Welcome to The Lens, a monthly-ish recurring series where I unpack the strategy and context behind the news in as few bullet points as possible. Â For past editions, check out Vol. 1, Vol. 2, Vol. 3, Vol. 4, and Vol. 5.
For Volume 6, let's talk about occupancy data.
Enjoy!
What happened?
In the last year, huge (for our industry) funding has flowed into occupancy data startupsâDensity has raised over $100M, VergeSense over $20M, Â Calumino at $9M to name a few
Space and workplace management software also continues to see large investments, acquisitions, and partnership announcementsâSee Schneider Electric+Planon, VTS+Rise Buildings, Spaceti+Cisco to name a few.
Memoori's 2020 report on occupancy data unpacked the "striking diversity and range of companies addressing this opportunity". They found 221 companies actively engaged in providing Occupancy Analytics or Location-Based Services in just the office buildings sector. They predicted the market growing at a Compound Annual Growth Rate (CAGR) of 21.5%.
Why?
Talking about occupancy data trends feels like we're talking about the entire industry's trends. That's because, as we discussed recently in the Nexus newsletter, for every type of building except data centers, occupancy data is the most important type of data. No other type of data can benefit so many building stakeholders.
Office space management and retail use cases were hot even before the pandemic. Energy codes were already requiring more and more occupancy sensors controlling lighting and HVAC.
The pandemic accelerated those trends and took them to another level entirely. Building owners and tenants now need to know questions like, "Should we let more people in or not?" and "How many people showed up last week?" They want more granular and real-time data. Tech can obviously help, even if it's simply visualizing sensor data, as shown below via Metrikus.
As shown above, it helps building owners (and tenant businesses) understand how their customers are using the space, from office workers to retail customers to school students. It helps optimize the Three S's that are vital to controlling building systems and getting to full autonomy: Setpoints, Schedules, and Sequences of Operations. It helps solve for the optimal balance between the competing goals of better IAQ and lower energy use. It can help improve occupants' experience and find the right space to suit their current needs. Â Memoori agrees... check out all the office building use cases they've identified:
Let's not forget Moore's Law and all the underlying technologies that are enabling innovation here. Device and protocol improvements are enabling 15-year battery life for wireless sensors. Some (see Disruptive Technologies) are so small that the installation instructions are simply "Peel and stick".
Although infrared occupancy sensors have been around for something like 25 years, new approaches are akin to a body heat camera feed. That data can then be analyzed for deeper, more granular insightsâe.g. activity type, did they fall down?, etc. See an example from newcomer Butlr below.
â
That's where the cloud comes in. Machine learning advancements help analyze those data streams and improve both accuracy and insights.
Thanks to edge computing advancements, those machine learning algorithms can also be pushed down into the building, so the data can be processed locally and ease privacy concerns.
The context
Let's look at this in terms of technology waves. The first wave was binary occupancy sensors that were integrated into the BAS or LCS. Obviously, this was very limited. But at least it allowed you to control systems based on occupancy.
The wave we're in now is the full-stack point solution wave. As I wrote back in May:
Densityâs core differentiator is that their sensors count humans anonymously and accurately. But that data stream is not what they sell. They sell the entire stack, from hardware to SaaS.
They want you to use their SaaS to âsimplify space planning, manage operating expenses, and meet new workplace expectations with a single platformâ, even though all of those things depend on more than just occupancy data. Theyâre not a part of the stack, they want to be the whole stack.
I see today's point solutions as the ultimate abuse of the word 'platform' (Youâre pulling sensor data to the cloud and displaying it... congratulations!). Where we sit today is that point solutions are providing only some of the value prop that occupancy data can provide. And building owners aren't willing to pay for the full value if they're not getting it.
That means that the subscription fee (for software or for the APIâDensity's fee is $800 per sensor per year) is looked at with scrutiny. I'm hearing lots of feedback from leading building owners on this:
âI'm not paying ongoing costs for a 'platform' Iâm not usingâ
Iâve never really had a problem with SaaS, as long as you meet the expectations that come with an ongoing subscription fee. You need to continue providing value and grow that value over time. But continuing to provide data from a sensor and displaying it in the cloud? That doesn't cut it, even if you're using fancy ML. Sorry.
Also included in this second wave is the birth of comprehensive overlay software. Building owners don't want a point solution for each data source (energy, IAQ, occupancy, etc). They want a comprehensive platform: many stakeholders engaged, many use cases provided. They're either going to build it themselves or buy it... eventually.
Across a portfolio, occupancy information might come from 15+ different sources (Access control systems, LEDs with sensors embedded, BAS-integrated occupancy sensors, LCS-integrated occupancy sensors, Mobile devices to wifi, Cameras, etc). An overlay software or independent data layer is needed to abstract away that complexity and determine that best-available occupancy info. As you all know, I'm bullish on this approach.
So what's next?
I think there are three distinct roles here. First, the sensor+analytics package. Second, the overlay that's responsible for determining the best available occupancy data and providing it in context with all the other data. Third, the overlay or overlays that engage the end-users and capture the full value of the data. I think wave three will see the best in each role win out.
For the sensor+analytics, I think that means an API-first approach. There are at at least 4 companies that are taking that approach and they're shown in the main image above: Xovis, Calumino, Irisys, Disruptive Technologies (I'd love to hear more if you know of any).
For the overlay, I think that either means an independent data layer approach or at least a flexible way to charge customers for only the value captured.
Wave three also needs to do more with the data. We need to get back to using all this amazing data to actually control systems better. That means better supervisory control. Wave two has gotten away from that.
Finally, from an energy standpoint, we need to start quantifying how well the performance of the building adapts to variable occupancy. As weâve discussed in the past (here and here), more data allows our smart building platforms to create better key performance indicators. KPIs allow the ability to zero in quickly on under-performing buildings and systems to actually do something with all that data.
What new KPIs are enabled by more prevalent occupancy and energy data? A 2020 study in Building and Environment explored just that. The authors argue that the typical commercial building just doesnât adapt its energy consumption to real-time occupancy very well, and KPIs to measure âadaptable building performanceâ are proposed to help change that.
They propose 3 sets of KPIs:
Metrics that focus on quantifying occupied vs. unoccupied energy use
Metrics that focus on measuring the utilization of different building systems at full capacity relative to equivalent occupancy at full capacity
Metrics that focus on using hourly occupancy as the normalizing factor for building performance reporting
Thatâs all for The Lens this month! Thanks for reading,
âJames
P.S. These are obviously just my opinions that I always welcome feedback on. Three questions for ya:
Did you like this? If not, let me know by hitting reply or leaving a comment on Nexus Connect.
Where am I wrong?
What news should I turn The Lens on next month?
P.P.S. Thanks to those of you who responded to my newsletter on this topic and gave me some feedback/content/help on the above.
Welcome to The Lens, a monthly-ish recurring series where I unpack the strategy and context behind the news in as few bullet points as possible. Â For past editions, check out Vol. 1, Vol. 2, Vol. 3, Vol. 4, and Vol. 5.
For Volume 6, let's talk about occupancy data.
Enjoy!
What happened?
In the last year, huge (for our industry) funding has flowed into occupancy data startupsâDensity has raised over $100M, VergeSense over $20M, Â Calumino at $9M to name a few
Space and workplace management software also continues to see large investments, acquisitions, and partnership announcementsâSee Schneider Electric+Planon, VTS+Rise Buildings, Spaceti+Cisco to name a few.
Memoori's 2020 report on occupancy data unpacked the "striking diversity and range of companies addressing this opportunity". They found 221 companies actively engaged in providing Occupancy Analytics or Location-Based Services in just the office buildings sector. They predicted the market growing at a Compound Annual Growth Rate (CAGR) of 21.5%.
Why?
Talking about occupancy data trends feels like we're talking about the entire industry's trends. That's because, as we discussed recently in the Nexus newsletter, for every type of building except data centers, occupancy data is the most important type of data. No other type of data can benefit so many building stakeholders.
Office space management and retail use cases were hot even before the pandemic. Energy codes were already requiring more and more occupancy sensors controlling lighting and HVAC.
The pandemic accelerated those trends and took them to another level entirely. Building owners and tenants now need to know questions like, "Should we let more people in or not?" and "How many people showed up last week?" They want more granular and real-time data. Tech can obviously help, even if it's simply visualizing sensor data, as shown below via Metrikus.
As shown above, it helps building owners (and tenant businesses) understand how their customers are using the space, from office workers to retail customers to school students. It helps optimize the Three S's that are vital to controlling building systems and getting to full autonomy: Setpoints, Schedules, and Sequences of Operations. It helps solve for the optimal balance between the competing goals of better IAQ and lower energy use. It can help improve occupants' experience and find the right space to suit their current needs. Â Memoori agrees... check out all the office building use cases they've identified:
Let's not forget Moore's Law and all the underlying technologies that are enabling innovation here. Device and protocol improvements are enabling 15-year battery life for wireless sensors. Some (see Disruptive Technologies) are so small that the installation instructions are simply "Peel and stick".
Although infrared occupancy sensors have been around for something like 25 years, new approaches are akin to a body heat camera feed. That data can then be analyzed for deeper, more granular insightsâe.g. activity type, did they fall down?, etc. See an example from newcomer Butlr below.
â
That's where the cloud comes in. Machine learning advancements help analyze those data streams and improve both accuracy and insights.
Thanks to edge computing advancements, those machine learning algorithms can also be pushed down into the building, so the data can be processed locally and ease privacy concerns.
The context
Let's look at this in terms of technology waves. The first wave was binary occupancy sensors that were integrated into the BAS or LCS. Obviously, this was very limited. But at least it allowed you to control systems based on occupancy.
The wave we're in now is the full-stack point solution wave. As I wrote back in May:
Densityâs core differentiator is that their sensors count humans anonymously and accurately. But that data stream is not what they sell. They sell the entire stack, from hardware to SaaS.
They want you to use their SaaS to âsimplify space planning, manage operating expenses, and meet new workplace expectations with a single platformâ, even though all of those things depend on more than just occupancy data. Theyâre not a part of the stack, they want to be the whole stack.
I see today's point solutions as the ultimate abuse of the word 'platform' (Youâre pulling sensor data to the cloud and displaying it... congratulations!). Where we sit today is that point solutions are providing only some of the value prop that occupancy data can provide. And building owners aren't willing to pay for the full value if they're not getting it.
That means that the subscription fee (for software or for the APIâDensity's fee is $800 per sensor per year) is looked at with scrutiny. I'm hearing lots of feedback from leading building owners on this:
âI'm not paying ongoing costs for a 'platform' Iâm not usingâ
Iâve never really had a problem with SaaS, as long as you meet the expectations that come with an ongoing subscription fee. You need to continue providing value and grow that value over time. But continuing to provide data from a sensor and displaying it in the cloud? That doesn't cut it, even if you're using fancy ML. Sorry.
Also included in this second wave is the birth of comprehensive overlay software. Building owners don't want a point solution for each data source (energy, IAQ, occupancy, etc). They want a comprehensive platform: many stakeholders engaged, many use cases provided. They're either going to build it themselves or buy it... eventually.
Across a portfolio, occupancy information might come from 15+ different sources (Access control systems, LEDs with sensors embedded, BAS-integrated occupancy sensors, LCS-integrated occupancy sensors, Mobile devices to wifi, Cameras, etc). An overlay software or independent data layer is needed to abstract away that complexity and determine that best-available occupancy info. As you all know, I'm bullish on this approach.
So what's next?
I think there are three distinct roles here. First, the sensor+analytics package. Second, the overlay that's responsible for determining the best available occupancy data and providing it in context with all the other data. Third, the overlay or overlays that engage the end-users and capture the full value of the data. I think wave three will see the best in each role win out.
For the sensor+analytics, I think that means an API-first approach. There are at at least 4 companies that are taking that approach and they're shown in the main image above: Xovis, Calumino, Irisys, Disruptive Technologies (I'd love to hear more if you know of any).
For the overlay, I think that either means an independent data layer approach or at least a flexible way to charge customers for only the value captured.
Wave three also needs to do more with the data. We need to get back to using all this amazing data to actually control systems better. That means better supervisory control. Wave two has gotten away from that.
Finally, from an energy standpoint, we need to start quantifying how well the performance of the building adapts to variable occupancy. As weâve discussed in the past (here and here), more data allows our smart building platforms to create better key performance indicators. KPIs allow the ability to zero in quickly on under-performing buildings and systems to actually do something with all that data.
What new KPIs are enabled by more prevalent occupancy and energy data? A 2020 study in Building and Environment explored just that. The authors argue that the typical commercial building just doesnât adapt its energy consumption to real-time occupancy very well, and KPIs to measure âadaptable building performanceâ are proposed to help change that.
They propose 3 sets of KPIs:
Metrics that focus on quantifying occupied vs. unoccupied energy use
Metrics that focus on measuring the utilization of different building systems at full capacity relative to equivalent occupancy at full capacity
Metrics that focus on using hourly occupancy as the normalizing factor for building performance reporting
Thatâs all for The Lens this month! Thanks for reading,
âJames
P.S. These are obviously just my opinions that I always welcome feedback on. Three questions for ya:
Did you like this? If not, let me know by hitting reply or leaving a comment on Nexus Connect.
Where am I wrong?
What news should I turn The Lens on next month?
P.P.S. Thanks to those of you who responded to my newsletter on this topic and gave me some feedback/content/help on the above.
Join Nexus Pro and get full access including invite-only member gatherings, access to the community chatroom Nexus Connect, networking opportunities, and deep dive essays.
.png)
.png)
