The famed New Yorker movie critic Pauline Kael wrote a book entitled I Lost It at the Movies. Far less famously, as a young journalist I too studied and wrote about film. Having had such an experience, if I were to write a book about movies and their effect on me, I might call it Everything I Know I Learned from the Movies, because I can say honestly that film has influenced my view of the world deeply, for better or worse. Mostly, I’d say for the better, because I firmly believe art, particularly film, can reveal and inform in an often more effective and lasting way than mere facts and science.
So when I spoke recently to John Gilbert, COO/EVP/CTO (I hear he does windows, too) of the Rudin Management Company, about the world of sensors, a movie reference he recalled particularly resonated with me: “I like to say that if the movie, The Graduate, was remade in 2019, Mr. Robinson wouldn’t have said ‘plastics’ to a young Dustin Hoffman, he would have said ‘sensors.’”
As art imitates life and life imitates art, Gilbert’s comment seemed perfect in defining the impact sensors have on virtually every aspect of our lives, most especially in PropTech, where the world of sensors play an ubiquitous, exciting and sometimes frustrating role as we work to make our offices, homes and other buildings smarter, more efficient and environmentally sounder.
Are sensors what we think they are?
If you ask someone what their five senses are, most people can immediately answer: hearing, sight, touch, smell and taste. However, how do we define sensors, a word which, like the term senses, is derived from the Latin word sensus, meaning sensation or feeling?
The Rudin portfolio is comprised of 18 residential buildings totaling 4.7 million square feet, 16 commercial office buildings totaling 10.2 msf and two condominiums under management totaling 241 residential units. So Gilbert has a big sand box of experience to draw from when he says: “To me sensors are really the nerve-endings of a building’s central nervous system. They are, and it doesn’t matter whether you’re collecting temperature data, oxygen data or CO2 data, those sensors really are the key piece, the key element of collecting that data.”
Connell McGill, CEO & Co-Founder of Enertiv, a leading startup that gathers and uses building energy data to create asset value, offered that, “PropTech sensors should be broadly defined as any device that captures data of a physical system or process. They come in all shapes and sizes; many are wireless, but some are connected directly to the system itself, such as equipment or the electrical infrastructure.”
Indeed, sensors of various types have been around for a long time. In the early 20th Century, Vaisala, a company founded in Finland, developed a sensor that could be born aloft in an air balloon to help measure weather patterns. Today, Vaisala is a global leader in environmental and industrial measurement, as well as a central player in Finland being a hotbed of PropTech innovation of all kinds.
The first electric thermostat, a kind of sensor, is credited to Warren Johnson (the founder of what is now know as Johnson Controls) and was brought to the market in 1883. Infrared sensors have been around since the late 1940s. By the 1950s, Samuel Bagno had invented a motion sensor for alarm systems using ultrasonic frequencies and the Doppler Effect. Today sensors touch us day and night, from the benignly common smoke detectors to high-tech night goggles that provide the US military with a decisive and deadly advantage over adversaries.
The need for sensors in the modern world, particularly in the built world, has long been great and growing with increasing rapidity according to all those involved in the technology. Sensors are used for a vast variety of data gathering, including air quality, temperature, lighting, energy flows and usage, building stability (think earthquake prediction), access and security, water control and much more.
According to the early stage PropTech venture capital firm MetaProp [Full disclosure: I’m a former Partner at MetaProp], the number of startups creating or using sensor data in PropTech has risen rapidly over Q4 2018 and has had an uptick in use in the following verticals, in Q1 2019:
- Energy Optimization (Think: HVAC, for instance)
- Water leak detection
- Room-level & Building-level space usage sensors (Think: The office in which you sit and how and where people move through a building.)
- Building maintenance (doors, elevators, etc.)
- Insurance (Think: Live and predictive monitoring for possible accidents.)
- Security
- Construction Safety (Think: Preventative monitoring to lower your insurance costs.)
Along with tracking sensor startup growth, MetaProp also uses sensors in its new Manhattan office. “We use sensors for all kind of insights at our office,” said Bryan Lyandvert, an Investor at the firm. “We use sensors to monitor our energy usage to make sure we aren’t wasting electricity excessively. We use sensors to make sure we make the most efficient use of our space when we have portfolio companies visit. We even use sensors to make ourselves more comfortable during this Polar Vortex by monitoring the differences in temperature by the windows and the walls.”
As wide as the use of sensors is, the depth and granularity of what they can measure in just one aspect of the built world is equally expansive. For example, Flo, a comprehensive home water monitor, conservation and alarm system, concentrates solely on that single element. Being expert in all sensor-based water issues in homes has proved to be a highly viable business proposition for the PropTech startup.
“Inside our Flo by Moen device there’s a temperature sensor, a pressure sensor and a flow sensor, which is all three of the fluid mechanical elements of the water system,’’ said Gabriel Halimi, CEO & Co-Founder of Flo. “So we can tell you an incredible amount about your water. What’s exciting about sensors in real estate is that they give you primary reads on certain things, but then it’s the kind of insights you can gather when you put all of these different data points together. So if you just take our sensors together, we can not only tell you a lot about your water, we can tell you a lot about occupancy in the space. When people are in their homes or in a building, they tend to use water and so you can tell a lot about the real estate by how the water is used. If you can differentiate between different kinds of fixtures in the home, we can tell if someone has an irrigation system, if they have a pool or if they have a bathtub and how often they use those things. When you look at the flushes of a toilet, those are pretty good indicators of whether people are actually home. It can also be an indicator for health.”
Taking all this water sensor-derived information together produces a rich set of insights, concludes Halimi.
Cui bono? Who benefits from these insights?
Gilbert points out that there are three stakeholders within the built environment that all need to have digital communication, a digital conversation if you will, with a building’s “brain.” These digital conversations happen between the building ownership and the mechanical core of the building, with the large and small corporations that lease space in buildings and the individuals who live, work or visit the property.
Only with the last group, does the individual’s smartphone (more on this later) in effect become a sensor, Gilbert said. “That’s where this whole tenant experience app comes into play. But I’m not going to give up on that infrastructure that’s already embedded in the building, I’m just going to use that additional data to augment it. So I don’t think you’ll ever rip out the existing sensor network and just use phones. You will just use that additional data as long as you have a brain that has the capacity to understand that data, learn from it and predict and proscribe from it.”
Without sensors, you don’t have an IoT. We look at it very simply. Buildings have always had a heart — engine room, boiler room and centralized mechanicals room. Now they have a brain.
“The granularity of data is the next frontier,” Gilbert continued. “Without sensors, you don’t have an IoT. We look at it very simply. Buildings have always had a heart — engine room, boiler room and centralized mechanicals room. Now they have a brain with our product Nantum, a central nervous system and that’s where the sensors come in. But you need all three of those pieces working together, just like our bodies.”
Speaking of sensors, ay, there’s the rub
Of course, as in everything, progress in sensors is not a straight line toward perfection. The night before Halimi spoke to me, he, his wife and their two-year-old twins were awakened for a couple of hours when their carbon monoxide sensors went off in the middle of the night. Eventually, the fire department arrived, determined everything was okay, but couldn’t detect what had set off the alarm.
Despite such hardware failures, there is no doubt of the ongoing enormous impact sensors have in real estate, said McGill. “It’s hard to overstate the impact of sensors, and the data they collect, on real estate. There are obvious applications, such as using data to catch when something is going on, but there are also enormous opportunities for leveraging massive amounts of historical data to drive decision making. Even the best run portfolios still rely on assumptions, rules of thumb and the opinion of their teams. The impact of having empirical evidence to justify decision making around capital investments, leasing, etcetera, will be huge.”
However, the seemingly endless stream of new sensors being created to monitor everything in and about property, all the while spewing out huge amounts of data, provides a daunting challenge of analyzing and integrating the information and insights gleaned into a useable format. As Gilbert succinctly put it: “Data is useless unless there’s a brain that can make sense of it and make it actionable.”
That brings us to a continual question that pops up across the PropTech spectrum: What will be the platform that integrates, clarifies and makes actionable the data from untold numbers of sensors and many multiples of unique producers?
Halimi sees more platforms coming to solve these issues and to take the lead. “You can look at Amazon, you can look at ADT, there’s Honeywell, different cable companies, all of which are developing their own platforms for things like smart blocks and smart thermostats. You name it. But I think ultimately the most exciting part is how you bring all these different sensors together and the insights that can be gleaned when you put all these sensors together.” What we can learn from this data about property and people will make people safer and give them an overall better experience in their property, he said.
Yet, to me the nagging question remains: Are we over-monitored by sensors without enough value produced by them to make the effort and costs worthwhile?
“The cost of sensors is falling,” McGill said. “The range of wireless networking is improving. The sensors themselves are getting smaller and more accurate. The future of the sensor business is bright. But the real value for landlords comes from the insights that the data enables. This requires software and algorithms, which is the other half of the equation. Neither can be successful without the other.”
In fact, the commitment from the incumbent real estate establishment is still to be fully determined when it comes to do the hard work of creating an algorithmic platform for one’s portfolio that can robustly and accurately digest, make clear and produce actionable options for the owner. In fairness to the landlords, the financial cost and time and labor commitment to creating a platform can be steep, especially when considering that the owner might have to rely on less-than-fully-proven hardware and software from an array of PropTech startups.
A less enthusiastic view of the impact of sensors on smart buildings is expressed by Robert Entin, EVP & Chief Information Officer for Vornado Realty Trust, a company that owns and manages more than 23 million square feet of LEED certified buildings.
Entin sees some potholes in the road ahead, “I think today there’s been a lot of excitement over sensor technology. It’s got a lot of wide potential applications, no question about that, so I don’t want to be a naysayer. What I want to say is that it’s yet to be demonstrated in practice. You have some case studies here and there that look really good, measuring where people are and occupants on the floor and regulating HVAC, things like that, so there’s live examples out there today, but clearly I don’t think there’s this massive deployment that’s occurred yet with a massive disruption.”
“But given how cheap they are and how prevalent they’re going to be I think they’re no doubt going to be a big part of the future, not only in PropTech, but in general society. The question then is how do sensors fit in with cameras. The thing with sensors is you know where people are coming and going. You don’t know the who. So maybe as the world gets more comfortable with the complete loss of privacy, if that happens, maybe when that happens, sensor technology will allow for more specific identification of individuals coming and going.”
“When I think of sensors as opposed to cameras, they’re cheap low bandwidth and therefore they serve a great purpose. But like everything that’s converging, think about 5G bandwidth in the buildings, whether it is 5G private LTE or 5G public space. The bandwidth inside buildings is going to be massive. So maybe the answer is the camera and facial recognition and visual processing will continue to get better and better. That doesn’t fully overlap with what a sensor may or may not do, but there is definitely some cross-over there to me.”
“It’s just like the argument that people said rotating memories will go away in place of RAM, they have been saying that for probably almost 30 years, because RAM is getting much cheaper.”
Despite all that, Entin sees applications of scale for sensors in the future. “I think we’re at the beginning of the sensor revolution, but I have no doubt that it’s just one of the areas of PropTech inside the building that continues to grow. I think we’re at 1.0. The people who have deployed them successfully are of course enthusiasts, as they should be. The first time a baby gets up and takes a few steps it’s a massive thing, but in a few years, when you seeing a baby running, jumping and playing in a jungle gym that’s really a lot of progress.”
More power, Scotty!
Another challenge involving sensors, and truth be told, all technology development is energy. We are told that blockchain and cryptocurrency mining sucks up gargantuan amounts of electricity in their verification process, making the former slower than it should be and both much more expensive and redundant (with questionable results) than is desirable. Similarly, sensors have a relatively high locational cost factor on the front end to go along with the previously mentioned data aggregation and actionability factor on the back end.
Gilbert pointed to Disruptive Technologies, an IoT company in Norway, as a great example of the how reduced-cost sensors could cause the technology to more rapidly proliferate. “We are not an investor in Disruptive Technologies,” said Gilbert, “we’re just a fan. If you can create an inexpensive, wireless sensor that’s battery operated, to me that’s the future. What you’re talking about is a buck a sensor a month. Now suddenly, you can put a pressure sensor in a chair and on the floor and you know exactly how many people on that floor, how many people are in their seats and how long they’ve been there. From a space management, space utilizations standpoint, that’s hugely valuable data.”
There are less business, but more fun uses for sensors, too, said Gilbert. “When we were at MIPIM in the south of France last year we had these sensors linked to Nantum, and I have to give the Norwegians credit, they were actually sensing how cold the beer was in the fridge.”
Train your bot replacement and clean out your desk
In this Age of Anxiety, everything technological innovation, even sensors, can and should be examined as a job creator or a job killer. Not surprisingly, PropTech proponents fall heavily into the former camp.
“I think there is plenty of opportunity [created by sensors]” said Halimi. “There will always be property damage, so it’s not like that work is going away in its entirety. I can tell you that in our case, people who are already handy, already using their hands to repair properties, it’s a pretty simple training to be the technicians we use day in and day out installing our devices and being trained to see what some early indicators are and how to fix them. It’s really not that far of a stretch. In fact we are already getting folks from the restoration industry who are seeing the writing on the wall and are coming to us to learn our technology.”
”Totally creators,” said Gilbert. “I think if you’re talking automation, it’s different from sensor technology, in my mind. I think automation is definitely designed to decrease labor costs by definition, ‘let’s automate stuff,’ where sensor technology is just going to create more ability to create applications that are relevant to people, which is more job creation, more company creation, more programmers, more data scientists, all that kind of stuff.”
However, predictions of an utopian future of easier, more interesting and plentiful jobs due to technology, without any social tension or displacement, have at least as often as not proven incorrect, as the current world of Trump voters, French Gilet Jaunes (Yellow Jackets) and other scared, angry and increasingly protesting workers can attest. Sensors might not have such a negative impact as they become more ubiquitous, but . . .
“I’m of the belief that the next industrial or technological revolution might be the first that has a net reduction in jobs,” said Entin. “A lot of people continue to sell [the idea that] in every technological revolution the jobs were shifted from less labor intensive to more meaningful, and I think that’s true. But those were all absent machine learning. And when you get to true AI and machine learning, we’re obviously [for example] not going to drive cars at all.”
Entin hypothesizes that the transportation industry will, according to some observers, see job creation in the design and building of autonomous vehicles, “but robots are going to be doing it,” he added. And even though someone has to program the robots, “over the next 20 or 30 years, my children’s children will be affected by [automation].
He notes that, “At every turn we’ve been wrong. Globalization happened quickly and disrupted a lot of industries, though not nearly as many as it will disrupt. So you have a lot of workers who are in their 50s and 60s and near retirement who had a deal with society and the deal was broken. It was no one’s fault, just the speed of technology.”
Well, perhaps we took a wrong turn on the road of analyzing sensors in PropTech, however interesting and important the detour may have been. Before we fall too far down this dystopian rabbit hole, let’s conclude with a bit of speculation about what other societal outcomes may result from the ability to place sensors everywhere and all the time and whether that will mean an aggregate gain or loss for the individual.
How do you say “Can you extend my credit limit” in Chinese?
So all these sensors are going to be all over the place and produce untold amounts of incredibly granular data about everything and everyone. Who will control that data?
On the PropTech level, such data could be manna from heaven for landlords, if it can be aggregated and its insights gleaned for productive, aka profitable, uses. However, sensors produce data that is of great value to homeowners, insurers, public and private security services and many more entities in our lifecycle, as well.
In a more macro view, as sensors are indispensable to the IoT, and as most technologists expect that all things will truly be interconnected in the not so distant future, it is worth asking how such a confluence of technology embedded in and beyond PropTech will be used by governments around the world.
Generally, the PropTech experts to whom I spoke for this piece felt that the US government was neither a help nor a hindrance to sensor development, but was operating under what I would term an umbrella of benign ignorance concerning all technology. (If you need proof of such a characterization, I would direct you to YouTube for the Congressional hearings on Facebook and Google. Spoiler alert: It is NSFW if you’re in technology, or just an adult.)
McGill’s view of this subject was typical: “There are a few incentives for landlords to deploy sensor technology, but government may be more behind the eight ball than real estate. In general, there are few barriers and marginal support, so it’s basically a wash.”
As sensors intrude further into our offices, homes, cars (okay, everywhere) privacy concerns arise, even in the face of voluminous evidence that we have all basically long ago surrendered our privacy in the name of convenience and consumerism.
Notably, the EU has addressed technology privacy issues more comprehensively than any other part of the world through its adoption of GDPR, which exhaustively delineates the privacy rights of individuals interacting with tech. However, similar to the continually evolving rules and regulations of governmental bodies like the US SEC, it remains to be seen if the EU regulators will keep up with, or in the opposing viewpoint, over-regulate, technology companies wanting as much leeway as the market and laws will allow.
Then there is the more nakedly authoritarian end of the governmental interference spectrum, China.
When I asked Gilbert if we in the US are already “over monitored” by our government or the private sector due to sensor proliferation, he replied: “I think the better way to ask this question is, are we using the data we are already collecting in an intelligent way?”
While Gilbert doesn’t think we are over-monitored in the US, he points to China as what has long been a state-controlled economy hugely empowered by the ability to intrude on its citizens through sensors and the data that is derived from them, resulting in a “social credit scoring” system.
“If you’ve ever been in China, you buy everything digitally,” said Gilbert. “There’s very little money that’s exchanged. It’s all either WeChat or the Alibaba product. The government is literally monitoring what you’re buying and giving you a score. So if you’re buying too much alcohol your score is going to go down, if you’re doing ‘good things’ with your money, your score is going to go up. If you’re in debt, your score goes down. If you don’t pay your taxes your score goes down. People with low scores, [the government limits] your mobility. You try to buy a train or plane ticket and they say no, your score is too low. It’s crazy.”
While the above-noted activity doesn’t sound all that far from what US private credit and banking companies do to the consumer on a regular basis, the actions coming from an authoritarian government are, of course, much more draconian and chilling.
Now, are you not happy that we have decided to end this story by noting that it has mostly been about harmless little PropTech sensors and not autocratic Big Brotherism? Me too.
Watch out for that rabbit hole.