. Cool. So let's continue on with these kind of sales-focused questions.
Um, so one thing I'm doing right now is creating a standard RFP for anyone who's procuring analytics software. And one of the challenges with that is creating some way to compare software like this FDD software in an apples-to-apples sort of way. So what are your thoughts on how that can best be done, and is it even possible at this point?
Alex Grace: [00:11:15] Yeah, it's a great question. It's definitely a challenge. I think it is possible. I think it's really about thinking beyond the initial... I have analytics up and running, or I have some basic fault detection up and running, but what is this gonna look like at scale, and what is this going to look like three years, five years from now? Meaning, what is the maintenance on the fault detection and diagnostics?
So I think something that's often not maybe given as much attention to as it needs to, is that buildings are dynamic. They're changing all the time, and the FDD needs to evolve with those operational changes. So you set up a bunch of static rules that were customized to a building, and then you change your sequence or you replace and air handler a year later. Who maintains that code set?
So, you know, are you getting costs associated with just standing up some initial FDD and hoping it works today? Well, that FDD may be producing a lot more false positives a year from now if you're not thinking about how that code set evolves and needs to be maintained. That's one point.
It is a challenge. You know, I think we love seeing people test multiple solutions for that reason. It just, it helps us because sometimes things can be abstract until you've had your hands in it, you know? So we've had clients-
James Dice: [00:12:29] Sorry, what was that? Test multiple solutions?
Alex Grace: [00:12:31] Yeah. So if you're looking at multiple FDD products, you know, you have a couple of buildings. Or you know, the most fun cases we've had is where someone actually has a historian that they're able to serve data to multiple FDD vendors and see what happens. I mean, that's a great way, but there's very few people that can do that. And you know, I don't recommend spending a year trying to figure out how you abstract your data in order to do that. You know, you want to get started and that's really the bottom line, right? Get started. You're going to learn a lot, you gotta move forward. And I think that requires leadership to really have that fail fast mentality. You know, I talk a lot about the University of Iowa because they've really done that well, you know, from the beginning, they had this mentality that we're going to test a solution, we're not going to be tied to it. We're going to learn a lot. And, and having that, making sure that your organization is a learning organization is, I think fundamental here. Yeah, so apples to apples is tricky, I'd say the majority of our biggest, and best clients have already done other types of fault detection before because they just get the complexity. They get that it's not as simple as just tagging everything and expecting things to work, that you have to really understand the operation, and that false positive is the Achilles heel of this industry, you know? So the ability to effectively prioritize and avoid false positives is critical.
And I do think that there's ways to write requirements that ensure that you're getting, for example, energy calculations accurate. And that rules are aware of other rules. And I know when you had Nick on, you guys talked about mass customization, so to have a custom solution that's going to degrade the second that you have it up and running, but you have something that is constantly evolving and code sets that are shared and parameterized in this mass customization approach. And another way, I like to say that just that the rules are aware of each other. So if you're implementing a bunch of rules, you know, I think that's key.
And I love the conversation you sparked on LinkedIn, James, around like, where's the IP? Is the IP in the rules themselves? And I just think that's a really important point from my perspective, that, or from the KGS standpoint, that, you know, identifying that a valve is leaking or a damper or stuck is not that challenging. There's no IP in the math associated with finding a leaky valve. The IP is how do you structure those code sets in umpteen different ways and hundreds of different scenarios with different types of mechanical configurations and sequence of operations, and how exactly is that economizer controlled that's controlling the damper? And what about dehumidification and all these things that you run into as you know from your background and your experience, right?
James Dice: [00:15:17] Yeah. And for everyone who hasn't seen that conversation, I basically asked all my contacts on LinkedIn, you know, what do you guys think about intellectual property when it comes to fault detection rules? And I'll put that conversation in the show notes. The responses were pretty fascinating to me.
So first of all, I want to key in on intellectual property because I think it relates to a conversation that you and I have had in the past, which is: where along the value chain can you commoditize this and where can't you?
Alex Grace: [00:15:50] Yeah. That is interesting. We hear a lot of conversation in the industry around fault detection focused on tagging. Tagging's important. Normalization's important. You know, Nick spoke more eloquently than I can on all the different things he's involved with related to the ASHRAE BACnet Committee and establishing industry standards that we want to be really heavily involved with as we think it's incredibly important to push things forward.
That being said, you know, for us, tagging is a starting point. It's not an ending point. I think a lot of people talk about it as an endpoint, and frankly, if you have very simple rules, it is an endpoint, right? I know what the data is, and therefore I'm running a full linear logic to say, you know, is something off here? Has it deviated?
You know, we sort of see that type of quote unquote fault detection as really alarming or advanced alarming potentially, but to get to the level of diagnostics, you have to understand a lot more than just, what are the points? You have to understand where equipment is. You have to understand where thermal elements are exactly in an air stream.
So not just understanding the points, but what is the order of those points in that air stream in an air handler? And then what are the sequences of operations that may or may not be reflected in the modes or the set points or the parameters that are visible from the points. They may be just written in a document somewhere, right?
So, and then having code sets that can dynamically determine based on the best available information. So that's a key thing that's a little bit harder to understand and requires some explanation, but it means that you want to have a situation where your rules don't just break because it's missing a variable.
And the way Clockworks diagnostics work is that it always looks for the best available information, meaning, okay, if this point's here, I'll use it. If this point is not here, can I use this other point or can I approximate with these two points or reference a inputed piece of metadata from a written sequence of operation to enter it?
So a really simple example would be, do I have a flow station? Yes, I have a CFM point. I'm going to use that in my calculation. No, there's no flow station. Okay. Reference fan speed and reference the rate of flow of the unit to approximate flow at any given point in time. Really simple example, but there's hundreds and hundreds of those.
James Dice: [00:18:13] Got it. Yeah, that's a great example of a place where intellectual property applies, versus the actual rule that you see when you're seeing the results of the fault detection that that might not be, I mean, cause those are pretty well understood by the industry at this point. Yeah. Thanks for taking us through that.
Um, you mentioned, a potential case study that you say you like to talk about, but I actually haven't heard the story of the University of Iowa. Could you kind of give us, you know, what's happened there?
Alex Grace: [00:18:45] Sure, yeah. I think I can because a lot of it's public, won an award from the Smart Energy Analytics Campaign, and they've won a number of awards. And Don Guckert, AVP of Facilities there has given a lot of public conferences, so I think it's... You know, we're always sensitive to making sure that we can speak there. And I should say that client is through a partner of ours Schneider Electric. So, that's important.
But just in terms of what Don speaks about in the industry that I think is really, really important is the way that he articulates institutional knowledge and shared knowledge, that I think is a really interesting point. So you hear a lot in our industry about institutional knowledge. You know, we're really into, rightly so, that technician has been here for 30 years and he walks into a chiller plant and based on a sound, he knows what's wrong, right?
And Don I hope he doesn't mind me stealing this quote from him where he says, let me tell you an example of institutional knowledge. He gives example of a technician who would walk around with pieces of wood. He knew exactly the right size piece of wood to jam in the damper to be able to fix a problem. And he goes, that's institutional knowledge. He's like, we don't need institutional knowledge. What we need is shared knowledge. And I just think that's a really powerful illustration of what we're talking about.
And so what is shared knowledge? It means that you have a single source of truth. It means that you have, in this case, a diagnostics platform, where the data doesn't lie. It's accurate. It's tuned to your operation and tuned to your sequences, and that a mechanical technician, a controls technician, an energy engineer, a sustainability professional, a design engineer on the capital planning side, they all have access to that same information based on what they need to do their job. And I think overall, as an industry, that's a really useful concept.
The reality is institutional knowledge is retiring. That's the reality that we're all facing. Right? And the next five years is more and more of that in 10 years more and more of that. So it's not really an either or, it's how do you replace institutional knowledge with shared knowledge and what are the organizational benefits associated with that? That I think is really powerful.
The other thing that's powerful, I think, from their story is just having many different stakeholders involved very early in the process. You know, I think something we may have talked about before, James, that oftentimes I think FDD gets siloed within a particular use case, and it's something Nick touched on a little bit as well. Right? You know, is it within commissioning? Is it within energy manager who's doing energy projects, or is it that shared knowledge tool that is spanning these different use cases? And it's also fine to get started with one use case, right? You don't want to boil the ocean. Totally fine to get started with one and then be aware of how you expand.
But for those that are higher up in an organization and thinking about this from a leadership perspective, mapping out your stakeholders, which is not a mystery. It doesn't require a whole, you know, consulting engagement. It's your controls technicians, it's your maintenance technicians, it's your commissioning agents or third party commissioning folks, and those on the design or capital planning team side. And how do you bring them together to derive value from a platform like Clockworks?
James Dice: [00:21:58] Got it. Okay, cool. Yeah, thanks for sharing that. It sounds like a really cool project. So a couple of last questions around sales. I'm just wondering from the perspective of you, you know, spreading the messages of analytics over the last however many years, many years, right? Um, what's it been like watching these buzzwords come into our industry? So, I'll start with IoT. I mean, when I think about Clockworks, you guys have been putting things on the internet for long before IoT was hot. So what's that been like for you and crafting a message around that?
Alex Grace: [00:22:34] Yeah. That is interesting. I mean, we're a pretty conservative company, so our marketing goes about as far as, you know, us doing podcasts with you, James. Right? So, so from that perspective, you know, we've always been a little bit skeptical of the hype cycle, and we try to not get too deep into it, right? We don't just shift our messaging immediately that, okay, now everything's IoT, or now everything's AI or whatever, right?
Um, but there's a lot to dive into there. So, you know, I think early on when IoT was coming out... So, okay. The whole conversation on IT OT convergence is totally fascinating. It's a 10 year old or 15 year old discussion and it's still totally relevant. Right? So I think early on when you saw the, the big IT players coming out, you know, I saw a presentation from Cisco, this is a number of years ago now, but to me it sounded like they'd never heard of a building automation system. In the way that they were talking about IoT, you know, it was like, we're gonna have these sensors, and they're going to talk to the centralized location, and you're going to be able to control things. I was like, um, excuse me. Right?
So that being said, obviously there's a lot of great IoT use cases that are out there. There's a lot of really exciting stuff. I mean, I know you talked about InfiSense, and we arere watching LoRaWAN, and we're watching that whole world. And there's a lot of really cool use cases that are out there, don't get me wrong. But I think my organization... I guess my reference point a lot of time with the types of clients that we work with, which those that have big facilities portfolios. They have large facilities teams. They have a lot of infrastructure. Before you start chasing IoT and new buzzwords, let's start using the data you have. Like you have already made millions and millions of dollars of investments in sensing technology, and that data is incredibly underutilized. So there's a massive opportunity to say, look, let's start with what you have. And then absolutely as there are more opportunities to bring in IoT, that's great.
But now I'm also talking from a little bit of an HVAC-centric perspective. You know, there's a lot of really cool use cases out there in janitorial or in other types of areas where, you know, having sensing that wasn't there before, it can be understandably a game changer.
James Dice: [00:24:49] Got it. Yeah. Cool. So how about machine learning? So I want you to talk about this from the perspective of how you're messaging it, but then also how you guys are thinking about it internally and whether to add these capabilities to your tools.
Alex Grace: [00:25:03] Sure. Yeah. Thanks for the question. I think there's a lot of, frankly, market education that is needed around that. I mean, we still sometimes see RFPs, for example, come out and say like, you know, do you have AI? And it's, you know, there's a lot there. So let me start. So, machine learning, we see a lot of potential in certain techniques and we're using it today for the main area of improving onboarding.
So we use natural language processing and machine learning to help onboard buildings more effectively. We have an incredible dataset. We have about 260,000 mechanical assets, meaning an air handler, a chiller, a boiler, a VAV, a pump in Clockworks today. And all of the data associated with those 260,000 equipment from about 380 million square feet, has all been totally normalized to a standard information model.
So based on that information that we have, when we ingest data from a new building, we have tools that will take that points list and help figure out what those points are based on everything we've seen before. And that is not just, have I seen that exact point name before? And match it to a database. It's obviously more complicated than that. And, our chief scientist can speak a lot more to the actual techniques we're using there, but what's exciting about it is it will continue to evolve. The more data you collect, the smarter you get, and we have a really significant dataset that will only continue to grow. So we see a lot of potential there for machine learning.
On the diagnostic side , it gets interesting. Right? So for those that are a little less familiar, you know, fundamentally machine learning, you're going to train a model on a certain dataset, and then you're going to identify deviations from that model, right?
We have yet to see examples of machine learning for diagnostics that get us closer to an accurate result than what we are already doing, which the academic term will be an expert system, which is a form of AI, an early form of AI, and hierarchal rule-based FDD, which is a fancy way of saying the rules are aware of each other, and getting to the root cause of the problem as much as possible.
So in other words, if you find a deviation-, if the air handler was in normal operation, and you trained a model on that normal operation, and now the air handler is in something different, it's deviated from that, okay. But can you say exactly what's wrong as a result of that deviation in a way that is more accurate than saying, I can see the temperatures on either side of this coil and the valve position is closed and therefore that valve is leaking? Or we're simultaneous heating and cooling, and I'm referencing my dehumidification sequence, and I can see that, yes, we should be dehumidifying, but we're dehumidifying too much. Like that degree of specificity and then adding the loads, looking at the excess kBtus, turning that into dollars, adding a comfort ranking or maintenance range to it, which is what we do today with our expert system. We're not seeing anything yet within machine learning that lets us do it better than that.
That being said, we have a lot of academic relationships. We monitor that world really closely. And one of the nice things about our technology stack and software as a service is that if something compelling comes out academia, out of the national labs, or out of our own research, we will build those ML techniques into Clockworks, but so far today, we're not seeing something there that is more compelling. But we are seeing a lot of excitement and have a lot of enthusiasm for what we're doing and continuing to evolve on the the onboarding and the natural language processing side. Long answer to a question. I hope that addresses-.
James Dice: [00:28:35] No, that's exactly what I was looking for. I love getting your perspective because you're intelligent about the technical side, but you're also grounded in the way that you're speaking to customers all the time on the business development side. So, and I've played that role for many years as well, and so I recognize someone else that I like to talk to in that regard.
Let's transition a little bit over to channels. I know it's something that you've been personally working on quite a bit. It sounds like you've been focused lately on getting Clockworks in the hands of service providers. So not exactly building owners, but people that serve building owners, so mechanical service providers, controls service providers, others. But let's start there. So what are the things that these firms should be thinking about when it comes to approaching this whole digitization phase of everyone's life cycle right now? I mean, we're all being digitized, whether we like it or not, so what are some of the things that these types of firms should be thinking about?
Alex Grace: [00:29:35] Yeah, absolutely. Cool. So there's a lot there. So all we have to do is Google digital transformation and we see just the massive amounts of information there, right? So every business is facing this to some degree or another. And it's really, I think it's important to reference that broad trend, right? Like this is a shift that's happening in the world, and the facilities industry, the mechanical industry, the controls industry are not, you know, are not insulated, right? Like, like this, this transformation has been happening, is happening and is accelerating. I think that's another important thing that the pace of change is accelerating.
So, I mean, you know, of course we have this fantastic global relationship with Schneider Electric, where we are a global technology partner and have been since, um, formally since 2012. And that's a really, really important relationship for us. And they've been really thought leaders from very early on about how their service strategy will change globally. And we're really proud to be a part of that transformation and a part of that story. And we've learned a lot over the years from that.
And then, you know, we're, we're working with a number of other types of partners today, both in the controls world, in the mechanical world, facilities management companies as well, is another important one. So just to talk about services for a second , broadly speaking, the old approach of: I have a checklist; I'm going to roll a truck to a customer site; I'm going to go down that PM list; and I grease bearings this frequently, and I change the belts that frequently, and filters on that timeframe. Or I'm time-and-materials, break-fix type focused, and I show up on site every once in a while, maybe one day a month, one day a quarter and my guy tells me what his biggest headaches are, maybe it's written on a piece of paper. And then I go and try and take care of those things. That model is fundamentally transforming and changing, and I really see that there's, there is a carrot and there's a stick, right?
So there's the carrot being the opportunity. The opportunity is for business leaders running the mechanical service business or control service business to realize that I have a major opportunity to differentiate the way we're doing things and to show customer value and to elevate the value of my services to a different level in the organization, that I'm not just thought of as someone that fixes things when there's a major problem, but I'm also thought of someone that helps me strategically understand the operation of my facilities, the risk factors that I'm facing, and where I can drive energy reduction, performance and long term value when it comes to capital planning and asset management as well. So there's a lot, there, but that transformation is occurring.
So, you know, that's the opportunity. And what I mean by that, okay, when you're up for renewal on that service agreement-, you have a $20,000 service agreement with the customer. And now you're up for renewal, and there's a new finance guy, and he says, well, what are we getting for this $20,000? Well, here's my checklist. You know, I checked a ton of boxes. We did all these things for you that was really important based on ASHRAE guideline, blah, blah, blah, blah. Okay. Does that guy know what you're talking about? Does he understand the value of that $20,000 agreement?
Now let's talk about a digital approach. So how does this transform? How does this change? I no longer only roll trucks reactively or based on a checklist. I now, when that guy goes onsite, when my technician appears, he knows exactly what the top priorities are in this building based on energy, what it's costing my customer, based on comfort, which is arguably the biggest cost, your employees themselves. Okay, during the normal COVID times when there are people in those buildings. And then thirdly, the mechanical severity of the problem. Can I fix something that otherwise is going to wake someone up at two in the morning and they're going to have to come into the building and have a massive headache on their hands? Right?
So now the technician shows up with these priorities. He knows what's in scope, based on his current agreement. He knows what's out of scope that he can now propose a work on. So obviously the massive business value there from the vendor side of being able to have a consistent ability to propose ROI-justified work, and from the customer's side, massively more valuable because they know that that technician is focused on the biggest risk factors for them, the biggest energy factors, and the biggest comfort factors for them.
And now let's flash forward to that same conversation a year in to your agreement. You're up for renewal with that finance person. You now have a dynamic dashboard or a static report as you wish saying, what did that $20,000 buy you? Here's exactly what we saved you last year as a result of doing predictive maintenance and continuous commissioning. Here's exactly the impact we had on your facility in terms of quantified maintenance value and quantified comfort value.
James Dice: [00:34:16] Wow. Yeah, that's powerful. And I've certainly seen, as someone that has been providing these types of software for a long time, I've seen what happens when you're providing, say an FDD solution for instance, and then you're asking the person that is on one of these service contracts to fix what comes out of the FDD, and there's these huge disconnects there between like, I'm not here to fix that. I'm actually here to do this other thing. And it's usually, like you said, find an issue, right? I'm not here to fix an issue. I'm here to survey. I'm gonna walk around and look at gauges. I'm gonna check filters.
And I think what you're saying here is that the new sort of service and sort of O&M process is we're actually focused on results.
Alex Grace: [00:35:04] Right.
James Dice: [00:35:04] With the time we're spending. Right?
Alex Grace: [00:35:07] Absolutely. In this line of, you know, let's spend less time identifying and more time fixing. And you know to be clear, obviously we're not replacing all the PMs, right? I mean, you're still gonna need to grease bearings. You're still gonna need to change filters. You're still gonna need to look at certain things. Absolutely. But there's a certain percent of that agreement and that labor that can be shifted to higher value activities. And that's the percent we target.
And we actually go to a pretty deep level on that. So we'll go through an exercise with partners. We'll look at their task list. How do you do maintenance today? And what are your typical agreements? And are you standardized? So that's the first step, right?
You have organizations that made acquisitions, and this branch does something this way, and this branch does something completely differently. You have others that have consolidated their approach to services and say, we have a three tier program. We have good, better, best, or you know, a T and M block hours, a PM, and full service.
So whatever that is, if you've already done that standardization, you're already a leg up. If you haven't, incorporating FDD into your plan could help you do that standardization. So we really go to this level of understanding: what are your service tiers that you have today? We go through a task by task basis. We assign hours to that with our partner. So yeah, if I don't need to check that thing because analytics is looking for it every single day because I have diagnostics running, that's going to save me 15 minutes per activity, cause I don't have to do that thing. And then you add up all those hours and you've come out with, Hey, it looks like I could actually offset a third of the hours here.
And you know, operations leaders can be a little bit threatened by that. I just want to make that comment. If they don't understand the big picture, which is that their guys are going to get way more work out of this than they ever had before. It's just that that work is going to be higher value.
The other thing is what can you do remotely, you know, really figuring out what can you do remotely versus what needs to be on site. And I think now in the age of, you know, coronavirus that's probably even more important than ever, but well after coronavirus has left us, that's still going to be really important for a business model perspective.
James Dice: [00:37:07] Great. Yeah, I agree. You mentioned the carrot and the stick or the carrot or the stick. So we've kind of painted a picture towards those that are going after the carrot. What about the people that are getting hit with the stick? What are the organizations that aren't quite keeping up with this digital transformation? What are you seeing happening to them?
Alex Grace: [00:37:28] Sure. I think that the organizations that are able to pick their head out of the noise and the day-to-day, I've got a million things on my to do list, and able to look into the future are seeing that there is a threat here. So there's the opportunity to really have better client outcomes, better client relationships, and derive additional pull-through revenue through their service organizations and provide just a better service for their customers.
And there's a stick of, if we don't do this, in five years is our organization going to be relevant? And that's a bold statement to make. But I really think that's how leaders of these companies are thinking and need to be thinking if they're not. Because imagine for a second that you're going to renew that service agreement, and you're doing things the exact same way you've always done them, and you're doing T & M or block hours, and they started getting calls from people that are offering the same value service from a similar price point with way more. With an analytics driven approach and reporting on that value to the customer at the end of the year . Where the customer value is just that much more explicit.
And then one thing I do want to circle back on, and the carrot that we're really focused on is I think a lot of companies are still thinking about FDD or a tool like Clockworks as another tool in the toolbox. So, you know, we talk about shifting from reactive to proactive maintenance. We also talk about shifting from reactive to proactive business model, right? So I want to bring that up, which is, you know what I mean by that is if you're an organization that is thinking about FDD just because one of your more sophisticated customers is asking about it and asking for your help, you might want to take a step back and say, do we want to continue to be reactive to customers asking us about this, or do we want to have a strategy?
So we are not, frankly, not as interested in partners that want to resell Clockworks as another tool. You know, okay, the customer asks about it, you know, here's something I can maybe talk to them about. We're much more interested in business model transformation, because that's what this industry, that's where things are going. And that requires a whole different range of discussions about understanding the partner organization, how they operationally deliver maintenance today, and the structures I was talking about: good, better, best; block hours, PMs, full service. And it also requires understanding how they go to market. You know, who's selling those service contracts? How are they incentivized? You know, are they thinking about building up a remote operation center, a network operation center? So it's the business side. It's the sales side. It's the operational delivery.
And for that reason, sometimes people are surprised, like they'll reach out to us with an RFP and say, Hey, we're evaluating five FDD vendors and they're sometimes surprised about how much we ask about their organizations and their businesses. And the reason is that we need to qualify our partners as well as partners qualify us, because we go deep. We go deep in understanding them and consulting on that business model transformation, because we think that's where the most value is on both sides.
James Dice: [00:40:27] Fascinating. So let's go through some of these business models.
So I have a list of about five different potential partners and channels of yours. So starting with these service contractors, what does the service contract of the digital world look like? So how are they packaging the technology in with their services into this new package?
Alex Grace: [00:40:49] Yeah, absolutely. So part of that equation I was saying before where you determine are there costs that can be offset? That's part of the equation. So let's say for example, you have, this is just hypothetical because it does depend and you need to dive into the details here to really be able to stay this, but as a hypothetical situation, you have a $20,000 service agreement. Let's say you're able to remove $5,000 in labor. And by remove, I mean offset. So you're gonna replace that with some diagnostic monitoring and you're gonna shift some of those hours to remote operation center approach.
And it doesn't mean you're starting with, you know, NASA space command. It means that you have a guy in the office, who is one of your most experienced technicians or an energy engineer, someone who really understands data and buildings and systems, and that person, by the way, can now be 10 times more valuable because you don't have them on customer sites one-to-one. You have them one-to-many, and they're now able to guide, with the diagnostic results, less experienced technicians. And there's a lot of operational value there as well. So you take that best guy, you put them in the office and you say, okay, we're gonna take an analytics-driven approach.
So you've got that $20,000 agreement, let's say $5,000 gets offset by a combination of hours internally from your experienced guy in the remote operation center and the cost of Clockworks to look at things. And then there's going to be some delta there that you're going to have to upsell. But the upsell is now way more value add, right? You're able to now talk about prioritization, about scoring that building in terms of metrics.
So every day, what are your top priorities relevant to energy, comfort, and maintenance, and how do you drive actions and quantify the results of the fix? You're now able to do that and show examples of that. And we're even seeing some organizations that are actually subsidizing that transformation. Meaning, are you willing to actually eat a little cost because you know that you're going to get way more project pull through work? And because you see it as a strategic priority to have this transformation occur for all the reasons we talked about related to the carrot and the stick.
So some combination there, and it's gonna look different for different orientations. You absolutely don't need to be eating it, but some are getting aggressive and saying, you know, this is important. We're just going to do this because we know it's gonna produce fruit and we know it's going to make us more profitable and expand our service business over time.
James Dice: [00:43:13] Fascinating. Okay, let's move on from the service contractors and go into commissioning firms and talk about the inherent one-time commissioning effort or one-time retrocommissioning effort business model versus the ongoing monitoring based commissioning model that analytics offers and presents as an opportunity. So can you talk a little bit about the carrot and stick for those guys and how the ones that are on top of digital transformation are transitioning their business models?
Alex Grace: [00:43:47] Yeah, sure thing. And this is a cool one cause I know you've got a lot of background here. Well, you have a lot of background in a lot of areas, but this'll be a fun one to talk about.
So the commissioning world has honestly been fascinating to me. And it's fascinating because I've always sort of wondered, why don't we have more commissioning firms that are using Clockworks and providing an ongoing service? And I saw, actually one proactive commissioning company that we're talking to right now turned me on to this report from the Building Commissioning Association, the BCA, that was really confirming for me. And it showed a study, and it was survey of their members, and the result of that survey was that over 90% of monitoring based commissioning projects are not going past 12 months. So they're ending after one year.
James Dice: [00:44:34] That is a staggering statistic for me to think about.
Alex Grace: [00:44:39] Yeah. Yeah. I thought that was incredibly fascinating. So I think the problems that I'm seeing the commissioning world are really business model based. They're obviously not technically based. I mean commissioning firms are incredibly well positioned in terms of skill sets of their people to drive an ongoing commissioning process with fault detection and diagnostics at its core. I mean incredibly well positioned.
And don't get me wrong, we're seeing plenty of companies that are using, I would say, toolkits in their commissioning toolset. So, okay. Rather than, you know, downloading data from the control system. First of all, there's still a lot of companies doing this: downloading data from the control system, spending three weeks messing around with it in Excel back in the office to come up with a report. Like, if you're still doing that, obviously that needs to change. I mean, that's just crazy.
But , and then there are companies that are changing that by going to, I'd say, a toolkit approach. So they're using an FDD tool where they can write their custom algorithms, because it helps them do their commissioning process. But they're still fundamentally project-based businesses, meaning: they do a project; they end it, whether it's instead of three months, now it's 12 months, it's still very limited; and then they move on. And the mentality is, well, the facilities teams are gonna pick it up from there. There is a big gap in the big market opportunity, I think, for commissioning firms to stay involved with their clients longer and to be adding more value on a continuous basis as it relates to operations and maintenance, by running FDD longer and continuing that process. I think it's not happening because it's not how project teams are based, and it's not how companies are structured or incentivized. So I think they're missing an opportunity.
And here's a different perspective to take. I firmly believe if you're a commissioning firm listening to this right now, you're working at a commissioning firm, five years from now, your best customers are all going to have fault detection and diagnostics. All of them. Do you want to be driving that and be a part of that solution or not? I really think that's what it comes down to. And I know that's a strong statement, but I'm going to make it here, talking with James Dice. I think it's really true.
You know, so if you're in this mentality of, well, we just don't know why customers aren't renewing and we're doing it for 12 months. Then you're missing something, because we have clients doing this for 10 years. And fundamentally, when organizations we work with start using Clockworks, they don't stop. Because how do you go back to reactive maintenance after you have a tool that tells you every single day where your priorities are?
And there's stuff that comes out of Clockworks that is more complicated, right? There's the leaking valves and the stuck dampers and the broken sensors, but there's also the: we could be staging our cooling towers more efficiently than we are, I need someone to rewrite the sequence; or our loops are all under loaded, Clockworks is telling me that the delta T in our chilled water loops is performing consistently bad, and it gives me a range of recommendations on how to fix that, but I still need an engineer now figure out what is the solution? What is the exact sequence I need to rewrite here to improve?
So I think sometimes people miss that, that there's a lot of engineering work that comes out of FDD if the FDD is actually doing the diagnostic piece and not just fault detection. So yeah, there's an opportunity there and there's the stick, which is all your clients are going to have this and how is your work going to change if they all do and you weren't a part of driving that solution?
James Dice: [00:48:09] Right. And I came from this world, so the way I see it is these companies have existed for a while. You know, commissioning and retrocommissioning have been, well-ingrained processes for a little bit. You know, there's still a lot of construction projects that don't use commissioning, which is a whole different conversation. There are still a lot of buildings that haven't been retrocommissioned, which is a totally different conversation. But in general, those two practices are pretty well -accepted as projects that are worth people's time, worth people's money.
But those firms that provide those services have a business model that is just like a construction project. It's a one-time event, right? So how are you seeing the business models for those companies transform to accommodate monitoring essentially?
Alex Grace: [00:48:56] Yeah, that's a great point. Okay, so practically speaking, you're doing commissioning, you're doing post-occupancy commissioning. Maybe you have FDD in there already to some degree. Um, this is where I think Clockworks does differentiate from some of the other tools because we do provide this O&M focus and we serve a lot of different stakeholder groups. But let's just say more generally, it doesn't have to be specific to Clockworks, you've got it running, and rather than going away to the next project, you now propose or you've proposed from the beginning that, how about we have a one day a month? So this agent that knows your system, has been testing things, has gone through the commissioning process, we're going to keep him on one day a month. We're going to check in on the information that the fault detection and diagnostics has produced. We're going to make sure you're not missing anything, and we're going to make sure any system retuning that's done continues to happen.
And I think, you know, at a minimum that's gonna make a lot of sense to the customer post-occupancy in terms of identifying things during the warranty period and holding vendors accountable. I'm talking about new construction now or major retrofit, but through that process, you know, you've got to make sure you're getting to the O&M folks, first of all, that you're not just seeing yourself as like, Oh, I'm just giving them an O&M manual and walking away, that you're really providing value of helping them see their issues, and that, you know, just because it's a new building certainly those of us who have done commissioning know that the problems don't stop. Right? So even after you've tuned things up, things are going to keep coming back.
It's also a difference in terms of portfolios, I would say. So that's the one-off building example, but there's also, if you're working with a portfolio and you're, for example, retrocommissioning a handful of buildings a year, moving your way through the portfolio, maybe you take a different approach.
If you look at the cost of that deep dive retrocommissioning every single year only touching a handful of the portfolio, versus what if I monitored everything every year, and then that same man hours I was spending doing my deep dive functional performance testing, I've shifted to letting analytics drive where to focus. So rather than testing a bunch of systems to figure out where the problems are, how about diagnostics tells me exactly where the problems are, and I spend those man hours helping engineer solutions across the portfolio? Because the cost benefit on that is dramatic. And it may not be even more costly from a customer perspective to do that.
Or the other thing to keep in mind is you don't have to run FDD on everything. You might decide, my customer has a 50 building portfolio. I'm going to monitor every chiller plant, every boiler plant, and every air handling unit across the portfolio. And then focus my retrocommissioning efforts really targeted on the buildings that are outliers based on that data analytics. And it's not just, you know, EUI cost per square foot. It's a much deeper dataset where I'm now seeing, maybe I'm looking at kW per tons, maybe I'm looking at kW per CFMs on ventilation system efficiency, and maybe I'm able to go even deeper and see exactly where sequences could be improved, where they're not being followed, where the resets that I engineered five years ago are no longer in place cause they've been over-ridden, and et cetera, et cetera, right.
James Dice: [00:52:03] Got it. Yeah. And something that I've-, so when I've designed these types of business models and these types of services in the past, I found this book called Subscribed to be really helpful for transitioning your proposals, transitioning all of your internal processes to go from this one-time, like percent complete invoice-based business model of these commissioning firms, to a subscription business model. You're basically turning your business into a subscription at that point. And these firms need to stop thinking about this in terms of a one time event. And I could get all up on my soapbox about that, but I'll stop.
Alex Grace: [00:52:41] I like that, James. And just one more thing there. I mean, I think that's a great point. I gotta check out that book; that sounds really useful. I just want to mention on the positive side, we have, we do have commissioning firms that are working with Clockworks. Uh, shout out to WSP in Boston, for example. But you know, organizations that are really differentiating what they're doing and they're going into the customer's sites.
So from a sales perspective, you're a commissioning firm, the power of going into a customer site in a competitive situation-, obviously, if the customer is totally race-to-the-bottom and focused, then yeah, this is maybe not what I'm talking about. But if you have a more proactive, forward-thinking customer, and you're able to walk into that meeting as we've seen some of our partners do and say, we do commissioning differently. Here's how we do it. You know, that's very powerful. And that we're going to have an ongoing O&M focus if you want that, you know . I think there's a big opportunity there to really differentiate your service in the market.
The other thing I want to just quickly mention is macroeconomic trends. So because I think that's been a factor lately also. Because commissioning firms are busy, like they're getting a ton of work, right? No one's, no one has guys sitting around at home starving for time.
So when construction is booming, when the economy is doing really well, and you've got new construction and commissioning projects left and right, why take the time to really take a step back and look at some of these things we're talking about? You know, I get that.
Well, the current reality of where we are with the current economic situation related to coronavirus, but I'll say more broadly, just economic cycles, macroeconomic cycles in general, there's going to be periods of construction slow down. There's going to be periods where we were taking a step back. Now is a phenomenal time to take a step back and think about that and where things are going. And I think that we're seeing that because I'm seeing inbound messages lately and I think a lot more actually than we had been before. And I think that's part of it is people just have a little more time to pick their head above the water from all the work they'd had and say, okay, where are things going? Where is the industry moving? How can I position our company to be where we need to be, you know, now and five years now?
James Dice: [00:54:44] That's fascinating. On this last, so we're kind of continuing on these different types of business models. So commissioning firms are often, sometimes part of, you mentioned WSP, part of larger MEP design firms. So how do all of these lessons apply to those types of businesses and what are you seeing for them right now?
Alex Grace: [00:55:04] Yeah, that's really cool. I think a couple things. One, every design engineer I've ever talked to is dying to know how their designs actually are performing in the field, because you never get that feedback as a design engineer, right? I mean, you have a model, you put the best, all of your knowledge into this design, but you don't really know how the operation is going.
You know, is it being operated the way it was designed or not? So that feedback is obviously very powerful , and we've seen that everyone wants that. You know, there's a, there's a message there when it comes to new construction from an owner perspective also, around you're spending millions, hundreds of millions of dollars, whatever the case might be, on this building and on these designs and on the install of these systems. For pennies on the dollar, do you want to ensure that you're getting what you paid for? Right? So , yeah, I think that is an interesting area. If commissioning firms that often are tied to larger MEP firms, if they're able to create that feedback loop to the design, that's incredibly powerful. And a tool like Clockworks, you know, would certainly allow you to do that, but that is something interesting to think about.
The other thing I'll mention is just around M&V specs. So we've seen M&V specs come out that, you know, I want a submeter every last thing in order to just see how my designs are performing or at least like compare that to the energy model. Don't get me wrong, that's fantastic. We love sub-metering. Put all the data acquisition out there that you can. But at the end of the day, you're still limited by, okay, performance is different than model. Now what? Like where's the problem? Right.
So I think, I think FDD needs to be thought of more in M&V than it is today. And there's a big opportunity there where it's to say, look, that's great you have a lots of sub-metering. That's great you're able to compare performance against your energy model. But can you figure out where the problem is specifically? Can you pinpoint it? And FDD can tell you exactly what's going on. Well, you know, these two things were reversed and your actuation is totally backwards in terms of how you're economizing or whatever.
James Dice: [00:57:14] Yeah, that's exactly right. And so we've covered most of what I was hoping to get through today. And so we're almost running out of time here, but I did want to hit in on something that we haven't really covered that I think is an important detail. So for all of these different service providers and for owners and their O&M teams, what happens when you have better diagnostics? In terms of when that guy needs to go turn that wrench or take that two by four out of the damper or fix that actuator.
What I get a sense from you guys is when you have better diagnostics, you have a better information model that produces those diagnostics, and that gives you more information that you need when you're going to go to fix that fault, whatever it is. So can you talk a little bit about what you're seeing on that side of things?
Alex Grace: [00:58:04] Sure. Just in terms of having better data to fix problems, yeah, absolutely. So, you know, we produce a range of possible causes, so we identify a problem, we prioritize it. Prioritization is key, obviously. We, you know, in 10 years plus of doing this, still haven't met a facilities team that's looking for another to-do list. Right.
But that being said, people do need priorities , and understanding what the real impacts of these problems are and how do I sort them accordingly? And then it's really important to have those possible causes where I can understand, is it-, first of all, that's critical to know: who do I send this issue to? Like is it a controls programming issue? Is it a mechanical issue? Or maybe I can't tell from the data. Maybe the diagnostic can say it's one of the two. In other words, the valve is leaking. Okay, let's go look at it. Either the valve is leaking or the temperature sensor was broken. That's always going to be a reality from the data.
So but what is the order of events? So first I'm going to go look at that mechanically speaking and see if the valve is in fact leaking. If it's not, then I send it over to the control shop. Now it might be the same person, it might be someone different, depending on the organization, they might be sending it to a vendor. Hey, Johnson Controls or Siemens or Honeywell or Schneider or whoever is coming tomorrow . Let me add this to the list of things, of broken sensors. Or just simply being able to search. So we have, you know, the search ability, so I can just pull up every faulty sensor in the building. And send that to a vendor.
So the possible causes things is really key. I think we're going to continue to evolve there. So when I mentioned like rules being aware of other rules, that's really important for that to be able to pinpoint the problem and not send people on wild goose chases. Again, false positives are the Achilles heel. You've got to have a structure. That one, allows you to drastically reduce false positives and two, when they come up, you can quickly change them. That's also key, because you're never gonna get 100% , and the process is important there.
But I, I'm really excited about areas that are kind of in our R & D bucket today as well, around what we call meta analysis, I'll just bring up quick. So we already have rules aware of other rules, structured into a hierarchy. The next level is you have diagnostics aware of other diagnostics, and so basically you keep building up layers of analytics capability, and we are continuing to evolve that, where you will get to a point where we have diagnostics on top of diagnostics and that means longer term degradations of equipment and patterns. That means more in deeper level of systems analysis. Um, again, we do a lot of that today, but there's there room to grow and it's some really cool R & D there.
James Dice: [01:00:28] Yeah, that's all awesome. I was hoping you would say that.
The one thing I was thinking about though, that I don't know that the service providers are thinking about is that when you have this common single source of truth that you were talking about earlier in the context of institutional versus shared knowledge, so that shared knowledge from my perspective, what I think is valuable for all of those service providers is that when you have, say, your fan motor is not working. Whatever the problem is that comes out of Clockworks or any FDD, that service provider is going to know the model number, the size of the motor, everything to do with that in that information model, before they leave the shop, right? So they're already knowing, like you said earlier, what their exact task list is when they get to the building. But the point would be that they already are ready to perform the exact task. And I think there's this, I don't know if that message is getting out there as far as the process efficiencies.
So it's not just about labor savings, it's also about everything to do with the process of performing these ongoing services.
Alex Grace: [01:01:36] Yeah, that's a great point. Sorry, I think I missed your question a little bit, so-
James Dice: [01:01:38] No, you were great.
Alex Grace: [01:01:40] So I think you're really touching on a key thing, which is the asset management side. So what are the datasets and the data silos that need to be combined, right? So you now have this fault detection information and a history of fault detection. So how many times has that valve leaked on that air handler, and how many excess kBtus has it cost you in the last three years, every time you've fixed it?
You also have the asset service history. So what are the lists of fixes and PMs and reactive break fix, and what are the labor and material costs associated with that? So being able to combine the datasets from an asset management tool or a CMS tool with fault detection, we see as really critical, and we are having that conversation with some service organizations that are larger, that are really thinking about the value of that data.
I mean, everyone's talking now from a business perspective, the values of data, data is the new oil, et cetera, et cetera. Right? And those that are thinking about like, what is the strategy there? So you have all this asset history, you have all this fault detection history. What are you providing?
So specifically to what you're saying, the process efficiency, absolutely. Not just showing on site with a prioritized list, but knowing what parts to bring. I mean, you can't underestimate that if you're rolling a truck, particularly if you're in a situation where you have to travel some distance. You know, if you have to send a guy an hour away or two hours away, even more critical that you know what the problem is and that you couldn't have fixed it remotely.
So imagine you fix the things that you can fix remotely, cause you have remote access and you're providing, for example, control service. And the things that you can't fix remotely, you know the exact sensor you need to bring or the exact actuator that needs to be replaced or whenever the tools are that are gonna allow you to pinpoint the fix that's needed.
James Dice: [01:03:18] Totally. Yeah. And so I can't help myself. I also have to ask about, so from the perspective of the owner, a lot of what I hear from owners in this regard is, is it going to integrate with my work order system? And I hear two perspectives, it seems like from the marketplace in this regard. One is, Oh yeah, we do it and it's happening all the time and it's an easy integration. And then on the other hand, I hear, I heard this in a meeting last week: No one's doing it. It's too difficult to do. It's too hard to write integrations for every CMMS or computerized maintenance management system. So what are you guys seeing for your clients and what have you guys built up as far as integrations with work orders from faults?
Alex Grace: [01:04:00] Yeah, that's super interesting. I'd say the truth is somewhere in the middle, from my perspective. So I'll say that off the bat, we are doing it successfully. We have work order integrations in place with multiple customers. And it's also true that just because you have Maximo, everyone's implementation of Maximo is different.
What fields specifically you're filling out and how those fields need to be mapped to combine an output from diagnostics from Clockworks with that system needs to be defined. So it's both-and. So fundamentally, if your work order system has an API, we can talk to it. If you don't, please upgrade to a system that does, you need to be thinking about that, right?
So, for example, if you have like a legacy platform, a lot of people are in process of upgrading from, for example, FAMIS local to FAMIS online or whatever the equivalent is , and there's a ton of tools out there. So we can do the integration, absolutely. There is a certain element of custom software development associated, because again, you have to define those fields and make sure that you're mapping that process appropriately. But we also have, once you do it once for a Maximo or AiM or FAMIS, that is replicable for us. But then it's the fields and the details that still need to be defined.
So basically what we're seeing is it's not something you do out of the gate because there are costs. You may not do it for the first building that you do, but if you've reached a certain degree of scale, everyone we know both in terms of our enterprise accounts and partner organizations are thinking about this, that they need to do it. Some of them already have. There's great process efficiency gains there that are useful. And also having those combined datasets, you know, as you pointed out around asset management can be powerful and is important to think about.
James Dice: [01:05:47] Cool. Yeah, I've got a sense we could talk all day, but I think we gotta wrap things up here. I want to thank you for spreading your knowledge around to the industry and letting me learn from you guys this week and last week. So I really appreciate it. Thanks for coming on the show.
Alex Grace: [01:06:02] Thanks so much, James. It's great to talk with you. Really enjoyed it.
James Dice: [01:06:05] All right, friends. Thanks for listening to this episode of the Nexus podcast. For more episodes like this and to get the weekly Nexus newsletter, please subscribe at nexus.substack.com. You can find show notes from this conversation there as well. As always, please reach out on LinkedIn with any thoughts on this episode. I'd love to hear from you. Have a great day.
Happy Thursday!
Welcome to the first Nexus deep dive exclusively for Nexus Pro members. It’s an honor to have you here. And let’s kick this off right: use the comments to introduce yourself and let us know what you thought of the episode and deep dive.
This deep dive is a follow up to my recent conversation with Alex Grace, VP of Business Development at KGS Buildings. I thoroughly enjoyed this conversation and want to share my takeaways and the full transcript with you below.
In case you missed it in your inbox, you can find the audio or video here:
Nexus site | Apple Podcasts | Spotify | YouTube | Add to other podcast apps
Enjoy!
Disclaimer: James is a researcher at the National Renewable Energy Laboratory (NREL). All opinions expressed via Nexus emails, podcasts, or the website belong solely to James. No resources from NREL are used to support Nexus. NREL does not endorse or support any aspect of Nexus.
Here’s an outline of today’s deep dive:
After these two conversations with KGS Buildings in back to back podcasts, it felt like we stumbled upon a potential new interview strategy: talk to different people from the same company and go deep into different facets of their business. In this case, Nick provided what sets KGS apart and where the analytics industry is headed. Alex provided the biz dev lens and a view of analytics-based digital transformation for all types of service providers.
My #1 takeaway from Alex’s wisdom happened when he got on his soapbox about commissioning providers. Here’s his powerful quote:
And here's a different perspective to take. I firmly believe if you're a commissioning firm listening to this right now, five years from now, your best customers are all going to have fault detection and diagnostics. All of them. Do you want to be driving that and be a part of that solution or not? I really think that's what it comes down to. And I know that's a strong statement, but I'm going to make it here, talking with James Dice. I think it's really true.
I obviously agree, but this has made me think…
From a building owner’s perspective, there needs to be a separate strategy from the service provider’s strategy. The owner needs to have a digitization approach that selects the best platform and then drives each service provider to provide a better and more digitized service using that single platform. I’ve seen what happens when multiple service providers serving a single building owner bring their preferred platform, and they’re not necessarily bringing all the other service providers into the solution with them. Chaos can ensue.
What do you think?
Alex Grace: [00:29:35] So just to talk about services for a second, broadly speaking, the old approach of: I have a checklist; I'm going to roll a truck to a customer site; I'm going to go down that PM list; and I grease bearings this frequently, and I change the belts that frequently, and filters on that timeframe. Or I'm time-and-materials, break-fix type focused, and I show up on site every once in a while, maybe one day a month, one day a quarter and my guy tells me what his biggest headaches are, maybe it's written on a piece of paper. And then I go and try and take care of those things. That model is fundamentally transforming and changing, and I really see that there's, there is a carrot and there's a stick, right?
So there's the carrot being the opportunity. The opportunity is for business leaders running the mechanical service business or control service business to realize that I have a major opportunity to differentiate the way we're doing things and to show customer value and to elevate the value of my services to a different level in the organization, that I'm not just thought of as someone that fixes things when there's a major problem, but I'm also thought of someone that helps me strategically understand the operation of my facilities, the risk factors that I'm facing, and where I can drive energy reduction, performance and long term value when it comes to capital planning and asset management as well. So there's a lot, there, but that transformation is occurring.
So, you know, that's the opportunity. And what I mean by that, okay, when you're up for renewal on that service agreement-, you have a $20,000 service agreement with the customer. And now you're up for renewal, and there's a new finance guy, and he says, well, what are we getting for this $20,000? Well, here's my checklist. You know, I checked a ton of boxes. We did all these things for you that was really important based on ASHRAE guideline, blah, blah, blah, blah. Okay. Does that guy know what you're talking about? Does he understand the value of that $20,000 agreement?
Now let's talk about a digital approach. So how does this transform? How does this change? I no longer only roll trucks reactively or based on a checklist. I now, when that guy goes onsite, when my technician appears, he knows exactly what the top priorities are in this building based on energy, what it's costing my customer, based on comfort, which is arguably the biggest cost, your employees themselves. Okay, during the normal COVID times when there are people in those buildings. And then thirdly, the mechanical severity of the problem. Can I fix something that otherwise is going to wake someone up at two in the morning and they're going to have to come into the building and have a massive headache on their hands? Right?
So now the technician shows up with these priorities. He knows what's in scope, based on his current agreement. He knows what's out of scope that he can now propose a work on. So obviously the massive business value there from the vendor side of being able to have a consistent ability to propose ROI-justified work, and from the customer's side, massively more valuable because they know that that technician is focused on the biggest risk factors for them, the biggest energy factors, and the biggest comfort factors for them.
And now let's flash forward to that same conversation a year into your agreement. You're up for renewal with that finance person. You now have a dynamic dashboard or a static report as you wish saying, what did that $20,000 buy you? Here's exactly what we saved you last year as a result of doing predictive maintenance and continuous commissioning. Here's exactly the impact we had on your facility in terms of quantified maintenance value and quantified comfort value.
Let's spend less time identifying and more time fixing. And you know to be clear, obviously we're not replacing all the PMs, right? I mean, you're still gonna need to grease bearings. You're still gonna need to change filters. You're still gonna need to look at certain things. Absolutely. But there's a certain percent of that agreement and that labor that can be shifted to higher value activities. And that's the percent we target.
And we actually go to a pretty deep level on that. So we'll go through an exercise with partners. We'll look at their task list. How do you do maintenance today? And what are your typical agreements? And are you standardized? So that's the first step, right?
You have organizations that made acquisitions, and this branch does something this way, and this branch does something completely differently. You have others that have consolidated their approach to services and say, we have a three tier program. We have good, better, best, or you know, a T and M block hours, a PM, and full service.
So whatever that is, if you've already done that standardization, you're already a leg up. If you haven't, incorporating FDD into your plan could help you do that standardization. So we really go to this level of understanding: what are your service tiers that you have today? We go through a task by task basis. We assign hours to that with our partner. So yeah, if I don't need to check that thing because analytics is looking for it every single day because I have diagnostics running, that's going to save me 15 minutes per activity, cause I don't have to do that thing. And then you add up all those hours and you've come out with, Hey, it looks like I could actually offset a third of the hours here.
And you know, operations leaders can be a little bit threatened by that. I just want to make that comment. If they don't understand the big picture, which is that their guys are going to get way more work out of this than they ever had before. It's just that that work is going to be higher value.
The other thing is what can you do remotely, you know, really figuring out what can you do remotely versus what needs to be on site. And I think now in the age of, you know, coronavirus that's probably even more important than ever, but well after coronavirus has left us, that's still going to be really important for a business model perspective.
James Dice: [00:37:07] Great. Yeah, I agree. You mentioned the carrot and the stick or the carrot or the stick. So we've kind of painted a picture of those that are going after the carrot. What about the people that are getting hit with the stick? What are the organizations that aren't quite keeping up with this digital transformation? What are you seeing happening to them?
Alex Grace: [00:37:28] Sure. I think that the organizations that are able to pick their head out of the noise and the day-to-day, I've got a million things on my to do list, and able to look into the future are seeing that there is a threat here. So there's the opportunity to really have better client outcomes, better client relationships, and derive additional pull-through revenue through their service organizations and provide just a better service for their customers.
And there's a stick of, if we don't do this, in five years is our organization going to be relevant? And that's a bold statement to make. But I really think that's how leaders of these companies are thinking and need to be thinking if they're not. Because imagine for a second that you're going to renew that service agreement, and you're doing things the exact same way you've always done them, and you're doing T & M or block hours, and they started getting calls from people that are offering the same value service from a similar price point with way more. With an analytics driven approach and reporting on that value to the customer at the end of the year . Where the customer value is just that much more explicit.
And then one thing I do want to circle back on, and the carrot that we're really focused on is I think a lot of companies are still thinking about FDD or a tool like Clockworks as another tool in the toolbox. So, you know, we talk about shifting from reactive to proactive maintenance. We also talk about shifting from reactive to proactive business model, right? So I want to bring that up, which is, you know what I mean by that is if you're an organization that is thinking about FDD just because one of your more sophisticated customers is asking about it and asking for your help, you might want to take a step back and say, do we want to continue to be reactive to customers asking us about this, or do we want to have a strategy?
So we are not, frankly, not as interested in partners that want to resell Clockworks as another tool. You know, okay, the customer asks about it, you know, here's something I can maybe talk to them about. We're much more interested in business model transformation, because that's what this industry, that's where things are going. And that requires a whole different range of discussions about understanding the partner organization, how they operationally deliver maintenance today, and the structures I was talking about: good, better, best; block hours, PMs, full service. And it also requires understanding how they go to market. You know, who's selling those service contracts? How are they incentivized? You know, are they thinking about building up a remote operation center, a network operation center? So it's the business side. It's the sales side. It's the operational delivery.
And for that reason, sometimes people are surprised, like they'll reach out to us with an RFP and say, Hey, we're evaluating five FDD vendors and they're sometimes surprised about how much we ask about their organizations and their businesses. And the reason is that we need to qualify our partners as well as partners qualify us, because we go deep. We go deep in understanding them and consulting on that business model transformation because we think that's where the most value is on both sides.
James Dice: [00:40:27] Fascinating. So let's go through some of these business models.
So I have a list of about five different potential partners and channels of yours. So starting with these service contractors, what does the service contract of the digital world look like? So how are they packaging the technology in with their services into this new package?
Alex Grace: [00:40:49] Yeah, absolutely. So part of that equation I was saying before where you determine are there costs that can be offset? That's part of the equation. So let's say for example, you have, this is just hypothetical because it does depend and you need to dive into the details here to really be able to stay this, but as a hypothetical situation, you have a $20,000 service agreement. Let's say you're able to remove $5,000 in labor. And by remove, I mean offset. So you're gonna replace that with some diagnostic monitoring and you're gonna shift some of those hours to remote operation center approach.
And it doesn't mean you're starting with, you know, NASA space command. It means that you have a guy in the office, who is one of your most experienced technicians or an energy engineer, someone who really understands data and buildings and systems, and that person, by the way, can now be 10 times more valuable because you don't have them on customer sites one-to-one. You have them one-to-many, and they're now able to guide, with the diagnostic results, less experienced technicians. And there's a lot of operational value there as well. So you take that best guy, you put them in the office and you say, okay, we're gonna take an analytics-driven approach.
So you've got that $20,000 agreement, let's say $5,000 gets offset by a combination of hours internally from your experienced guy in the remote operation center and the cost of Clockworks to look at things. And then there's going to be some delta there that you're going to have to upsell. But the upsell is now way more value add, right? You're able to now talk about prioritization, about scoring that building in terms of metrics.
So every day, what are your top priorities relevant to energy, comfort, and maintenance, and how do you drive actions and quantify the results of the fix? You're now able to do that and show examples of that. And we're even seeing some organizations that are actually subsidizing that transformation. Meaning, are you willing to actually eat a little cost because you know that you're going to get way more project pull through work? And because you see it as a strategic priority to have this transformation occur for all the reasons we talked about related to the carrot and the stick.
So some combination there, and it's gonna look different for different orientations. You absolutely don't need to be eating it, but some are getting aggressive and saying, you know, this is important. We're just going to do this because we know it's gonna produce fruit and we know it's going to make us more profitable and expand our service business over time.
James Dice: [01:00:28] The one thing I was thinking about though, that I don't know that the service providers are thinking about is that when you have this common single source of truth that you were talking about earlier in the context of institutional versus shared knowledge, so that shared knowledge from my perspective, what I think is valuable for all of those service providers is that when you have, say, your fan motor is not working. Whatever the problem is that comes out of Clockworks or any FDD, that service provider is going to know the model number, the size of the motor, everything to do with that in that information model, before they leave the shop, right? So they're already knowing, like you said earlier, what their exact task list is when they get to the building. But the point would be that they already are ready to perform the exact task. And I think there's this, I don't know if that message is getting out there as far as the process efficiencies.
So it's not just about labor savings, it's also about everything to do with the process of performing these ongoing services.
Alex Grace: [01:01:40] So I think you're really touching on a key thing, which is the asset management side. So what are the datasets and the data silos that need to be combined, right? So you now have this fault detection information and a history of fault detection. So how many times has that valve leaked on that air handler, and how many excess kBtus has it cost you in the last three years, every time you've fixed it?
You also have the asset service history. So what are the lists of fixes and PMs and reactive break fix, and what are the labor and material costs associated with that? So being able to combine the datasets from an asset management tool or a CMS tool with fault detection, we see as really critical, and we are having that conversation with some service organizations that are larger, that are really thinking about the value of that data.
I mean, everyone's talking now from a business perspective, the values of data, data is the new oil, et cetera, et cetera. Right? And those that are thinking about like, what is the strategy there? So you have all this asset history, you have all this fault detection history. What are you providing?
So specifically to what you're saying, the process efficiency, absolutely. Not just showing on site with a prioritized list, but knowing what parts to bring. I mean, you can't underestimate that if you're rolling a truck, particularly if you're in a situation where you have to travel some distance. You know, if you have to send a guy an hour away or two hours away, even more critical that you know what the problem is and that you couldn't have fixed it remotely.
So imagine you fix the things that you can fix remotely, cause you have remote access and you're providing, for example, control service. And the things that you can't fix remotely, you know the exact sensor you need to bring or the exact actuator that needs to be replaced or whenever the tools are that are gonna allow you to pinpoint the fix that's needed.
James Dice: [00:05:12] How about from an energy efficiency standpoint? So when I think about an energy efficiency project, you're doing some sort of study, right? And then you're installing some sort of ECMs, energy conservation measures, and then you're realizing savings after that. But what I've found is with building owners that have a hard time wrapping their heads around the Software as a Service fee, because it doesn't produce savings in and of itself usually, and so how are you guys helping to justify those ongoing fees? Because as we know, they're vital for maintaining the savings that come out of energy efficiency projects, but they're not necessarily required to implement energy efficiency projects themselves.
Alex Grace: [00:05:56] Right. Yeah. There's a couple interesting pieces there that you touched on for sure. So, some changes that we're seeing... I mean, you're absolutely right. There's a fundamental market education piece there, which is about energy drift, which those in the energy auditing, commissioning world know well, right? That just because you implemented a beautiful reset schedule, it doesn't mean it's still being followed a day later, much less six months later, much less two years later, right?
So probably how we've been from an M&V perspective, calculating savings and thinking that low cost, no cost measures actually just persist, is probably not right. That's one point.
And then secondly, you know, I think that on the maintenance side of the house, this really isn't as much of a concern, but on the energy project side of the house, it is more. And really the obvious point is valves don't stop leaking; dampers don't stop sticking; sensors don't stop failing. So everyone, again, on the O&M and the maintenance world knows that very well, but from an energy projects perspective, you may have the idea that you go and implement something, you squeeze the savings out, and you walk away. And it's just simply not the case.
So you know, it's, it's why our angle is really... Our clients and our partners are looking to drive predictive maintenance outcomes. They're looking to implement continuous commissioning processes for all the reasons that we know: because it improves operational efficiency, because their people can be focused on what matters most rather than just reacting and going down a checklist. They're zeroed in on the issues that are costing them the most amount of money, that are having the greatest maintenance impact or having the biggest impact to building occupants, the building comfort. So you know, that's why they're doing it. And then the natural byproduct of that process is energy saved.
And I like to talk about it and frame it in that way: that continuous energy savings is there, of course it is, but for a lot of our users, it's not the primary driver, although it may be very important to get that initial budget approval. And we have a client that will give a speech on stage and say, you know, we didn't care about energy. We're completely focused on maintenance optimization. We're having trouble retaining staff. You know, people are retiring. There's a major labor shortage in the industry, as we all know. We need to make sure that our people can really be laser-focused on having the biggest impact and getting more from our existing teams.
And then you'll say, oh, but yeah, we also saved eight hundred thousand dollars. So it's like the energy savings is there, whether you're focused on it or not.
Here’s a summary of their program:
James Dice: [00:43:13] Fascinating. Okay, let's move on from the service contractors and go into commissioning firms and talk about the inherent one-time commissioning effort or one-time retrocommissioning effort business model versus the ongoing monitoring based commissioning model that analytics offers and presents as an opportunity. So can you talk a little bit about the carrot and stick for those guys and how the ones that are on top of digital transformation are transitioning their business models?
Alex Grace: [00:43:47] Yeah, sure thing. And this is a cool one cause I know you've got a lot of background here. Well, you have a lot of background in a lot of areas, but this'll be a fun one to talk about.
So the commissioning world has honestly been fascinating to me. And it's fascinating because I've always sort of wondered, why don't we have more commissioning firms that are using Clockworks and providing an ongoing service? And I saw, actually one proactive commissioning company that we're talking to right now turned me on to this report from the Building Commissioning Association, the BCA, that was really confirming for me. And it showed a study, and it was survey of their members, and the result of that survey was that over 90% of monitoring based commissioning projects are not going past 12 months. So they're ending after one year.
James Dice: [00:44:34] That is a staggering statistic for me to think about.
Alex Grace: [00:44:39] Yeah. Yeah. I thought that was incredibly fascinating. So I think the problems that I'm seeing the commissioning world are really business model based. They're obviously not technically based. I mean commissioning firms are incredibly well positioned in terms of skill sets of their people to drive an ongoing commissioning process with fault detection and diagnostics at its core. I mean incredibly well positioned.
And don't get me wrong, we're seeing plenty of companies that are using, I would say, toolkits in their commissioning toolset. So, okay. Rather than, you know, downloading data from the control system. First of all, there's still a lot of companies doing this: downloading data from the control system, spending three weeks messing around with it in Excel back in the office to come up with a report. Like, if you're still doing that, obviously that needs to change. I mean, that's just crazy.
But , and then there are companies that are changing that by going to, I'd say, a toolkit approach. So they're using an FDD tool where they can write their custom algorithms, because it helps them do their commissioning process. But they're still fundamentally project-based businesses, meaning: they do a project; they end it, whether it's instead of three months, now it's 12 months, it's still very limited; and then they move on. And the mentality is, well, the facilities teams are gonna pick it up from there. There is a big gap in the big market opportunity, I think, for commissioning firms to stay involved with their clients longer and to be adding more value on a continuous basis as it relates to operations and maintenance, by running FDD longer and continuing that process. I think it's not happening because it's not how project teams are based, and it's not how companies are structured or incentivized. So I think they're missing an opportunity.
And here's a different perspective to take. I firmly believe if you're a commissioning firm listening to this right now, you're working at a commissioning firm, five years from now, your best customers are all going to have fault detection and diagnostics. All of them. Do you want to be driving that and be a part of that solution or not? I really think that's what it comes down to. And I know that's a strong statement, but I'm going to make it here, talking with James Dice. I think it's really true.
You know, so if you're in this mentality of, well, we just don't know why customers aren't renewing and we're doing it for 12 months. Then you're missing something, because we have clients doing this for 10 years. And fundamentally, when organizations we work with start using Clockworks, they don't stop. Because how do you go back to reactive maintenance after you have a tool that tells you every single day where your priorities are?
And there's stuff that comes out of Clockworks that is more complicated, right? There's the leaking valves and the stuck dampers and the broken sensors, but there's also the: we could be staging our cooling towers more efficiently than we are, I need someone to rewrite the sequence; or our loops are all under loaded, Clockworks is telling me that the delta T in our chilled water loops is performing consistently bad, and it gives me a range of recommendations on how to fix that, but I still need an engineer now figure out what is the solution? What is the exact sequence I need to rewrite here to improve?
So I think sometimes people miss that, that there's a lot of engineering work that comes out of FDD if the FDD is actually doing the diagnostic piece and not just fault detection. So yeah, there's an opportunity there and there's the stick, which is all your clients are going to have this and how is your work going to change if they all do and you weren't a part of driving that solution?
James Dice: [00:48:09] Right. And I came from this world, so the way I see it is these companies have existed for a while. You know, commissioning and retrocommissioning have been, well-ingrained processes for a little bit. You know, there's still a lot of construction projects that don't use commissioning, which is a whole different conversation. There are still a lot of buildings that haven't been retrocommissioned, which is a totally different conversation. But in general, those two practices are pretty well -accepted as projects that are worth people's time, worth people's money.
But those firms that provide those services have a business model that is just like a construction project. It's a one-time event, right? So how are you seeing the business models for those companies transform to accommodate monitoring essentially?
Alex Grace: [00:48:56] Yeah, that's a great point. Okay, so practically speaking, you're doing commissioning, you're doing post-occupancy commissioning. Maybe you have FDD in there already to some degree. Um, this is where I think Clockworks does differentiate from some of the other tools because we do provide this O&M focus and we serve a lot of different stakeholder groups. But let's just say more generally, it doesn't have to be specific to Clockworks, you've got it running, and rather than going away to the next project, you now propose or you've proposed from the beginning that, how about we have a one day a month? So this agent that knows your system, has been testing things, has gone through the commissioning process, we're going to keep him on one day a month. We're going to check in on the information that the fault detection and diagnostics has produced. We're going to make sure you're not missing anything, and we're going to make sure any system retuning that's done continues to happen.
And I think, you know, at a minimum that's gonna make a lot of sense to the customer post-occupancy in terms of identifying things during the warranty period and holding vendors accountable. I'm talking about new construction now or major retrofit, but through that process, you know, you've got to make sure you're getting to the O&M folks, first of all, that you're not just seeing yourself as like, Oh, I'm just giving them an O&M manual and walking away, that you're really providing value of helping them see their issues, and that, you know, just because it's a new building certainly those of us who have done commissioning know that the problems don't stop. Right? So even after you've tuned things up, things are going to keep coming back.
Alex Grace: [00:48:56] If you're working with a portfolio and you're, for example, retrocommissioning a handful of buildings a year, moving your way through the portfolio, maybe you take a different approach.
If you look at the cost of that deep dive retrocommissioning every single year only touching a handful of the portfolio, versus what if I monitored everything every year, and then that same man hours I was spending doing my deep dive functional performance testing, I've shifted to letting analytics drive where to focus. So rather than testing a bunch of systems to figure out where the problems are, how about diagnostics tells me exactly where the problems are, and I spend those man hours helping engineer solutions across the portfolio? Because the cost benefit on that is dramatic. And it may not be even more costly from a customer perspective to do that.
Or the other thing to keep in mind is you don't have to run FDD on everything. You might decide, my customer has a 50 building portfolio. I'm going to monitor every chiller plant, every boiler plant, and every air handling unit across the portfolio. And then focus my retrocommissioning efforts really targeted on the buildings that are outliers based on that data analytics. And it's not just, you know, EUI cost per square foot. It's a much deeper dataset where I'm now seeing, maybe I'm looking at kW per tons, maybe I'm looking at kW per CFMs on ventilation system efficiency, and maybe I'm able to go even deeper and see exactly where sequences could be improved, where they're not being followed, where the resets that I engineered five years ago are no longer in place cause they've been over-ridden, and et cetera, et cetera, right.
I just want to mention on the positive side, we have, we do have commissioning firms that are working with Clockworks. Uh, shout out to WSP in Boston, for example. But you know, organizations that are really differentiating what they're doing and they're going into the customer's sites.
So from a sales perspective, you're a commissioning firm, the power of going into a customer site in a competitive situation-, obviously, if the customer is totally race-to-the-bottom and focused, then yeah, this is maybe not what I'm talking about. But if you have a more proactive, forward-thinking customer, and you're able to walk into that meeting as we've seen some of our partners do and say, we do commissioning differently. Here's how we do it. You know, that's very powerful. And that we're going to have an ongoing O&M focus if you want that, you know . I think there's a big opportunity there to really differentiate your service in the market.
The other thing I want to just quickly mention is macroeconomic trends. So because I think that's been a factor lately also. Because commissioning firms are busy, like they're getting a ton of work, right? No one's, no one has guys sitting around at home starving for time.
So when construction is booming, when the economy is doing really well, and you've got new construction and commissioning projects left and right, why take the time to really take a step back and look at some of these things we're talking about? You know, I get that.
Well, the current reality of where we are with the current economic situation related to coronavirus, but I'll say more broadly, just economic cycles, macroeconomic cycles in general, there's going to be periods of construction slow down. There's going to be periods where we were taking a step back. Now is a phenomenal time to take a step back and think about that and where things are going. And I think that we're seeing that because I'm seeing inbound messages lately and I think a lot more actually than we had been before. And I think that's part of it is people just have a little more time to pick their head above the water from all the work they'd had and say, okay, where are things going? Where is the industry moving? How can I position our company to be where we need to be, you know, now and five years now?
James Dice: [00:54:44] That's fascinating. On this last, so we're kind of continuing on these different types of business models. So commissioning firms are often, sometimes part of, you mentioned WSP, part of larger MEP design firms. So how do all of these lessons apply to those types of businesses and what are you seeing for them right now?
Alex Grace: [00:55:04] Yeah, that's really cool. I think a couple things. One, every design engineer I've ever talked to is dying to know how their designs actually are performing in the field, because you never get that feedback as a design engineer, right? I mean, you have a model, you put the best, all of your knowledge into this design, but you don't really know how the operation is going.
You know, is it being operated the way it was designed or not? So that feedback is obviously very powerful , and we've seen that everyone wants that. You know, there's a, there's a message there when it comes to new construction from an owner perspective also, around you're spending millions, hundreds of millions of dollars, whatever the case might be, on this building and on these designs and on the install of these systems. For pennies on the dollar, do you want to ensure that you're getting what you paid for? Right? So , yeah, I think that is an interesting area. If commissioning firms that often are tied to larger MEP firms, if they're able to create that feedback loop to the design, that's incredibly powerful. And a tool like Clockworks, you know, would certainly allow you to do that, but that is something interesting to think about.
The other thing I'll mention is just around M&V specs. So we've seen M&V specs come out that, you know, I want a submeter every last thing in order to just see how my designs are performing or at least like compare that to the energy model. Don't get me wrong, that's fantastic. We love sub-metering. Put all the data acquisition out there that you can. But at the end of the day, you're still limited by, okay, performance is different than model. Now what? Like where's the problem? Right.
So I think, I think FDD needs to be thought of more in M&V than it is today. And there's a big opportunity there where it's to say, look, that's great you have a lots of sub-metering. That's great you're able to compare performance against your energy model. But can you figure out where the problem is specifically? Can you pinpoint it? And FDD can tell you exactly what's going on. Well, you know, these two things were reversed and your actuation is totally backwards in terms of how you're economizing or whatever.
James Dice: [00:10:33] Um, so one thing I'm doing right now is creating a standard RFP for anyone who's procuring analytics software. And one of the challenges with that is creating some way to compare software like this FDD software in an apples-to-apples sort of way. So what are your thoughts on how that can best be done, and is it even possible at this point?
Alex Grace: [00:11:15] Yeah, it's a great question. It's definitely a challenge. I think it is possible. I think it's really about thinking beyond the initial... I have analytics up and running, or I have some basic fault detection up and running, but what is this gonna look like at scale, and what is this going to look like three years, five years from now? Meaning, what is the maintenance on the fault detection and diagnostics?
So I think something that's often not maybe given as much attention to as it needs to, is that buildings are dynamic. They're changing all the time, and the FDD needs to evolve with those operational changes. So you set up a bunch of static rules that were customized to a building, and then you change your sequence or you replace and air handler a year later. Who maintains that code set?
So, you know, are you getting costs associated with just standing up some initial FDD and hoping it works today? Well, that FDD may be producing a lot more false positives a year from now if you're not thinking about how that code set evolves and needs to be maintained. That's one point.
It is a challenge. You know, I think we love seeing people test multiple solutions for that reason. It just, it helps us because sometimes things can be abstract until you've had your hands in it, you know? So we've had clients-
James Dice: [00:12:29] Sorry, what was that? Test multiple solutions?
Alex Grace: [00:12:31] Yeah. So if you're looking at multiple FDD products, you know, you have a couple of buildings. Or you know, the most fun cases we've had is where someone actually has a historian that they're able to serve data to multiple FDD vendors and see what happens. I mean, that's a great way, but there's very few people that can do that. And you know, I don't recommend spending a year trying to figure out how you abstract your data in order to do that. You know, you want to get started and that's really the bottom line, right? Get started. You're going to learn a lot, you gotta move forward. And I think that requires leadership to really have that fail fast mentality. You know, I talk a lot about the University of Iowa because they've really done that well, you know, from the beginning, they had this mentality that we're going to test a solution, we're not going to be tied to it. We're going to learn a lot. And, and having that, making sure that your organization is a learning organization is, I think fundamental here. Yeah, so apples to apples is tricky, I'd say the majority of our biggest, and best clients have already done other types of fault detection before because they just get the complexity. They get that it's not as simple as just tagging everything and expecting things to work, that you have to really understand the operation, and that false positive is the Achilles heel of this industry, you know? So the ability to effectively prioritize and avoid false positives is critical.
And I do think that there's ways to write requirements that ensure that you're getting, for example, energy calculations accurate. And that rules are aware of other rules. And I know when you had Nick on, you guys talked about mass customization, so to have a custom solution that's going to degrade the second that you have it up and running, but you have something that is constantly evolving and code sets that are shared and parameterized in this mass customization approach. And another way, I like to say that just that the rules are aware of each other. So if you're implementing a bunch of rules, you know, I think that's key.
And I love the conversation you sparked on LinkedIn, James, around like, where's the IP? Is the IP in the rules themselves? And I just think that's a really important point from my perspective, that, or from the KGS standpoint, that, you know, identifying that a valve is leaking or a damper or stuck is not that challenging. There's no IP in the math associated with finding a leaky valve. The IP is how do you structure those code sets in umpteen different ways and hundreds of different scenarios with different types of mechanical configurations and sequence of operations, and how exactly is that economizer controlled that's controlling the damper? And what about dehumidification and all these things that you run into as you know from your background and your experience, right?
We hear a lot of conversation in the industry around fault detection focused on tagging. Tagging's important. Normalization's important. You know, Nick spoke more eloquently than I can on all the different things he's involved with related to the ASHRAE BACnet Committee and establishing industry standards that we want to be really heavily involved with as we think it's incredibly important to push things forward.
That being said, you know, for us, tagging is a starting point. It's not an ending point. I think a lot of people talk about it as an endpoint, and frankly, if you have very simple rules, it is an endpoint, right? I know what the data is, and therefore I'm running a full linear logic to say, you know, is something off here? Has it deviated?
You know, we sort of see that type of quote unquote fault detection as really alarming or advanced alarming potentially, but to get to the level of diagnostics, you have to understand a lot more than just, what are the points? You have to understand where equipment is. You have to understand where thermal elements are exactly in an air stream.
So not just understanding the points, but what is the order of those points in that air stream in an air handler? And then what are the sequences of operations that may or may not be reflected in the modes or the set points or the parameters that are visible from the points. They may be just written in a document somewhere, right?
So, and then having code sets that can dynamically determine based on the best available information. So that's a key thing that's a little bit harder to understand and requires some explanation, but it means that you want to have a situation where your rules don't just break because it's missing a variable.
And the way Clockworks diagnostics work is that it always looks for the best available information, meaning, okay, if this point's here, I'll use it. If this point is not here, can I use this other point or can I approximate with these two points or reference a inputed piece of metadata from a written sequence of operation to enter it?
So a really simple example would be, do I have a flow station? Yes, I have a CFM point. I'm going to use that in my calculation. No, there's no flow station. Okay. Reference fan speed and reference the rate of flow of the unit to approximate flow at any given point in time. Really simple example, but there's hundreds and hundreds of those.
Alex Grace: [00:18:45] And Don I hope he doesn't mind me stealing this quote from him where he says, let me tell you an example of institutional knowledge. He gives example of a technician who would walk around with pieces of wood. He knew exactly the right size piece of wood to jam in the damper to be able to fix a problem. And he goes, that's institutional knowledge. He's like, we don't need institutional knowledge. What we need is shared knowledge. And I just think that's a really powerful illustration of what we're talking about.
And so what is shared knowledge? It means that you have a single source of truth. It means that you have, in this case, a diagnostics platform, where the data doesn't lie. It's accurate. It's tuned to your operation and tuned to your sequences, and that a mechanical technician, a controls technician, an energy engineer, a sustainability professional, a design engineer on the capital planning side, they all have access to that same information based on what they need to do their job. And I think overall, as an industry, that's a really useful concept.
The reality is institutional knowledge is retiring. That's the reality that we're all facing. Right? And the next five years is more and more of that in 10 years more and more of that. So it's not really an either or, it's how do you replace institutional knowledge with shared knowledge and what are the organizational benefits associated with that? That I think is really powerful.
The other thing that's powerful, I think, from their story is just having many different stakeholders involved very early in the process. You know, I think something we may have talked about before, James, that oftentimes I think FDD gets siloed within a particular use case, and it's something Nick touched on a little bit as well. Right? You know, is it within commissioning? Is it within energy manager who's doing energy projects, or is it that shared knowledge tool that is spanning these different use cases? And it's also fine to get started with one use case, right? You don't want to boil the ocean. Totally fine to get started with one and then be aware of how you expand.
But for those that are higher up in an organization and thinking about this from a leadership perspective, mapping out your stakeholders, which is not a mystery. It doesn't require a whole, you know, consulting engagement. It's your controls technicians, it's your maintenance technicians, it's your commissioning agents or third party commissioning folks, and those on the design or capital planning team side. And how do you bring them together to derive value from a platform like Clockworks?
James Dice: [00:21:58] Got it. Okay, cool. Yeah, thanks for sharing that. It sounds like a really cool project. So a couple of last questions around sales. I'm just wondering from the perspective of you, you know, spreading the messages of analytics over the last however many years, many years, right? Um, what's it been like watching these buzzwords come into our industry? So, I'll start with IoT. I mean, when I think about Clockworks, you guys have been putting things on the internet for long before IoT was hot. So what's that been like for you and crafting a message around that?
Alex Grace: [00:22:34] Yeah. That is interesting. I mean, we're a pretty conservative company, so our marketing goes about as far as, you know, us doing podcasts with you, James. Right? So, so from that perspective, you know, we've always been a little bit skeptical of the hype cycle, and we try to not get too deep into it, right? We don't just shift our messaging immediately that, okay, now everything's IoT, or now everything's AI or whatever, right?
Um, but there's a lot to dive into there. So, you know, I think early on when IoT was coming out... So, okay. The whole conversation on IT OT convergence is totally fascinating. It's a 10 year old or 15 year old discussion and it's still totally relevant. Right? So I think early on when you saw the, the big IT players coming out, you know, I saw a presentation from Cisco, this is a number of years ago now, but to me it sounded like they'd never heard of a building automation system. In the way that they were talking about IoT, you know, it was like, we're gonna have these sensors, and they're going to talk to the centralized location, and you're going to be able to control things. I was like, um, excuse me. Right?
So that being said, obviously there's a lot of great IoT use cases that are out there. There's a lot of really exciting stuff. I mean, I know you talked about InfiSense, and we arere watching LoRaWAN, and we're watching that whole world. And there's a lot of really cool use cases that are out there, don't get me wrong. But I think my organization... I guess my reference point a lot of time with the types of clients that we work with, which those that have big facilities portfolios. They have large facilities teams. They have a lot of infrastructure. Before you start chasing IoT and new buzzwords, let's start using the data you have. Like you have already made millions and millions of dollars of investments in sensing technology, and that data is incredibly underutilized. So there's a massive opportunity to say, look, let's start with what you have. And then absolutely as there are more opportunities to bring in IoT, that's great.
But now I'm also talking from a little bit of an HVAC-centric perspective. You know, there's a lot of really cool use cases out there in janitorial or in other types of areas where, you know, having sensing that wasn't there before, it can be understandably a game changer.
Alex Grace: [00:25:03] I think there's a lot of, frankly, market education that is needed around that. I mean, we still sometimes see RFPs, for example, come out and say like, you know, do you have AI? And it's, you know, there's a lot there. So let me start. So, machine learning, we see a lot of potential in certain techniques and we're using it today for the main area of improving onboarding.
So we use natural language processing and machine learning to help onboard buildings more effectively. We have an incredible dataset. We have about 260,000 mechanical assets, meaning an air handler, a chiller, a boiler, a VAV, a pump in Clockworks today. And all of the data associated with those 260,000 equipment from about 380 million square feet, has all been totally normalized to a standard information model.
So based on that information that we have, when we ingest data from a new building, we have tools that will take that points list and help figure out what those points are based on everything we've seen before. And that is not just, have I seen that exact point name before? And match it to a database. It's obviously more complicated than that. And, our chief scientist can speak a lot more to the actual techniques we're using there, but what's exciting about it is it will continue to evolve. The more data you collect, the smarter you get, and we have a really significant dataset that will only continue to grow. So we see a lot of potential there for machine learning.
On the diagnostic side , it gets interesting. Right? So for those that are a little less familiar, you know, fundamentally machine learning, you're going to train a model on a certain dataset, and then you're going to identify deviations from that model, right?
We have yet to see examples of machine learning for diagnostics that get us closer to an accurate result than what we are already doing, which the academic term will be an expert system, which is a form of AI, an early form of AI, and hierarchal rule-based FDD, which is a fancy way of saying the rules are aware of each other, and getting to the root cause of the problem as much as possible.
So in other words, if you find a deviation-, if the air handler was in normal operation, and you trained a model on that normal operation, and now the air handler is in something different, it's deviated from that, okay. But can you say exactly what's wrong as a result of that deviation in a way that is more accurate than saying, I can see the temperatures on either side of this coil and the valve position is closed and therefore that valve is leaking? Or we're simultaneous heating and cooling, and I'm referencing my dehumidification sequence, and I can see that, yes, we should be dehumidifying, but we're dehumidifying too much. Like that degree of specificity and then adding the loads, looking at the excess kBtus, turning that into dollars, adding a comfort ranking or maintenance range to it, which is what we do today with our expert system. We're not seeing anything yet within machine learning that lets us do it better than that.
James Dice: [01:03:18] Totally. Yeah. And so I can't help myself. I also have to ask about, so from the perspective of the owner, a lot of what I hear from owners in this regard is, is it going to integrate with my work order system? And I hear two perspectives, it seems like from the marketplace in this regard. One is, Oh yeah, we do it and it's happening all the time and it's an easy integration. And then on the other hand, I hear, I heard this in a meeting last week: No one's doing it. It's too difficult to do. It's too hard to write integrations for every CMMS or computerized maintenance management system. So what are you guys seeing for your clients and what have you guys built up as far as integrations with work orders from faults?
Alex Grace: [01:04:00] Yeah, that's super interesting. I'd say the truth is somewhere in the middle, from my perspective. So I'll say that off the bat, we are doing it successfully. We have work order integrations in place with multiple customers. And it's also true that just because you have Maximo, everyone's implementation of Maximo is different.
What fields specifically you're filling out and how those fields need to be mapped to combine an output from diagnostics from Clockworks with that system needs to be defined. So it's both-and. So fundamentally, if your work order system has an API, we can talk to it. If you don't, please upgrade to a system that does, you need to be thinking about that, right?
So, for example, if you have like a legacy platform, a lot of people are in process of upgrading from, for example, FAMIS local to FAMIS online or whatever the equivalent is , and there's a ton of tools out there. So we can do the integration, absolutely. There is a certain element of custom software development associated, because again, you have to define those fields and make sure that you're mapping that process appropriately. But we also have, once you do it once for a Maximo or AiM or FAMIS, that is replicable for us. But then it's the fields and the details that still need to be defined.
So basically what we're seeing is it's not something you do out of the gate because there are costs. You may not do it for the first building that you do, but if you've reached a certain degree of scale, everyone we know both in terms of our enterprise accounts and partner organizations are thinking about this, that they need to do it. Some of them already have. There's great process efficiency gains there that are useful. And also having those combined datasets, you know, as you pointed out around asset management can be powerful and is important to think about.
What did you think about these highlights? Let us know in the comments.
Note: transcript was created using an imperfect machine learning tool and lightly edited by a human (so you can get the gist). Please forgive errors!
James Dice: [00:00:00] Hello friends. Welcome to Nexus, the smart buildings technology podcast for smart humans. I'm your host, James Dice. If we haven't met before, I write a weekly newsletter on this same topic. It's also called Nexus. Each week I share what I've learned, my opinions, and what I'm excited about in the quickly evolving world of intelligent buildings. Readers have called Nexus the best way to stay up to date on the future of this industry without all the marketing fluff. You can check it out and subscribe at nexus.substack.com or click the link in the show notes. Since starting the Nexus newsletter, many of you have reached out to me wanting to talk shop, and we have. After a few weeks of those wonderful conversations, I realized I needed to record and share them with our growing community.
So here we are, the Nexus podcast is born. This is our chance to explore and learn with the brightest in our industry together.
One more quick note before we get to this week's episode. I'm a researcher at the National Renewable Energy Laboratory, otherwise known as NREL. All opinions expressed on this podcast belong solely to me or the guest. No resources from NREL are used to support Nexus. NREL does not endorse or support any aspect of Nexus.
Let's get to it then. Episode four is a conversation with Alex Grace, Vice President of Business Development at KGS Buildings. This episode is complementary to last week's episode with Nick Gajewski, Alex's CEO, and I think you'll like it just as much. These guys are like the dynamic duo of fault detection and diagnostics.
I picked Alex's brain on a range of topics such as: keys to selling software as a service in the buildings industry; how the leading mechanical controls and commissioning service providers are digitizing their offerings and business models; and where today's buzzwords like IoT, machine learning, and AI fit in with more traditional analytics like FDD; and much, much more.
You can find Alex online, on LinkedIn and at KGSBuildings.com both of these links can be found in the show notes on nexus.substack.com. Without further ado, please enjoy Nexus podcast episode four with Alex. Grace.
Hello, Alex, welcome to the podcast.
Alex Grace: [00:02:05] Hi James, great to be with you.
James Dice: [00:02:07] Please introduce yourself and your role, at KGS.
Alex Grace: [00:02:12] Sure thing. My name is Alex Grace. I'm the Vice President of Business Development at KGS Buildings, and been with the company about seven and a half years now.
James Dice: [00:02:21] Great. And for those who don't know, what is KGS Buildings, and what is Clockworks?
Alex Grace: [00:02:27] Sure, yeah. So KGS Buildings is a fault detection and diagnostics and performance monitoring company, and we are totally focused on FDD. So that is all that we do, and Clockworks is our product. And that is the platform to be able to consistently diagnose building issues and help service providers and end users figure out what's going on in their buildings.
James Dice: [00:02:50] Great. And, just for everyone who's listening to this and hasn't listened to the episode from last week, with Nick Gajewski, we went deep into Clockworks, and I recommend checking out that conversation. So let's jump into the business development side of things at KGS. First question, how do you think selling software as a service is different for buildings than it is in other industries?
Alex Grace: [00:03:16] Sure, yeah. That's interesting. I think in some ways the facilities world, you know, was a little late to the game of SaaS, right? Um, you know, Software as a Service disrupted a lot of industries first, I would say. So, you know, if we think about when's the last time you went and actually loaded a CD in your personal life onto your harddrive to install some software or installed anything locally? I mean, everything we do now is, is through a browser, right? And delivered in a SaaS format. I mean, there's no, it's, it's all about Microsoft 365, right? Or Adobe, or you name the product in any other realm, you know, it's pretty much all Software as a Service.
Um, so I think we were a little late to that, but I think at this point, there's a lot of comfort with that model, especially because, again, in the rest of our lives, it's sort of a foregone conclusion.
James Dice: [00:04:06] Got it. What about from the budgeting standpoint? So are building owners getting used to paying for things in this ongoing way versus... I think about a building owner. They're buying a construction project. They're buying a retrofit project. And those are two separate budgets, right? From operations to capital. So how do they think about that from the standpoint of buying software these days?
Alex Grace: [00:04:29] Yeah, that's a good point. It's definitely still a challenge sometimes. I mean, different organizations deal with it differently, I'd say different sectors, but for sure, you know, we're in the category of an operations budget. That being said, we do have plenty of clients that will, especially with new construction, sort of bury the cost of monitoring for a couple of years in a new construction budget as a way to fund the initial piece, particularly as they're thinking about their commissioning budget or an M&V budget, or, you know, you can end up in different categories depending on what makes sense.
Um, but for the most part, you know, being able to build that into O&M , is key, and you're right, can be a challenge sometimes for people to realize that there's operational tools that need to be in that category.
James Dice: [00:05:12] Totally. And how about from an energy efficiency standpoint? So when I think about an energy efficiency project, you're doing some sort of study, right? And then you're installing some sort of ECMs, energy conservation measures, and then you're realizing savings after that. But what I've found is with building owners that have a hard time wrapping their heads around the Software as a Service fee, because it doesn't produce savings in and of itself usually, and so how are you guys helping to justify those ongoing fees? Because as we know, they're vital for maintaining the savings that come out of energy efficiency projects, but they're not necessarily required to implement energy efficiency projects themselves.
Alex Grace: [00:05:56] Right. Yeah. There's a couple interesting pieces there that you touched on for sure. So, some changes that we're seeing... I mean, you're absolutely right. There's a fundamental market education piece there, which is about energy drift, which those in the energy auditing, commissioning world know well, right? That just because you implemented a beautiful reset schedule, it doesn't mean it's still being followed a day later, much less six months later, much less two years later, right?
So probably how we've been from an M&V perspective, calculating savings and thinking that low cost, no cost measures actually just persist, is probably not right. That's one point.
And then secondly, you know, I think that on the maintenance side of the house, this really isn't as much of a concern, but on the energy project side of the house, it is more. And really the obvious point is valves don't stop leaking; dampers don't stop sticking; sensors don't stop failing. So everyone, again, on the O&M and the maintenance world knows that very well, but from an energy projects perspective, you may have the idea that you go and implement something, you squeeze the savings out, and you walk away. And it's just simply not the case.
So you know, it's, it's why our angle is really... Our clients and our partners are looking to drive predictive maintenance outcomes. They're looking to implement continuous commissioning processes for all the reasons that we know: because it improves operational efficiency, because their people can be focused on what matters most rather than just reacting and going down a checklist. They're zeroed in on the issues that are costing them the most amount of money, that are having the greatest maintenance impact or having the biggest impact to building occupants, the building comfort. So you know, that's why they're doing it. And then the natural byproduct of that process is energy saved.
And I like to talk about it and frame it in that way: that continuous energy savings is there, of course it is, but for a lot of our users, it's not the primary driver, although it may be very important to get that initial budget approval. And we have a client that will give a speech on stage and say, you know, we didn't care about energy. We're completely focused on maintenance optimization. We're having trouble retaining staff. You know, people are retiring. There's a major labor shortage in the industry, as we all know. We need to make sure that our people can really be laser-focused on having the biggest impact and getting more from our existing teams.
And then you'll say, oh, but yeah, we also saved eight hundred thousand dollars. So it's like the energy savings is there, whether you're focused on it or not.
James Dice: [00:08:25] Right. That's fascinating. And I think guys like you can really help the industry, and that's why I wanted to ask you this. We can really help the industry explain these new services and processes and technology a lot better because I think they're misunderstood in many ways.
Alex Grace: [00:08:41] Yeah. And James, one thing that came to mind there, just like as an interesting question that I think used to come up more, but it still comes up sometimes is this idea that I think maybe comes more from the commissioning approach where you're like, Hey, let's put fault detection on a building, and then let's move it, right? We do it for a year, and then let's move it to another building. And that's, that's a little bit from that same mentality, right, of sort of missing the fact that that mechanical degradation doesn't stop.
James Dice: [00:09:07] Yeah. And I've seen monitoring based commissioning incentive programs structured, where it's literally incentivizing firms to do exactly what you just described. I know the one in Chicago with ComEd is exactly like that. It's providing a one year incentive for something that's supposed to be an ongoing part of operations from year zero to infinity, right?
Alex Grace: [00:09:31] Totally. Yeah, and in some ways, I think that those programs are actually hurting the industry.
James Dice: [00:09:38] I agree.
Alex Grace: [00:09:38] And what you see consistently is where those programs came out of, are utilities that already incentivized a retrocommissioning program. They were used to doing that. They had the process down, and then they said, okay, now we're going to change that RCx incentive to an MBCx incentive, and we extend it from maybe two months to a year, but that's it. So like those that already had that model, because they already had a list of vendors that were approved, it was sort of an easy way to bring in the monitoring based conditioning idea. Whereas there are, I think, really shining examples that've taken a different approach, like NYSERDA's Real Time Energy Management is, I think, the most forward-thinking for a number of reasons. Mainly because they're incentivizing for five years.
So that's a whole different animal. But yeah, I do think that comes out of... It used to be pretty consistently that where those programs have appeared have been where they had a history of funding retrocommissioning, and then they just kind of barely shifted the model.
James Dice: [00:10:33] Yeah. And that just seems to be, to me, misunderstanding the opportunity, and kind of leads, like you said, to hurt rather than help. Cool. So let's continue on with these kind of sales-focused questions.
Um, so one thing I'm doing right now is creating a standard RFP for anyone who's procuring analytics software. And one of the challenges with that is creating some way to compare software like this FDD software in an apples-to-apples sort of way. So what are your thoughts on how that can best be done, and is it even possible at this point?
Alex Grace: [00:11:15] Yeah, it's a great question. It's definitely a challenge. I think it is possible. I think it's really about thinking beyond the initial... I have analytics up and running, or I have some basic fault detection up and running, but what is this gonna look like at scale, and what is this going to look like three years, five years from now? Meaning, what is the maintenance on the fault detection and diagnostics?
So I think something that's often not maybe given as much attention to as it needs to, is that buildings are dynamic. They're changing all the time, and the FDD needs to evolve with those operational changes. So you set up a bunch of static rules that were customized to a building, and then you change your sequence or you replace and air handler a year later. Who maintains that code set?
So, you know, are you getting costs associated with just standing up some initial FDD and hoping it works today? Well, that FDD may be producing a lot more false positives a year from now if you're not thinking about how that code set evolves and needs to be maintained. That's one point.
It is a challenge. You know, I think we love seeing people test multiple solutions for that reason. It just, it helps us because sometimes things can be abstract until you've had your hands in it, you know? So we've had clients-
James Dice: [00:12:29] Sorry, what was that? Test multiple solutions?
Alex Grace: [00:12:31] Yeah. So if you're looking at multiple FDD products, you know, you have a couple of buildings. Or you know, the most fun cases we've had is where someone actually has a historian that they're able to serve data to multiple FDD vendors and see what happens. I mean, that's a great way, but there's very few people that can do that. And you know, I don't recommend spending a year trying to figure out how you abstract your data in order to do that. You know, you want to get started and that's really the bottom line, right? Get started. You're going to learn a lot, you gotta move forward. And I think that requires leadership to really have that fail fast mentality. You know, I talk a lot about the University of Iowa because they've really done that well, you know, from the beginning, they had this mentality that we're going to test a solution, we're not going to be tied to it. We're going to learn a lot. And, and having that, making sure that your organization is a learning organization is, I think fundamental here. Yeah, so apples to apples is tricky, I'd say the majority of our biggest, and best clients have already done other types of fault detection before because they just get the complexity. They get that it's not as simple as just tagging everything and expecting things to work, that you have to really understand the operation, and that false positive is the Achilles heel of this industry, you know? So the ability to effectively prioritize and avoid false positives is critical.
And I do think that there's ways to write requirements that ensure that you're getting, for example, energy calculations accurate. And that rules are aware of other rules. And I know when you had Nick on, you guys talked about mass customization, so to have a custom solution that's going to degrade the second that you have it up and running, but you have something that is constantly evolving and code sets that are shared and parameterized in this mass customization approach. And another way, I like to say that just that the rules are aware of each other. So if you're implementing a bunch of rules, you know, I think that's key.
And I love the conversation you sparked on LinkedIn, James, around like, where's the IP? Is the IP in the rules themselves? And I just think that's a really important point from my perspective, that, or from the KGS standpoint, that, you know, identifying that a valve is leaking or a damper or stuck is not that challenging. There's no IP in the math associated with finding a leaky valve. The IP is how do you structure those code sets in umpteen different ways and hundreds of different scenarios with different types of mechanical configurations and sequence of operations, and how exactly is that economizer controlled that's controlling the damper? And what about dehumidification and all these things that you run into as you know from your background and your experience, right?
James Dice: [00:15:17] Yeah. And for everyone who hasn't seen that conversation, I basically asked all my contacts on LinkedIn, you know, what do you guys think about intellectual property when it comes to fault detection rules? And I'll put that conversation in the show notes. The responses were pretty fascinating to me.
So first of all, I want to key in on intellectual property because I think it relates to a conversation that you and I have had in the past, which is: where along the value chain can you commoditize this and where can't you?
Alex Grace: [00:15:50] Yeah. That is interesting. We hear a lot of conversation in the industry around fault detection focused on tagging. Tagging's important. Normalization's important. You know, Nick spoke more eloquently than I can on all the different things he's involved with related to the ASHRAE BACnet Committee and establishing industry standards that we want to be really heavily involved with as we think it's incredibly important to push things forward.
That being said, you know, for us, tagging is a starting point. It's not an ending point. I think a lot of people talk about it as an endpoint, and frankly, if you have very simple rules, it is an endpoint, right? I know what the data is, and therefore I'm running a full linear logic to say, you know, is something off here? Has it deviated?
You know, we sort of see that type of quote unquote fault detection as really alarming or advanced alarming potentially, but to get to the level of diagnostics, you have to understand a lot more than just, what are the points? You have to understand where equipment is. You have to understand where thermal elements are exactly in an air stream.
So not just understanding the points, but what is the order of those points in that air stream in an air handler? And then what are the sequences of operations that may or may not be reflected in the modes or the set points or the parameters that are visible from the points. They may be just written in a document somewhere, right?
So, and then having code sets that can dynamically determine based on the best available information. So that's a key thing that's a little bit harder to understand and requires some explanation, but it means that you want to have a situation where your rules don't just break because it's missing a variable.
And the way Clockworks diagnostics work is that it always looks for the best available information, meaning, okay, if this point's here, I'll use it. If this point is not here, can I use this other point or can I approximate with these two points or reference a inputed piece of metadata from a written sequence of operation to enter it?
So a really simple example would be, do I have a flow station? Yes, I have a CFM point. I'm going to use that in my calculation. No, there's no flow station. Okay. Reference fan speed and reference the rate of flow of the unit to approximate flow at any given point in time. Really simple example, but there's hundreds and hundreds of those.
James Dice: [00:18:13] Got it. Yeah, that's a great example of a place where intellectual property applies, versus the actual rule that you see when you're seeing the results of the fault detection that that might not be, I mean, cause those are pretty well understood by the industry at this point. Yeah. Thanks for taking us through that.
Um, you mentioned, a potential case study that you say you like to talk about, but I actually haven't heard the story of the University of Iowa. Could you kind of give us, you know, what's happened there?
Alex Grace: [00:18:45] Sure, yeah. I think I can because a lot of it's public, won an award from the Smart Energy Analytics Campaign, and they've won a number of awards. And Don Guckert, AVP of Facilities there has given a lot of public conferences, so I think it's... You know, we're always sensitive to making sure that we can speak there. And I should say that client is through a partner of ours Schneider Electric. So, that's important.
But just in terms of what Don speaks about in the industry that I think is really, really important is the way that he articulates institutional knowledge and shared knowledge, that I think is a really interesting point. So you hear a lot in our industry about institutional knowledge. You know, we're really into, rightly so, that technician has been here for 30 years and he walks into a chiller plant and based on a sound, he knows what's wrong, right?
And Don I hope he doesn't mind me stealing this quote from him where he says, let me tell you an example of institutional knowledge. He gives example of a technician who would walk around with pieces of wood. He knew exactly the right size piece of wood to jam in the damper to be able to fix a problem. And he goes, that's institutional knowledge. He's like, we don't need institutional knowledge. What we need is shared knowledge. And I just think that's a really powerful illustration of what we're talking about.
And so what is shared knowledge? It means that you have a single source of truth. It means that you have, in this case, a diagnostics platform, where the data doesn't lie. It's accurate. It's tuned to your operation and tuned to your sequences, and that a mechanical technician, a controls technician, an energy engineer, a sustainability professional, a design engineer on the capital planning side, they all have access to that same information based on what they need to do their job. And I think overall, as an industry, that's a really useful concept.
The reality is institutional knowledge is retiring. That's the reality that we're all facing. Right? And the next five years is more and more of that in 10 years more and more of that. So it's not really an either or, it's how do you replace institutional knowledge with shared knowledge and what are the organizational benefits associated with that? That I think is really powerful.
The other thing that's powerful, I think, from their story is just having many different stakeholders involved very early in the process. You know, I think something we may have talked about before, James, that oftentimes I think FDD gets siloed within a particular use case, and it's something Nick touched on a little bit as well. Right? You know, is it within commissioning? Is it within energy manager who's doing energy projects, or is it that shared knowledge tool that is spanning these different use cases? And it's also fine to get started with one use case, right? You don't want to boil the ocean. Totally fine to get started with one and then be aware of how you expand.
But for those that are higher up in an organization and thinking about this from a leadership perspective, mapping out your stakeholders, which is not a mystery. It doesn't require a whole, you know, consulting engagement. It's your controls technicians, it's your maintenance technicians, it's your commissioning agents or third party commissioning folks, and those on the design or capital planning team side. And how do you bring them together to derive value from a platform like Clockworks?
James Dice: [00:21:58] Got it. Okay, cool. Yeah, thanks for sharing that. It sounds like a really cool project. So a couple of last questions around sales. I'm just wondering from the perspective of you, you know, spreading the messages of analytics over the last however many years, many years, right? Um, what's it been like watching these buzzwords come into our industry? So, I'll start with IoT. I mean, when I think about Clockworks, you guys have been putting things on the internet for long before IoT was hot. So what's that been like for you and crafting a message around that?
Alex Grace: [00:22:34] Yeah. That is interesting. I mean, we're a pretty conservative company, so our marketing goes about as far as, you know, us doing podcasts with you, James. Right? So, so from that perspective, you know, we've always been a little bit skeptical of the hype cycle, and we try to not get too deep into it, right? We don't just shift our messaging immediately that, okay, now everything's IoT, or now everything's AI or whatever, right?
Um, but there's a lot to dive into there. So, you know, I think early on when IoT was coming out... So, okay. The whole conversation on IT OT convergence is totally fascinating. It's a 10 year old or 15 year old discussion and it's still totally relevant. Right? So I think early on when you saw the, the big IT players coming out, you know, I saw a presentation from Cisco, this is a number of years ago now, but to me it sounded like they'd never heard of a building automation system. In the way that they were talking about IoT, you know, it was like, we're gonna have these sensors, and they're going to talk to the centralized location, and you're going to be able to control things. I was like, um, excuse me. Right?
So that being said, obviously there's a lot of great IoT use cases that are out there. There's a lot of really exciting stuff. I mean, I know you talked about InfiSense, and we arere watching LoRaWAN, and we're watching that whole world. And there's a lot of really cool use cases that are out there, don't get me wrong. But I think my organization... I guess my reference point a lot of time with the types of clients that we work with, which those that have big facilities portfolios. They have large facilities teams. They have a lot of infrastructure. Before you start chasing IoT and new buzzwords, let's start using the data you have. Like you have already made millions and millions of dollars of investments in sensing technology, and that data is incredibly underutilized. So there's a massive opportunity to say, look, let's start with what you have. And then absolutely as there are more opportunities to bring in IoT, that's great.
But now I'm also talking from a little bit of an HVAC-centric perspective. You know, there's a lot of really cool use cases out there in janitorial or in other types of areas where, you know, having sensing that wasn't there before, it can be understandably a game changer.
James Dice: [00:24:49] Got it. Yeah. Cool. So how about machine learning? So I want you to talk about this from the perspective of how you're messaging it, but then also how you guys are thinking about it internally and whether to add these capabilities to your tools.
Alex Grace: [00:25:03] Sure. Yeah. Thanks for the question. I think there's a lot of, frankly, market education that is needed around that. I mean, we still sometimes see RFPs, for example, come out and say like, you know, do you have AI? And it's, you know, there's a lot there. So let me start. So, machine learning, we see a lot of potential in certain techniques and we're using it today for the main area of improving onboarding.
So we use natural language processing and machine learning to help onboard buildings more effectively. We have an incredible dataset. We have about 260,000 mechanical assets, meaning an air handler, a chiller, a boiler, a VAV, a pump in Clockworks today. And all of the data associated with those 260,000 equipment from about 380 million square feet, has all been totally normalized to a standard information model.
So based on that information that we have, when we ingest data from a new building, we have tools that will take that points list and help figure out what those points are based on everything we've seen before. And that is not just, have I seen that exact point name before? And match it to a database. It's obviously more complicated than that. And, our chief scientist can speak a lot more to the actual techniques we're using there, but what's exciting about it is it will continue to evolve. The more data you collect, the smarter you get, and we have a really significant dataset that will only continue to grow. So we see a lot of potential there for machine learning.
On the diagnostic side , it gets interesting. Right? So for those that are a little less familiar, you know, fundamentally machine learning, you're going to train a model on a certain dataset, and then you're going to identify deviations from that model, right?
We have yet to see examples of machine learning for diagnostics that get us closer to an accurate result than what we are already doing, which the academic term will be an expert system, which is a form of AI, an early form of AI, and hierarchal rule-based FDD, which is a fancy way of saying the rules are aware of each other, and getting to the root cause of the problem as much as possible.
So in other words, if you find a deviation-, if the air handler was in normal operation, and you trained a model on that normal operation, and now the air handler is in something different, it's deviated from that, okay. But can you say exactly what's wrong as a result of that deviation in a way that is more accurate than saying, I can see the temperatures on either side of this coil and the valve position is closed and therefore that valve is leaking? Or we're simultaneous heating and cooling, and I'm referencing my dehumidification sequence, and I can see that, yes, we should be dehumidifying, but we're dehumidifying too much. Like that degree of specificity and then adding the loads, looking at the excess kBtus, turning that into dollars, adding a comfort ranking or maintenance range to it, which is what we do today with our expert system. We're not seeing anything yet within machine learning that lets us do it better than that.
That being said, we have a lot of academic relationships. We monitor that world really closely. And one of the nice things about our technology stack and software as a service is that if something compelling comes out academia, out of the national labs, or out of our own research, we will build those ML techniques into Clockworks, but so far today, we're not seeing something there that is more compelling. But we are seeing a lot of excitement and have a lot of enthusiasm for what we're doing and continuing to evolve on the the onboarding and the natural language processing side. Long answer to a question. I hope that addresses-.
James Dice: [00:28:35] No, that's exactly what I was looking for. I love getting your perspective because you're intelligent about the technical side, but you're also grounded in the way that you're speaking to customers all the time on the business development side. So, and I've played that role for many years as well, and so I recognize someone else that I like to talk to in that regard.
Let's transition a little bit over to channels. I know it's something that you've been personally working on quite a bit. It sounds like you've been focused lately on getting Clockworks in the hands of service providers. So not exactly building owners, but people that serve building owners, so mechanical service providers, controls service providers, others. But let's start there. So what are the things that these firms should be thinking about when it comes to approaching this whole digitization phase of everyone's life cycle right now? I mean, we're all being digitized, whether we like it or not, so what are some of the things that these types of firms should be thinking about?
Alex Grace: [00:29:35] Yeah, absolutely. Cool. So there's a lot there. So all we have to do is Google digital transformation and we see just the massive amounts of information there, right? So every business is facing this to some degree or another. And it's really, I think it's important to reference that broad trend, right? Like this is a shift that's happening in the world, and the facilities industry, the mechanical industry, the controls industry are not, you know, are not insulated, right? Like, like this, this transformation has been happening, is happening and is accelerating. I think that's another important thing that the pace of change is accelerating.
So, I mean, you know, of course we have this fantastic global relationship with Schneider Electric, where we are a global technology partner and have been since, um, formally since 2012. And that's a really, really important relationship for us. And they've been really thought leaders from very early on about how their service strategy will change globally. And we're really proud to be a part of that transformation and a part of that story. And we've learned a lot over the years from that.
And then, you know, we're, we're working with a number of other types of partners today, both in the controls world, in the mechanical world, facilities management companies as well, is another important one. So just to talk about services for a second , broadly speaking, the old approach of: I have a checklist; I'm going to roll a truck to a customer site; I'm going to go down that PM list; and I grease bearings this frequently, and I change the belts that frequently, and filters on that timeframe. Or I'm time-and-materials, break-fix type focused, and I show up on site every once in a while, maybe one day a month, one day a quarter and my guy tells me what his biggest headaches are, maybe it's written on a piece of paper. And then I go and try and take care of those things. That model is fundamentally transforming and changing, and I really see that there's, there is a carrot and there's a stick, right?
So there's the carrot being the opportunity. The opportunity is for business leaders running the mechanical service business or control service business to realize that I have a major opportunity to differentiate the way we're doing things and to show customer value and to elevate the value of my services to a different level in the organization, that I'm not just thought of as someone that fixes things when there's a major problem, but I'm also thought of someone that helps me strategically understand the operation of my facilities, the risk factors that I'm facing, and where I can drive energy reduction, performance and long term value when it comes to capital planning and asset management as well. So there's a lot, there, but that transformation is occurring.
So, you know, that's the opportunity. And what I mean by that, okay, when you're up for renewal on that service agreement-, you have a $20,000 service agreement with the customer. And now you're up for renewal, and there's a new finance guy, and he says, well, what are we getting for this $20,000? Well, here's my checklist. You know, I checked a ton of boxes. We did all these things for you that was really important based on ASHRAE guideline, blah, blah, blah, blah. Okay. Does that guy know what you're talking about? Does he understand the value of that $20,000 agreement?
Now let's talk about a digital approach. So how does this transform? How does this change? I no longer only roll trucks reactively or based on a checklist. I now, when that guy goes onsite, when my technician appears, he knows exactly what the top priorities are in this building based on energy, what it's costing my customer, based on comfort, which is arguably the biggest cost, your employees themselves. Okay, during the normal COVID times when there are people in those buildings. And then thirdly, the mechanical severity of the problem. Can I fix something that otherwise is going to wake someone up at two in the morning and they're going to have to come into the building and have a massive headache on their hands? Right?
So now the technician shows up with these priorities. He knows what's in scope, based on his current agreement. He knows what's out of scope that he can now propose a work on. So obviously the massive business value there from the vendor side of being able to have a consistent ability to propose ROI-justified work, and from the customer's side, massively more valuable because they know that that technician is focused on the biggest risk factors for them, the biggest energy factors, and the biggest comfort factors for them.
And now let's flash forward to that same conversation a year in to your agreement. You're up for renewal with that finance person. You now have a dynamic dashboard or a static report as you wish saying, what did that $20,000 buy you? Here's exactly what we saved you last year as a result of doing predictive maintenance and continuous commissioning. Here's exactly the impact we had on your facility in terms of quantified maintenance value and quantified comfort value.
James Dice: [00:34:16] Wow. Yeah, that's powerful. And I've certainly seen, as someone that has been providing these types of software for a long time, I've seen what happens when you're providing, say an FDD solution for instance, and then you're asking the person that is on one of these service contracts to fix what comes out of the FDD, and there's these huge disconnects there between like, I'm not here to fix that. I'm actually here to do this other thing. And it's usually, like you said, find an issue, right? I'm not here to fix an issue. I'm here to survey. I'm gonna walk around and look at gauges. I'm gonna check filters.
And I think what you're saying here is that the new sort of service and sort of O&M process is we're actually focused on results.
Alex Grace: [00:35:04] Right.
James Dice: [00:35:04] With the time we're spending. Right?
Alex Grace: [00:35:07] Absolutely. In this line of, you know, let's spend less time identifying and more time fixing. And you know to be clear, obviously we're not replacing all the PMs, right? I mean, you're still gonna need to grease bearings. You're still gonna need to change filters. You're still gonna need to look at certain things. Absolutely. But there's a certain percent of that agreement and that labor that can be shifted to higher value activities. And that's the percent we target.
And we actually go to a pretty deep level on that. So we'll go through an exercise with partners. We'll look at their task list. How do you do maintenance today? And what are your typical agreements? And are you standardized? So that's the first step, right?
You have organizations that made acquisitions, and this branch does something this way, and this branch does something completely differently. You have others that have consolidated their approach to services and say, we have a three tier program. We have good, better, best, or you know, a T and M block hours, a PM, and full service.
So whatever that is, if you've already done that standardization, you're already a leg up. If you haven't, incorporating FDD into your plan could help you do that standardization. So we really go to this level of understanding: what are your service tiers that you have today? We go through a task by task basis. We assign hours to that with our partner. So yeah, if I don't need to check that thing because analytics is looking for it every single day because I have diagnostics running, that's going to save me 15 minutes per activity, cause I don't have to do that thing. And then you add up all those hours and you've come out with, Hey, it looks like I could actually offset a third of the hours here.
And you know, operations leaders can be a little bit threatened by that. I just want to make that comment. If they don't understand the big picture, which is that their guys are going to get way more work out of this than they ever had before. It's just that that work is going to be higher value.
The other thing is what can you do remotely, you know, really figuring out what can you do remotely versus what needs to be on site. And I think now in the age of, you know, coronavirus that's probably even more important than ever, but well after coronavirus has left us, that's still going to be really important for a business model perspective.
James Dice: [00:37:07] Great. Yeah, I agree. You mentioned the carrot and the stick or the carrot or the stick. So we've kind of painted a picture towards those that are going after the carrot. What about the people that are getting hit with the stick? What are the organizations that aren't quite keeping up with this digital transformation? What are you seeing happening to them?
Alex Grace: [00:37:28] Sure. I think that the organizations that are able to pick their head out of the noise and the day-to-day, I've got a million things on my to do list, and able to look into the future are seeing that there is a threat here. So there's the opportunity to really have better client outcomes, better client relationships, and derive additional pull-through revenue through their service organizations and provide just a better service for their customers.
And there's a stick of, if we don't do this, in five years is our organization going to be relevant? And that's a bold statement to make. But I really think that's how leaders of these companies are thinking and need to be thinking if they're not. Because imagine for a second that you're going to renew that service agreement, and you're doing things the exact same way you've always done them, and you're doing T & M or block hours, and they started getting calls from people that are offering the same value service from a similar price point with way more. With an analytics driven approach and reporting on that value to the customer at the end of the year . Where the customer value is just that much more explicit.
And then one thing I do want to circle back on, and the carrot that we're really focused on is I think a lot of companies are still thinking about FDD or a tool like Clockworks as another tool in the toolbox. So, you know, we talk about shifting from reactive to proactive maintenance. We also talk about shifting from reactive to proactive business model, right? So I want to bring that up, which is, you know what I mean by that is if you're an organization that is thinking about FDD just because one of your more sophisticated customers is asking about it and asking for your help, you might want to take a step back and say, do we want to continue to be reactive to customers asking us about this, or do we want to have a strategy?
So we are not, frankly, not as interested in partners that want to resell Clockworks as another tool. You know, okay, the customer asks about it, you know, here's something I can maybe talk to them about. We're much more interested in business model transformation, because that's what this industry, that's where things are going. And that requires a whole different range of discussions about understanding the partner organization, how they operationally deliver maintenance today, and the structures I was talking about: good, better, best; block hours, PMs, full service. And it also requires understanding how they go to market. You know, who's selling those service contracts? How are they incentivized? You know, are they thinking about building up a remote operation center, a network operation center? So it's the business side. It's the sales side. It's the operational delivery.
And for that reason, sometimes people are surprised, like they'll reach out to us with an RFP and say, Hey, we're evaluating five FDD vendors and they're sometimes surprised about how much we ask about their organizations and their businesses. And the reason is that we need to qualify our partners as well as partners qualify us, because we go deep. We go deep in understanding them and consulting on that business model transformation, because we think that's where the most value is on both sides.
James Dice: [00:40:27] Fascinating. So let's go through some of these business models.
So I have a list of about five different potential partners and channels of yours. So starting with these service contractors, what does the service contract of the digital world look like? So how are they packaging the technology in with their services into this new package?
Alex Grace: [00:40:49] Yeah, absolutely. So part of that equation I was saying before where you determine are there costs that can be offset? That's part of the equation. So let's say for example, you have, this is just hypothetical because it does depend and you need to dive into the details here to really be able to stay this, but as a hypothetical situation, you have a $20,000 service agreement. Let's say you're able to remove $5,000 in labor. And by remove, I mean offset. So you're gonna replace that with some diagnostic monitoring and you're gonna shift some of those hours to remote operation center approach.
And it doesn't mean you're starting with, you know, NASA space command. It means that you have a guy in the office, who is one of your most experienced technicians or an energy engineer, someone who really understands data and buildings and systems, and that person, by the way, can now be 10 times more valuable because you don't have them on customer sites one-to-one. You have them one-to-many, and they're now able to guide, with the diagnostic results, less experienced technicians. And there's a lot of operational value there as well. So you take that best guy, you put them in the office and you say, okay, we're gonna take an analytics-driven approach.
So you've got that $20,000 agreement, let's say $5,000 gets offset by a combination of hours internally from your experienced guy in the remote operation center and the cost of Clockworks to look at things. And then there's going to be some delta there that you're going to have to upsell. But the upsell is now way more value add, right? You're able to now talk about prioritization, about scoring that building in terms of metrics.
So every day, what are your top priorities relevant to energy, comfort, and maintenance, and how do you drive actions and quantify the results of the fix? You're now able to do that and show examples of that. And we're even seeing some organizations that are actually subsidizing that transformation. Meaning, are you willing to actually eat a little cost because you know that you're going to get way more project pull through work? And because you see it as a strategic priority to have this transformation occur for all the reasons we talked about related to the carrot and the stick.
So some combination there, and it's gonna look different for different orientations. You absolutely don't need to be eating it, but some are getting aggressive and saying, you know, this is important. We're just going to do this because we know it's gonna produce fruit and we know it's going to make us more profitable and expand our service business over time.
James Dice: [00:43:13] Fascinating. Okay, let's move on from the service contractors and go into commissioning firms and talk about the inherent one-time commissioning effort or one-time retrocommissioning effort business model versus the ongoing monitoring based commissioning model that analytics offers and presents as an opportunity. So can you talk a little bit about the carrot and stick for those guys and how the ones that are on top of digital transformation are transitioning their business models?
Alex Grace: [00:43:47] Yeah, sure thing. And this is a cool one cause I know you've got a lot of background here. Well, you have a lot of background in a lot of areas, but this'll be a fun one to talk about.
So the commissioning world has honestly been fascinating to me. And it's fascinating because I've always sort of wondered, why don't we have more commissioning firms that are using Clockworks and providing an ongoing service? And I saw, actually one proactive commissioning company that we're talking to right now turned me on to this report from the Building Commissioning Association, the BCA, that was really confirming for me. And it showed a study, and it was survey of their members, and the result of that survey was that over 90% of monitoring based commissioning projects are not going past 12 months. So they're ending after one year.
James Dice: [00:44:34] That is a staggering statistic for me to think about.
Alex Grace: [00:44:39] Yeah. Yeah. I thought that was incredibly fascinating. So I think the problems that I'm seeing the commissioning world are really business model based. They're obviously not technically based. I mean commissioning firms are incredibly well positioned in terms of skill sets of their people to drive an ongoing commissioning process with fault detection and diagnostics at its core. I mean incredibly well positioned.
And don't get me wrong, we're seeing plenty of companies that are using, I would say, toolkits in their commissioning toolset. So, okay. Rather than, you know, downloading data from the control system. First of all, there's still a lot of companies doing this: downloading data from the control system, spending three weeks messing around with it in Excel back in the office to come up with a report. Like, if you're still doing that, obviously that needs to change. I mean, that's just crazy.
But , and then there are companies that are changing that by going to, I'd say, a toolkit approach. So they're using an FDD tool where they can write their custom algorithms, because it helps them do their commissioning process. But they're still fundamentally project-based businesses, meaning: they do a project; they end it, whether it's instead of three months, now it's 12 months, it's still very limited; and then they move on. And the mentality is, well, the facilities teams are gonna pick it up from there. There is a big gap in the big market opportunity, I think, for commissioning firms to stay involved with their clients longer and to be adding more value on a continuous basis as it relates to operations and maintenance, by running FDD longer and continuing that process. I think it's not happening because it's not how project teams are based, and it's not how companies are structured or incentivized. So I think they're missing an opportunity.
And here's a different perspective to take. I firmly believe if you're a commissioning firm listening to this right now, you're working at a commissioning firm, five years from now, your best customers are all going to have fault detection and diagnostics. All of them. Do you want to be driving that and be a part of that solution or not? I really think that's what it comes down to. And I know that's a strong statement, but I'm going to make it here, talking with James Dice. I think it's really true.
You know, so if you're in this mentality of, well, we just don't know why customers aren't renewing and we're doing it for 12 months. Then you're missing something, because we have clients doing this for 10 years. And fundamentally, when organizations we work with start using Clockworks, they don't stop. Because how do you go back to reactive maintenance after you have a tool that tells you every single day where your priorities are?
And there's stuff that comes out of Clockworks that is more complicated, right? There's the leaking valves and the stuck dampers and the broken sensors, but there's also the: we could be staging our cooling towers more efficiently than we are, I need someone to rewrite the sequence; or our loops are all under loaded, Clockworks is telling me that the delta T in our chilled water loops is performing consistently bad, and it gives me a range of recommendations on how to fix that, but I still need an engineer now figure out what is the solution? What is the exact sequence I need to rewrite here to improve?
So I think sometimes people miss that, that there's a lot of engineering work that comes out of FDD if the FDD is actually doing the diagnostic piece and not just fault detection. So yeah, there's an opportunity there and there's the stick, which is all your clients are going to have this and how is your work going to change if they all do and you weren't a part of driving that solution?
James Dice: [00:48:09] Right. And I came from this world, so the way I see it is these companies have existed for a while. You know, commissioning and retrocommissioning have been, well-ingrained processes for a little bit. You know, there's still a lot of construction projects that don't use commissioning, which is a whole different conversation. There are still a lot of buildings that haven't been retrocommissioned, which is a totally different conversation. But in general, those two practices are pretty well -accepted as projects that are worth people's time, worth people's money.
But those firms that provide those services have a business model that is just like a construction project. It's a one-time event, right? So how are you seeing the business models for those companies transform to accommodate monitoring essentially?
Alex Grace: [00:48:56] Yeah, that's a great point. Okay, so practically speaking, you're doing commissioning, you're doing post-occupancy commissioning. Maybe you have FDD in there already to some degree. Um, this is where I think Clockworks does differentiate from some of the other tools because we do provide this O&M focus and we serve a lot of different stakeholder groups. But let's just say more generally, it doesn't have to be specific to Clockworks, you've got it running, and rather than going away to the next project, you now propose or you've proposed from the beginning that, how about we have a one day a month? So this agent that knows your system, has been testing things, has gone through the commissioning process, we're going to keep him on one day a month. We're going to check in on the information that the fault detection and diagnostics has produced. We're going to make sure you're not missing anything, and we're going to make sure any system retuning that's done continues to happen.
And I think, you know, at a minimum that's gonna make a lot of sense to the customer post-occupancy in terms of identifying things during the warranty period and holding vendors accountable. I'm talking about new construction now or major retrofit, but through that process, you know, you've got to make sure you're getting to the O&M folks, first of all, that you're not just seeing yourself as like, Oh, I'm just giving them an O&M manual and walking away, that you're really providing value of helping them see their issues, and that, you know, just because it's a new building certainly those of us who have done commissioning know that the problems don't stop. Right? So even after you've tuned things up, things are going to keep coming back.
It's also a difference in terms of portfolios, I would say. So that's the one-off building example, but there's also, if you're working with a portfolio and you're, for example, retrocommissioning a handful of buildings a year, moving your way through the portfolio, maybe you take a different approach.
If you look at the cost of that deep dive retrocommissioning every single year only touching a handful of the portfolio, versus what if I monitored everything every year, and then that same man hours I was spending doing my deep dive functional performance testing, I've shifted to letting analytics drive where to focus. So rather than testing a bunch of systems to figure out where the problems are, how about diagnostics tells me exactly where the problems are, and I spend those man hours helping engineer solutions across the portfolio? Because the cost benefit on that is dramatic. And it may not be even more costly from a customer perspective to do that.
Or the other thing to keep in mind is you don't have to run FDD on everything. You might decide, my customer has a 50 building portfolio. I'm going to monitor every chiller plant, every boiler plant, and every air handling unit across the portfolio. And then focus my retrocommissioning efforts really targeted on the buildings that are outliers based on that data analytics. And it's not just, you know, EUI cost per square foot. It's a much deeper dataset where I'm now seeing, maybe I'm looking at kW per tons, maybe I'm looking at kW per CFMs on ventilation system efficiency, and maybe I'm able to go even deeper and see exactly where sequences could be improved, where they're not being followed, where the resets that I engineered five years ago are no longer in place cause they've been over-ridden, and et cetera, et cetera, right.
James Dice: [00:52:03] Got it. Yeah. And something that I've-, so when I've designed these types of business models and these types of services in the past, I found this book called Subscribed to be really helpful for transitioning your proposals, transitioning all of your internal processes to go from this one-time, like percent complete invoice-based business model of these commissioning firms, to a subscription business model. You're basically turning your business into a subscription at that point. And these firms need to stop thinking about this in terms of a one time event. And I could get all up on my soapbox about that, but I'll stop.
Alex Grace: [00:52:41] I like that, James. And just one more thing there. I mean, I think that's a great point. I gotta check out that book; that sounds really useful. I just want to mention on the positive side, we have, we do have commissioning firms that are working with Clockworks. Uh, shout out to WSP in Boston, for example. But you know, organizations that are really differentiating what they're doing and they're going into the customer's sites.
So from a sales perspective, you're a commissioning firm, the power of going into a customer site in a competitive situation-, obviously, if the customer is totally race-to-the-bottom and focused, then yeah, this is maybe not what I'm talking about. But if you have a more proactive, forward-thinking customer, and you're able to walk into that meeting as we've seen some of our partners do and say, we do commissioning differently. Here's how we do it. You know, that's very powerful. And that we're going to have an ongoing O&M focus if you want that, you know . I think there's a big opportunity there to really differentiate your service in the market.
The other thing I want to just quickly mention is macroeconomic trends. So because I think that's been a factor lately also. Because commissioning firms are busy, like they're getting a ton of work, right? No one's, no one has guys sitting around at home starving for time.
So when construction is booming, when the economy is doing really well, and you've got new construction and commissioning projects left and right, why take the time to really take a step back and look at some of these things we're talking about? You know, I get that.
Well, the current reality of where we are with the current economic situation related to coronavirus, but I'll say more broadly, just economic cycles, macroeconomic cycles in general, there's going to be periods of construction slow down. There's going to be periods where we were taking a step back. Now is a phenomenal time to take a step back and think about that and where things are going. And I think that we're seeing that because I'm seeing inbound messages lately and I think a lot more actually than we had been before. And I think that's part of it is people just have a little more time to pick their head above the water from all the work they'd had and say, okay, where are things going? Where is the industry moving? How can I position our company to be where we need to be, you know, now and five years now?
James Dice: [00:54:44] That's fascinating. On this last, so we're kind of continuing on these different types of business models. So commissioning firms are often, sometimes part of, you mentioned WSP, part of larger MEP design firms. So how do all of these lessons apply to those types of businesses and what are you seeing for them right now?
Alex Grace: [00:55:04] Yeah, that's really cool. I think a couple things. One, every design engineer I've ever talked to is dying to know how their designs actually are performing in the field, because you never get that feedback as a design engineer, right? I mean, you have a model, you put the best, all of your knowledge into this design, but you don't really know how the operation is going.
You know, is it being operated the way it was designed or not? So that feedback is obviously very powerful , and we've seen that everyone wants that. You know, there's a, there's a message there when it comes to new construction from an owner perspective also, around you're spending millions, hundreds of millions of dollars, whatever the case might be, on this building and on these designs and on the install of these systems. For pennies on the dollar, do you want to ensure that you're getting what you paid for? Right? So , yeah, I think that is an interesting area. If commissioning firms that often are tied to larger MEP firms, if they're able to create that feedback loop to the design, that's incredibly powerful. And a tool like Clockworks, you know, would certainly allow you to do that, but that is something interesting to think about.
The other thing I'll mention is just around M&V specs. So we've seen M&V specs come out that, you know, I want a submeter every last thing in order to just see how my designs are performing or at least like compare that to the energy model. Don't get me wrong, that's fantastic. We love sub-metering. Put all the data acquisition out there that you can. But at the end of the day, you're still limited by, okay, performance is different than model. Now what? Like where's the problem? Right.
So I think, I think FDD needs to be thought of more in M&V than it is today. And there's a big opportunity there where it's to say, look, that's great you have a lots of sub-metering. That's great you're able to compare performance against your energy model. But can you figure out where the problem is specifically? Can you pinpoint it? And FDD can tell you exactly what's going on. Well, you know, these two things were reversed and your actuation is totally backwards in terms of how you're economizing or whatever.
James Dice: [00:57:14] Yeah, that's exactly right. And so we've covered most of what I was hoping to get through today. And so we're almost running out of time here, but I did want to hit in on something that we haven't really covered that I think is an important detail. So for all of these different service providers and for owners and their O&M teams, what happens when you have better diagnostics? In terms of when that guy needs to go turn that wrench or take that two by four out of the damper or fix that actuator.
What I get a sense from you guys is when you have better diagnostics, you have a better information model that produces those diagnostics, and that gives you more information that you need when you're going to go to fix that fault, whatever it is. So can you talk a little bit about what you're seeing on that side of things?
Alex Grace: [00:58:04] Sure. Just in terms of having better data to fix problems, yeah, absolutely. So, you know, we produce a range of possible causes, so we identify a problem, we prioritize it. Prioritization is key, obviously. We, you know, in 10 years plus of doing this, still haven't met a facilities team that's looking for another to-do list. Right.
But that being said, people do need priorities , and understanding what the real impacts of these problems are and how do I sort them accordingly? And then it's really important to have those possible causes where I can understand, is it-, first of all, that's critical to know: who do I send this issue to? Like is it a controls programming issue? Is it a mechanical issue? Or maybe I can't tell from the data. Maybe the diagnostic can say it's one of the two. In other words, the valve is leaking. Okay, let's go look at it. Either the valve is leaking or the temperature sensor was broken. That's always going to be a reality from the data.
So but what is the order of events? So first I'm going to go look at that mechanically speaking and see if the valve is in fact leaking. If it's not, then I send it over to the control shop. Now it might be the same person, it might be someone different, depending on the organization, they might be sending it to a vendor. Hey, Johnson Controls or Siemens or Honeywell or Schneider or whoever is coming tomorrow . Let me add this to the list of things, of broken sensors. Or just simply being able to search. So we have, you know, the search ability, so I can just pull up every faulty sensor in the building. And send that to a vendor.
So the possible causes things is really key. I think we're going to continue to evolve there. So when I mentioned like rules being aware of other rules, that's really important for that to be able to pinpoint the problem and not send people on wild goose chases. Again, false positives are the Achilles heel. You've got to have a structure. That one, allows you to drastically reduce false positives and two, when they come up, you can quickly change them. That's also key, because you're never gonna get 100% , and the process is important there.
But I, I'm really excited about areas that are kind of in our R & D bucket today as well, around what we call meta analysis, I'll just bring up quick. So we already have rules aware of other rules, structured into a hierarchy. The next level is you have diagnostics aware of other diagnostics, and so basically you keep building up layers of analytics capability, and we are continuing to evolve that, where you will get to a point where we have diagnostics on top of diagnostics and that means longer term degradations of equipment and patterns. That means more in deeper level of systems analysis. Um, again, we do a lot of that today, but there's there room to grow and it's some really cool R & D there.
James Dice: [01:00:28] Yeah, that's all awesome. I was hoping you would say that.
The one thing I was thinking about though, that I don't know that the service providers are thinking about is that when you have this common single source of truth that you were talking about earlier in the context of institutional versus shared knowledge, so that shared knowledge from my perspective, what I think is valuable for all of those service providers is that when you have, say, your fan motor is not working. Whatever the problem is that comes out of Clockworks or any FDD, that service provider is going to know the model number, the size of the motor, everything to do with that in that information model, before they leave the shop, right? So they're already knowing, like you said earlier, what their exact task list is when they get to the building. But the point would be that they already are ready to perform the exact task. And I think there's this, I don't know if that message is getting out there as far as the process efficiencies.
So it's not just about labor savings, it's also about everything to do with the process of performing these ongoing services.
Alex Grace: [01:01:36] Yeah, that's a great point. Sorry, I think I missed your question a little bit, so-
James Dice: [01:01:38] No, you were great.
Alex Grace: [01:01:40] So I think you're really touching on a key thing, which is the asset management side. So what are the datasets and the data silos that need to be combined, right? So you now have this fault detection information and a history of fault detection. So how many times has that valve leaked on that air handler, and how many excess kBtus has it cost you in the last three years, every time you've fixed it?
You also have the asset service history. So what are the lists of fixes and PMs and reactive break fix, and what are the labor and material costs associated with that? So being able to combine the datasets from an asset management tool or a CMS tool with fault detection, we see as really critical, and we are having that conversation with some service organizations that are larger, that are really thinking about the value of that data.
I mean, everyone's talking now from a business perspective, the values of data, data is the new oil, et cetera, et cetera. Right? And those that are thinking about like, what is the strategy there? So you have all this asset history, you have all this fault detection history. What are you providing?
So specifically to what you're saying, the process efficiency, absolutely. Not just showing on site with a prioritized list, but knowing what parts to bring. I mean, you can't underestimate that if you're rolling a truck, particularly if you're in a situation where you have to travel some distance. You know, if you have to send a guy an hour away or two hours away, even more critical that you know what the problem is and that you couldn't have fixed it remotely.
So imagine you fix the things that you can fix remotely, cause you have remote access and you're providing, for example, control service. And the things that you can't fix remotely, you know the exact sensor you need to bring or the exact actuator that needs to be replaced or whenever the tools are that are gonna allow you to pinpoint the fix that's needed.
James Dice: [01:03:18] Totally. Yeah. And so I can't help myself. I also have to ask about, so from the perspective of the owner, a lot of what I hear from owners in this regard is, is it going to integrate with my work order system? And I hear two perspectives, it seems like from the marketplace in this regard. One is, Oh yeah, we do it and it's happening all the time and it's an easy integration. And then on the other hand, I hear, I heard this in a meeting last week: No one's doing it. It's too difficult to do. It's too hard to write integrations for every CMMS or computerized maintenance management system. So what are you guys seeing for your clients and what have you guys built up as far as integrations with work orders from faults?
Alex Grace: [01:04:00] Yeah, that's super interesting. I'd say the truth is somewhere in the middle, from my perspective. So I'll say that off the bat, we are doing it successfully. We have work order integrations in place with multiple customers. And it's also true that just because you have Maximo, everyone's implementation of Maximo is different.
What fields specifically you're filling out and how those fields need to be mapped to combine an output from diagnostics from Clockworks with that system needs to be defined. So it's both-and. So fundamentally, if your work order system has an API, we can talk to it. If you don't, please upgrade to a system that does, you need to be thinking about that, right?
So, for example, if you have like a legacy platform, a lot of people are in process of upgrading from, for example, FAMIS local to FAMIS online or whatever the equivalent is , and there's a ton of tools out there. So we can do the integration, absolutely. There is a certain element of custom software development associated, because again, you have to define those fields and make sure that you're mapping that process appropriately. But we also have, once you do it once for a Maximo or AiM or FAMIS, that is replicable for us. But then it's the fields and the details that still need to be defined.
So basically what we're seeing is it's not something you do out of the gate because there are costs. You may not do it for the first building that you do, but if you've reached a certain degree of scale, everyone we know both in terms of our enterprise accounts and partner organizations are thinking about this, that they need to do it. Some of them already have. There's great process efficiency gains there that are useful. And also having those combined datasets, you know, as you pointed out around asset management can be powerful and is important to think about.
James Dice: [01:05:47] Cool. Yeah, I've got a sense we could talk all day, but I think we gotta wrap things up here. I want to thank you for spreading your knowledge around to the industry and letting me learn from you guys this week and last week. So I really appreciate it. Thanks for coming on the show.
Alex Grace: [01:06:02] Thanks so much, James. It's great to talk with you. Really enjoyed it.
James Dice: [01:06:05] All right, friends. Thanks for listening to this episode of the Nexus podcast. For more episodes like this and to get the weekly Nexus newsletter, please subscribe at nexus.substack.com. You can find show notes from this conversation there as well. As always, please reach out on LinkedIn with any thoughts on this episode. I'd love to hear from you. Have a great day.
Head over to Nexus Connect and see what’s new in the community. Don’t forget to check out the latest member-only events.
Go to Nexus ConnectJoin Nexus Pro and get full access including invite-only member gatherings, access to the community chatroom Nexus Connect, networking opportunities, and deep dive essays.
Sign Up
Nexus Labs exists to move the smart buildings industry forward, without the fluff.
On Wednesday, we send out the renowned (and free) Nexus Newsletter.