"The challenge is how do you get out of that sea of data into making decisions, with AI, that are actually going to build the value, not the drudgery of analyzing."
Listen to the full episode.
Adam Hooper: All right, Tom, thanks for joining us at the CREtech 2019 show today.
Tom Arnold: Thanks for having me.
Adam: Perfect, so before we talk about Gridium, why don’t you tell us a little bit about your background, how you got into this energy space and AI and we’ll start from there.
Tom: Yeah, I started actually in hard tech doing data networking and very interested in how things work. And about 20 years ago, we started a company that started to do small energy projects on dairy farms and wind farms and that was my exposure to the grid and I’ve always been fascinated in the grid and how certain things like buildings interact with the grid.
Adam: So energy projects on small farms.
Tom: Quite literally, like dairies and piggeries. That was 20 years ago. Now I’m looking at condenser loops and lights and peak demand charges and chillers–
Adam: All that good stuff.
Tom: So a very different world. But we have about 2,000 buildings across the United States that use our software and our job is to help them get more from what they’ve currently got, either lower expenses or generate new revenue opportunities.
Adam: Good, and so how has that landscape changed from small farm projects looking at energy consumption there to what you guys are doing today?
Tom: You know, it’s the same themes. To really understand energy, you got to understand the prices, you got to understand the data and you’ve got to use that data to make smart decisions about how you operate.
Adam: Yeah, and that’s a consistent theme that we’ve been talking about here and on the show is the availability of data today is infinitely greater than it was even just a handful of years ago in the real estate space. I’m assuming in the energy space, similar, just the availability of data–
Tom: Absolutely, almost every building in the world now has a smart meter on it. Your work order data is accessible, as well. It’s actually a sea of data. So the challenge is how do you get out of that into making decisions and that’s where AI comes in because you can have a sea of data and deploy AI against it, get the diagnostics and recommendations that you want and focus your time on doing the things that are actually going to build the value rather than the drudgery of analyzing.
Adam: Okay, so AI, artificial intelligence, that gets talked about a lot in the industry. Why don’t we go 50,000 foot, tell the listeners what is artificial intelligence, at least how you guys use it. What are maybe some of the misconceptions? Are you turning buildings into machines that are going to come eat us? What is AI in your perspective?
Tom: AI is a big umbrella term for a lot of tools. We use the lower level machine learning techniques that are well established in the industry. There’s nothing going on here that’s out of Terminator 2 or anything. We use a bunch of pattern recognition. For example, one very simple thing that we do is we have taught the computer how to read a load curve and how to match a particular pattern that we’re looking for such as a chiller or hard start or a cycling of a unit and feed a message back to an operator that says, “Hey, Thursday it was 58 degrees and the building was cycling.” And that tells the operator, wait a second. My deadband’s probably not wide enough on the control system. Let me go fix that. Now, if you were to sift through hundreds of pages of load curves, you wouldn’t find that on your own. So it’s really about taking that data, turning it into diagnostics and linking it back up with an operator so that it can be corrected and build value for the–
Adam: And so that’s ultimately the goal of AI or machine learning is being able to sift through vast amounts of data that a human wouldn’t even necessarily be able to find those patterns or just doesn’t have the manpower or time, energy, effort to go find those.
Tom: Yeah, I think Minna said it on the panel as well. She said it very well. It’s basically looking to take the human out of the awful part of the job and put the human back on the high value part of the job.
Adam: Automating a lot of those, the grunt work basically so that we can still use our human brain to make insights from that data.
Adam: So what does a building owner get from this? Is it primarily for, with Gridium, is it primarily for building owners? Is there a benefit for the tenants–
Tom: It’s for whoever runs the building. Our typical user is a property manager or a chief engineer team. It’s a very easy to use SaaS application. There’s nothing to install on the building, so we don’t have to touch the building and we’re going to essentially leverage all the existing data streams to supply that advice to you. From there, we often find pay dirt. So a part of onboarding a new client is doing a series of audits that look at your operations, it looks at your rates, your bills, your energy choices, And it’s very common for us to find very strong savings right out of the gate. And that gets you into the Gridium program. And then from there, we continually optimize.
Adam: And so this is a more efficient way to manage energy, so it’s a savings, cost savings. So there is benefit to the users of the space, too. More efficient usage, obviously, is going to lower their bills as well.
Tom: Yeah, exactly.
Adam: As you said, you have about 2,000 buildings.
Tom: That’s correct. It’s about 200 million square feet all across the U.S.
Adam: And so given your start in small farms, are you going for individual building owners or is this more institutional scale roll out? Who’s the target user right now for the product?
Tom: These types of solutions only really work if it’s a partnership. And so even in a big portfolio, we start at the bottom of the account. We’ll go meet chief engineers, we start to understand buildings and we develop that trust. And that’s very important because if you don’t do that, if you try to roll proptech out from the top, it often never really penetrates the organization. If it’s a bottom up and has support from everybody in the organization as a useful part of their work, then you have the case for persistent value.
Adam: Good, any geographic preference? Do you see more savings in markets that have more…
Tom: We’re saving money in Seattle, which is seven cent energy. And they’re saving in Hawaii which is 29 cent energy. There are lots of tips and tricks that are unique to each utility and how it’s priced. For instance, a lot of our customers love our predictive demand alerts because demand charges are such a huge part of the building expense. And that’s going to work in almost every single territory just ’cause of the way commercial energy runs.
Adam: What are some interesting things that you guys have seen from again operating in Phoenix versus operating in Seattle? Are there any insights that you guys have learned through the system that you might not have expected?
Tom: Well, one fascinating trend right now is that lots of commercial companies, commercial real estate companies feel like the cities are going to make them turn their buildings into all electric buildings. So a market like Seattle is fascinating because it is actually almost an all electric market. You see these massive literally megawatt scale boilers in high-rises. And that’s very different from Phoenix where it’s so dry that all you need to do is run the cooling tower and you can get away with a lower energy expense.
Adam: You mentioned the cities trying to bring maybe different regulatory compliance things into place. How much of this is a tool to be able to future proof legacy systems or how does that work playing in with the dynamic from the regulatory side?
Tom: A lot of what we work with our bigger clients on is the energy strategy, understanding where the policy is going and also understanding what the opportunities are. So last week, we just announced Gridium Alpha. We purchased a unit of a French energy company. And Gridium Alpha helps building owners connect their wasted energy with the grid. So we finance, we originate, we do the project in the building and then we allow the building owner to sell that energy back to the grid. So this is a new initiative but I think smart owners who understand that buildings are a part of the solution to both climate change and changes on the grid. And they’re positioning themselves to benefit from those linkages.
Adam: And so tell me about that acquisition now. So this was you bought a unit of a power company?
Tom: Yeah, Engie is a large French energy company. They were actually, we were selling software to them for this unit. They decided to make a strategic shift and we said, listen, we want to bring that unit over to Gridium.
Adam: And was that in the plans or that just happened?
Tom: It literally just happened.
Adam: So you didn’t plan on buying part of an energy company when you started this?
Tom: No, but it’s quite interesting and you’ll hear a lot more about it. We’ve got some new announcements queued up this fall.
Adam: So you were selling the software to them–
Tom: That’s correct and now we’re going to do the software and the financing and bring that in to the Gridium customer base.
Adam: And so are energy companies and producers, are they a potential user of the software going forward?
Tom: If you’re interested in retrofits, yeah, essentially. Here’s a good example. We have a program upcoming with a utility where they are trying to get rid of a power plant in a metropolitan area in a very congested part of the grid. And they had asked us to go procure energy waste from the buildings next to that power plant. And so we can link those incentives and make it work with a commercial real estate structure so that the entire grid benefits.
Tom: And we, just to bring it back to AI, you couldn’t do that five years ago because the AI models on buildings weren’t sophisticated or stable or tested enough and now they are. And so the efficiency resource trend is what we’re getting on here, which is you’ve got supply resources on the grid and you also are going to have efficiency resources on the grid.
Adam: And so how does this scale or how big does this get? I mean, again, if you’re potentially working with the energy companies themselves, how integrated does that system get from production all the way down to utilization in the space?
Tom: I think there’s a huge opportunity. If you look at what has to happen to achieve the state policy goals, buildings are 40% of our energy use. And if buildings aren’t part of that equation on the reduction, that’s more and more renewables that we have to build on the other side to meet the state obligations. So it’s going to be a lot cheaper to cut reduction due to LEDs, upgrade the chillers, get the air distribution right in the buildings. I mean, we still have pneumatics in a huge percentage of the US buildings. Get all those retrofits done. That’s going to be a lot cheaper than building big solar power plants in the desert.
Adam: In terms of just reducing the total energy usage.
Tom: Exactly, for the grid and backhauling that to the city centers.
Adam: And what sort to efficiencies or savings are you guys seeing in some of these cases?
Tom: So in these cases, where we can actually bring money to the building and finance things up to 30, 40% total energy use reduction.
Adam: That’s fantastic. Yeah. And that’s something that’s got to be pretty effective to the bottom line, too.
Tom: Yeah, exactly. So essentially that allows us to finance your sustainability goals but without you touching the capital stack on the building.
Adam: Sounds like a pretty good scenario. Perfect, well, so how can listeners learn more about Gridium?
Tom: Just visit gridium.com, there’s a bunch of resources there, ask your chief if he’s heard of Gridium and if not, let’s come see him or her.
Adam: Perfect, all right, well Tom, thank you for joining us today. Enjoy the rest of the show.
Tom: Thanks, Adam, appreciate it.