Stable Auto

ROHAN PURI: My name is Rohan Puri.

JAMIE SCHIEL: I'm Jamie Schiel. I'm one of the founders of Stable Auto.

ROHAN PURI: We build SAS software that helps EV charging station operators predict the utilization of their charging site before they put all the money down to make it work.

JAMIE SCHIEL: My co-founder Rohan and I, we met at the MIT Media Lab working at the intersection of machine learning and sensing.

ROHAN PURI: Yeah so, my co-founder Jamie and I are originally from the Media Lab. We were researchers there in Camera Culture group, working at the intersection of sensing, machine learning. And we noticed the trend of this convergence of autonomous vehicles and electric vehicles, and the pure lack of charging infrastructure that was available to support those fleets and those types of cars.

And there's a sort of classic chicken and egg problem of, we've got to get the infrastructure in the ground first and then we got to get the-- then the vehicles will arrive. But now we kind of have the opposite problem, where we have some infrastructure in the ground, and it's not really being utilized well enough and so people are leaving the business. And we're just never going to get there to a net zero climate economy if we don't make that work.

JAMIE SCHIEL: A lot has been said about the enormous number of EV chargers that need to be installed for transportation to become fully electric. Something that isn't often discussed is many of the chargers that are already out there are seldom used.

So clearly we need more chargers, but they need to be in smarter locations to better serve that demand. And better performing stations is a win for the operators of those charging stations, a win for consumers, and ultimately for the planet.

ROHAN PURI: EV charging station utilization, especially for fast chargers, is incredibly, incredibly low. Basically, what that means is that people aren't using those chargers nearly as much as they need to to justify the investment for those chargers. A typical fast charger, like a super charger, costs about $100,000 all in get installed in the ground. And when people are using it maybe 20, 30, 45 minutes a day, you start to figure out that maybe it's not worth that investment at all.

So we have a very similar problem in a very tangential industry, in the retail industry. There's no end to the amount of data science and analytics that goes into where you put the next Starbucks, or McDonald's, or Subway, and you take into consideration a bunch of things. The demographics in the area, the traffic volumes, competitors nearby, so why can't the same thing exist for charging? That's basically what we're doing.

So the existing solutions out there basically are simulations. They're trying to look at what is the likely vehicle penetration rates in 2025, and 2030, and 2035. Make a whole set of assumptions, and then they kind of figure out, OK, based on how many vehicles we think are going to be there, and where they're going to go, we think we need this much charging over 10 years.

The reason that's tricky is that it's just sort of a business model assumption. It's not really based on historical data. So we actually use actual charger utilization data, historical information about how charging sites have actually performed, and then use that to predict how a new charging site will perform accordingly.

JAMIE SCHIEL: When we started, it wasn't at all clear that utilization could be predicted from things like traffic, local amenities, EV sales, down to the street address level. And it turns out it can be, but it's incredibly complex. And we've built all of that complexity into a very, very easy to use software package.

Anyone who's interested in profitable EV charging investments can use this software, whether they have understanding of machine learning or not. So that's chargepoint operators, utilities, local governments, lenders, and really anyone who's interested in a fully electric future and the well-being of the planet.

ROHAN PURI: We've proven that utilization can be predicted. You can actually try and predict what the utilization will be, especially in that first year or two of operation. Now of course, a lot's going to change over a 10 year period, and so you need some more assumptions to sort of extrapolate it out further.

But we've proved it's not a crapshoot. You can actually predict what the utilization will be for a new site, and so it's the responsible thing to do. Before you put a million dollars into a 10 charger site, just make sure that this thing will actually be utilized and will actually have the impact do you think it will.

Anyone who is owning, operating, or considering hosting EV charging stations should care about this. Basically what this means is, is this investment worth your time? Is this thing ever going to pay off, and is it going to bring any value to you?

If you're owning and operating, you're typically a charging network, a charging provider, or utility. You're putting the infrastructure in the ground, you want to make sure people are going to use it, and that you're actually going to make some money on it, or you're going to pay off on that asset.

If you're a site host, and you're a retail location like a grocery store, or a mall, or shopping center, you want to make sure that you're giving up these parking spots, it's actually going to bring-- that's actually going to add some value to your customer base and that those charging stations will be well used.