2020 STEX25 Accelerator Startups Day 2 - Startup Lightning Talks with Q&A, Session 1

MARCUS DALLHOF: Welcome. This is day two of the STEX25 showcase. Today, we're going to feature eight startups from the realms of sustainability, energy, advanced materials, and life science. My name is Marcus Dahllof. I'm a program director at the MIT Startup Exchange.

STEX25 is our industry accelerator. It's a 12-month program that features 25 of the most prominent, the most high-potential startups at MIT. But in addition, we have 1,900 startups that are part of the MIT Startup Exchange platform. These startups come from across all departments at MIT, and they employ all types of technologies to solve many different use cases across many different industries.

The agenda for today is as follows. We're going to start with four startup lightning talks. Each startup is going to present for about four minutes. We're going to take about three minutes of Q&A.

The first batch of startups include PolyJoule, Syzygy Plasmonics, Uncountable, and Stable. The same four startups are then going to go into panel discussion to talk about startup and corporate collaboration broadly. And we're going to repeat that structure with the following four startups-- Vaxess, Kytopen, Sweetwater Energy, and Mory, formerly known as Cambridge Crops.

We aim to wrap up around 12:35 PM Eastern Time. This is a very fast-paced event. We're not going to cover anything in depth. Rather, the objective here is to be the spark for the beginning of a conversation for the beginning of a partnership. Should you have any questions, comments, or feedback or want to engage with me after the event, you can reach me at MarcusD@mit.edu.

At this point, we're going to go to our first startup. The presenter is Eli Paster. He's the CEO of PolyJoule. Eli, please go ahead.

ELI PASTER: My name is Eli Paster. I am the CEO of PolyJoule. We're building high-power, organic batteries for the electricity grid. Glad to join you all. We're seeing batteries more and more on the electricity grid today. You see them in substations. You see them next to [INAUDIBLE]. You see them next to wind farms.

The problem is the majority of those batteries are lithium-ion. And we think that lithium-ion batteries are not the solution. They're still too expensive. They have the tendency to catch on fire. They put a deep economic and ecological footprint around the world. And you don't see domestic manufacturing of them. They're exclusive to certain regions of the world.

PolyJoule is trying to change that story. First of all, we try to cut costs everywhere that we can. So the cell, the pack, the system level, everything is defined by our chemistry. And the costs come down to that.

Secondly, we don't use lithium. We don't use cobalt. We don't use lead. We're trying to come up with an environmentally benign and an ecologically benign battery. That means it won't catch on fire. It won't poison the waters. And it's something that's recyclable and sustainable over the long term.

How do we do this? We do this through organic chemistry. So the majority of the stuff is organically derived. That means carbon derivatives.

And on top of that, because it's a safer battery, we save time, we save money, and we save space. You don't need air conditioning. You don't need heaters. This thing works robustly for 20-plus years.

But look at that use case for PolyJoule. So PolyJoule's high-power batteries are actually great in industrial applications. That could be server farms. That could be machine shops. That could be manufacturing facilities. That could be heating and venting and cooling for large refrigeration operations.

In our first particular case, we're deploying at an agricultural location where there's a solar farm. The solar peaks around the middle of the day, but the peak loads actually happen early in the morning and early in the afternoon. And PolyJoule batteries are essentially meant to capture all the possible solar energy and use the high-power capabilities to deliver power when it's needed most. That's in the morning and the PM.

Why did they choose PolyJoule over an off-the-shelf lithium-ion solution? The bottom line is because they're in agriculture, because they're around animals, because safety and environment is a core part of their features. They want something that's going to last longer and doesn't put environmental damage or damage the animals.

We're looking for partners in all types of areas. We like to say that batteries are a bit like salt or bacon. If you're not a vegetarian, they go well with everything. So that means industrial applications, wastewater treatment, telecom, smart cities, fast charging applications. They could be used behind the grid. They can be used at residential side.

The bottom line is if you're going to see a future where there is a grid that's distributed everywhere, it's got to be safe. It's got to last a long period of time. It's got to be environmentally benign. And it's got to be recyclable. And that's what PolyJoule is all about. Thank you very much.

ARIANA: Great. Thanks, Eli. So let's have some audience questions. First of all, could you elaborate a bit if there's any automotive applications for this, the PolyJoule?

ELI PASTER: The biggest automotive application is not for transportation. It's actually for EV charging. So imagine the equivalent of a gas station that charges up six vehicles at a time. That's an enormous amount of energy that you need to pump into EVs. So we're proposing that you put large PolyJoule batteries on there that can handle the load without having to recreate the electricity for it to handle this type of [INAUDIBLE]

ARIANA: Great. So we see a lot of energy storage startups coming up with a lot of different technologies. What makes PolyJoule approach a little different and unique?

ELI PASTER: Couple of different things. One, we don't manufacture our own batteries. We're the key to the chemists, and everything else we do is outsourced. So we're capital light.

Number two, we don't try to put the cart before the horse. So you haven't heard about us. We've been around for a long time, but the reality is we haven't announced coming out of stealth until we have an actual project deployed. So stay tuned over the next three or four months, and you'll will actually see real battery systems operating in a grid.

ARIANA: Exciting. Great. And what are some of the big trends towards long duration? Does that fall under PolyJoule?

ELI PASTER: PolyJoule's more on the short duration side. We see the grid as sort of forking in the road. There's a long duration, centralized electricity distribution network. And then there's more short duration. Short for us means 15 minutes to six hours.

Our real objective is to move quicker. So it's not a massive capital expense project that takes six years to build. It's low-hanging fruit that you can get into the hands of end users immediately.

ARIANA: OK. We're having a lot of questions coming in, so that's great. How fast can it be charged?

ELI PASTER: It's bi-directional. It can be charged and discharged in as quickly as 15 minutes. 100% depth of discharge.

ARIANA: OK. And what scale can you build by now? How do you see that?

ELI PASTER: 50 kilowatt hours, and it's really a matter of dialing it in. So we do not do our own manufacturing. So literally, if we had to go up to 500 kilowatt hours or 5 megawatt hours, it would be six months of planning and operations. That's it.

ARIANA: And one last question connected to that, then. How do you work with manufacturing partners?

ELI PASTER: Right now, it's more on the contract manufacturing and long-term objectives. So for us, for the manufacturing facilities, we're small potatoes. We take up three hours out of their week. And we say, can we get in the queue for more than that? They say absolutely. We're willing to grow with you.

ARIANA: OK. Actually, we have time for one more. Have you thought of other distribution outlets-- consumer applications, et cetera?

ELI PASTER: Absolutely. The real question for us is where to spend our resources, because we're a small company. So you will see us coming out to solar distributors for residential in about a year and a half. But right now, we're focused on larger projects where there's a greater need and a greater turnaround for the actual company.

ARIANA: OK. Thanks, Eli. Back to you, Marcus.

MARCUS DALLHOF: Thanks, Ariana. Thanks, Eli. Our next presenter is Trevor Best, CEE-- sorry, CEO-- of Syzygy Plasmonics.

TREVOR BEST: Thank you very much, Marcus. So my name is Trevor Best. I am the CEO of Syzygy Plasmonics. The Engine, which is a leading venture capital fund backed by MIT, co-led our last round. And this is why we're here presenting today. My co-founder and I started Syzygy with one premise-- drive sustainability through the energy and petrochem industries by changing how the world performs chemical reactions.

Our technology is broadly applicable. But we've decided to focus on disrupting the hydrogen market first. There is more excitement around hydrogen right now than any point in recent decades. However, there's still a number of key issues that need to be addressed.

First, hydrogen is still too expensive to compete with gasoline and other industrial feedstocks on price alone. Second, there are really only two options for getting hydrogen. Those are traditional steam methane reforming or electrolysis. Despite decades of research, no one has accomplished the huge leap, the dramatic reduction in cost, that's required to disrupt this industry.

And finally, hydrogen is not easy to decarbonise. When performing water electrolysis, you must use renewable electricity, or you don't make a positive impact on emissions at all. Yes, renewables are rapidly rising. But even at the current pace, it will be decades before electrolysis alone can meet the world's clean hydrogen needs.

Hydrogen from methane, the current approach of choice, requires very costly carbon capture to decarbonise. And then there are not many viable solutions for what you can do with the carbon dioxide after you capture it.

Syzygy's technology can solve these problems. We have cracked the code on photocatalysts-- that is, using light instead of heat to perform chemical reactions. Here, you see a picture of our revolutionary photocatalyst, which, I'm proud to say, our catalyst actually made the front page of Nature Catalysis last week because of its unique properties.

What this catalyst does is allow us to build a chemical reactor that is unlike anything else in the market today. Our first product is going to be a small-scale, on-site hydrogen production system that creates zero-emission hydrogen from methane. Because we use light to drive our reaction instead of heat, our system can be built out a very low-cost materials such as aluminum, glass, and plastic. This dramatically reduces our capital costs.

Our system will also be very simple and easy to operate. For example, when you turn the lights on, the system starts producing hydrogen in about a minute. When you turn them off, it stops.

Low-temperature operation and starting or stopping in under a minute helps us do things like grid following, makes maintenance a piece of cake. These items plus others combine to dramatically reduce our operating costs. Even better, we will use a fraction of the energy required by other hydrogen-producing techniques such as electrolysis.

Many of you are probably thinking, how can you make zero-emission hydrogen from methane? What do you do with the carbon? Well, it's simple. We turn it into something else. Our technology is also uniquely positioned to take the waste CO2 stream and process it back into useful products. We will likely start with green methanol but can make other things like DME, methyl formate, and so on.

So when you think about our business model, we will install a piece of equipment. We will sell hydrogen to the end user. And then we will sell the green methanol or other product to the market with zero emissions going into the atmosphere. This is what we're working towards.

Good examples of end users that would be interested in our hydrogen are those wishing to decarbonise transportation or material handling, those wanting to decarbonise small-scale fertilizer production, and those wanting to decarbonise their large-scale petrochem operations. Through our approach, we anticipate being able to get below $5 per kilogram dispensed into a vehicle. And we even see a pathway to $1 per kilogram zero-emission hydrogen for large-scale operations.

So things to remember about Syzygy-- we can make zero-emission hydrogen from methane. We have an incredible competitive edge in processing CO2 back into useful products. And we are actively looking for partners on demo units and field trials.

If you are a company that is interested in zero-emission hydrogen or CO2 processing and are curious about participating in our demo next year, please talk to me. If you're interested in field trials at your location in 2022, please talk to me. Thank you for your time, and I welcome any questions.

ARIANA: Great. Thanks, Trevor. We have a lot of questions coming up. First, when making hydrogen, how does Syzygy handle the carbon from methane?

TREVOR BEST: So we actually have another grant from the National Science Foundation right now for taking CO2 and turning it into syngas. Our catalyst is very resilient. And so at this time, we are experimenting with taking the waste gas stream from our hydrogen reactor, which contains CO2 and other things, and feeding it directly into another reactor that turns it into syngas that is a perfect ratio to turn it into something else, like methanol. We're calling this hydrogen plus. You get green hydrogen plus something else, and that carbon gets sequestered into the other product, which can then be sold to market with no emissions going into the atmosphere.

ARIANA: Fantastic. Another question-- have you looked at modular mobile applications, like automotive, et cetera?

TREVOR BEST: Absolutely. We are actually thinking about entering the market with material handling. And that would be forklifts, package trucks, et cetera, expanding into more transportation applications such as buses, semis, et cetera. And then once we have a foothold there, then expanding into large-scale industrial operations like ammonia, petrochem, et cetera.

ARIANA: OK. Where is the energy from when your CO2 turns to syngas?

TREVOR BEST: So that energy will also come from electricity. And so our entire system is powered with renewable electricity. And yeah, we use the renewable electricity to reform the methane. And then we use it again to convert the CO2 into syngas.

The total footprint of that system from an electricity standpoint should be lower than an electrolyzer to do both functions. And so instead of just getting hydrogen, you get hydrogen plus something else for a lower energy cost.

ARIANA: OK. Now, scalability-- how do you plan to scale this technology over time? And what's the maximum scale that you can achieve or foresee?

TREVOR BEST: So right now, we are in the process of moving from bench scale to a demonstration system. We are planning on building our demonstration system at the beginning of next year and deploying it next summer. That is planned at 20 kilograms of hydrogen per day.

After this, going into 2022, we'll be doing our field trials, which we're calling our beta unit, which is our MVP. And we are planning to 150 to 500 kilograms per day for that, which would be good for commercial transportation applications. After this, we do not see any limit on the scalability of our system. It is easy to mass manufacture. It is very easy to operate. And we see strong benefits at all scales for hydrogen consumers.

ARIANA: Great. Thank you.

MARCUS DALLHOF: Next speaker is Will Tashman, co-founder and CEO-- sorry, CRO, meaning Chief Revenue Officer-- at Uncountable.

WILL TASHMAN: Good afternoon and good morning, everybody. Like Marcus said, my name is Will Tashman. I'm co-founder and chief revenue officer of Uncountable. Uncountable is the leading lab informatics platform for material and chemical companies.

So one of the largest challenges in R&D today, whether you're making batteries, new coatings, new paints, new adhesives, is the diversity of data options for all of your different storages and all of your different metrics. The process of actually manufacturing these materials can be quite complex.

At the core, you start with raw materials. Raw materials have physical attributes and batch history and prices and regulatory. You move into formulations or compositions, which have various levels of complexity themselves. You have very advanced processing conditions. And you have analytical results that take up your majority of your performance conditions.

However, there are different data storage solutions and data management solutions for each one of these types of fields that creates a really hectic and scattered-- the placing of data throughout the system, which, one, prevents any real sense of collaboration between colleagues and also creates a massive data management task for the scientists. These are expert PhDs who you're spending a lot of money on. They spend a good amount of their time just digging through their own data rather than actually learning from it and trying to become more chemically proficient.

So the Uncountable platform solves that challenge by centralizing all data structures in the R&D environment. We provide a software solution that allows R&D teams to store all chemical and physical and material properties in one place where scientists can formulate. They can capture test results, capture test observations, and create reports directly in the platform. And this platform is completely searchable and filterable all across all parties and permissions granted to enable that ease of information sharing.

Now, with a centralized and structured set of data, we can move right into a set of advanced analysis tools. We'll allow scientists to prevent [INAUDIBLE] from formatting data and go right into learning from their data. And then finally, with this, again, structured foundation of data, we can leverage these advanced ML models which will allow scientists to predict how new experiments might perform and even suggest the next set of experiments to run given a set of priorities or topics.

In essence, the Uncountable platform allows R&D teams to do more with less. We want to create a single source of truth for R&D. So rather than 10,000 spreadsheets and 5,000 notebooks and all the different PDFs and reports that exist in the world, if we can find a way to connect all those different sources of data together, you create a much more cohesive ecosystem which will allow you to bring on new scientists faster. So as you hire new team members, they aren't digging through old lab notebooks. They're now being brought into an immersive knowledge-based environment.

It will also allow you to capture expertise from older scientists on their way out who are closer to retirement than to the beginning of their careers. This is obviously a bigger challenge now for many chemical companies that will see a lot of baby boomer-type scientists on their way out.

But in the end, this is about saving scientists and R&D teams time, money, and resources during development. We help our customers deliver products to market faster, move faster, and create new products and create better products than that.

As an example of a type of deployment and customer we work with, we worked with this large public engineering materials manufacturer. They had four different locations across three different business lines. And they were developing a bunch of different types of materials, both films, elastomers, electronic materials.

And previously, they had run everything either out of ad hoc, in-home systems or spreadsheets. And so by deploying the platform, we configured it. We trained their users and got them up and running and storing all their data and completely [INAUDIBLE] their workflows in the Uncountable platform in less than three months.

Since then, we've expanded to several different locations and added to the teams. But overall, the organization found that there was a massive reduction in the time spent of managing one's own data, which is usually somewhere between 30% and 40% of a scientist's time per week. That goes from maybe 15 to 20 hours down to just two or three hours.

And then there's a certain amount of time saved from sharing data across colleagues. Things that weren't possible before now become possible, so I can dig into these clearly marked experiments that are from different people across the world. And using the AI tools built into the platform, they were able to save many different mini trials over the traditional either DOE or ad hoc experimentation method used in many different environments today.

So Uncountable seeks new partnerships and new companies who wish to digitize and move to a more collaborative, structured data environment for their R&D resources. We have customers across the board from ceramics, batteries, and 3D printing to adhesives, coatings, paints, and pharmaceuticals.

ARIANA: Great, Will. Thanks a lot. So questions coming up for you-- there seem to be other single truth databases. How are you different or unique?

WILL TASHMAN: Yeah, absolutely. One, we've created a very flexible work environment that connects a lot of different systems together. We haven't really seen a tool that reaches across all the different aspects of a material development cycle from the raw materials and raw material lots all the way into processing condition and results in a way that's easy to use.

You're sort of stuck in these free-form flexible tools, or you have these very over-constrained SQL databases. There isn't a lot of wiggle room in between. The Uncountable solution provides that.

And our expertise is actually in ML. Both of our other founders have that expertise in ML. So we developed the data platform to provide the best data towards those ML engines. So not only is it easy to use, but it provides the information needed to develop these materials faster.

ARIANA: Great. How do you see handling the more complex requirements that pharma development has?

WILL TASHMAN: Yeah. So the platform is completely configurable. So it can work for a multi-step battery process. It can also work for a two-part adhesive. And so being able to have a flexible, configurable system allows us to deploy our platform to many different business units or material types.

ARIANA: OK. And why would you say that digitalization is so much more important today for large industrials?

WILL TASHMAN: Yeah, I think everyone's aware of the critical importance to digitalization. Obviously, during the COVID era, the ability to collaborate seamlessly with colleagues is almost a necessity. But in general, scientists and R&D teams want to do more with less. And that's what software tools are allowed to do is not replace scientists but enable them to focus on the more critical chemistry aspects, because their challenges are only getting harder, not easier.

ARIANA: A last question. How do you ensure data privacy and security?

WILL TASHMAN: Absolutely. Privacy and security are our number one priority as a company. We recognize that this is the crown jewels of each of our customers' information. We have very, very strict policies in place both internally and with our own software infrastructure to prevent any information leakage. We store all of our customers' data completely separately, so there's no overlap. And we go through the industry best practices like SOC 2 Type 2 audits to ensure that we're maintaining a very high level of security in addition to other things like whitelisting IPs, two-factor authentication, and least privilege access.

ARIANA: OK. And on size, what is a typical organization where Uncountable makes the most sense?

WILL TASHMAN: Good question. As you can see from our previous customer list, we have customers from small startups all the way to large, 20,000-person international organizations. We always follow a crawl, walk, run type of deployment to our customers.

But most of our customers are in the-- I'll call it 1 billion to 10 billion market cap range, which usually means there are somewhere between 1,000 and 5,000 total employees. And the R&D team might be 200 to 800 people.

ARIANA: OK, thanks a lot. And Marcus, back to you.

MARCUS DALLHOF: Thanks, Ariana. Thank you, Will. The next speaker is Rohan Puri, co-founder and CEO of Stable.

ROHAN PURI: Thank you so much, Marcus, and good morning. At Stable, we're about building the foundation of electric future. And we do that through infrastructure optimization. So for us, it's all about accelerating an introduction of commercial electric fleets to the market.

As Marcus said, I'm co-founder and CEO of Stable. My co-founder, Jamie, and I are originally from MIT's Media Lab. We worked in a group called Camera Culture there alongside many different OEMs working on their next generation vehicles.

The central problem around commercial electric fleets and their infrastructure needs is that you meet a lot of infrastructure to serve a lot of vehicles. And infrastructure is not cheap. Existing public infrastructure is totally inadequate for commercial fleet use. They're poorly located and offer no guarantee of availability. And that's because we put almost all of our chargers at homes, offices, and shopping centers. Again, great for the general public, not very useful for commercial fleets doing taxi, delivery, ride share, and things like that.

The other problem in infrastructure is it's complex and expensive. It's hard to figure out how many charges you need, how many should be slow, how many should be fast. And it's costly to own and operate those chargers. A DC fast charger, for example, ranges anywhere from $30,000 to $60,000 and can pull as much power in a single day as the average American home pulls in six months.

This picture is a great picture. It shows BYD's charging facility in Shenzhen, China. If you didn't already know this, Shenzhen's taxi fleet is already 99% electric today. And this is how charging is conducted in a centralized hub.

What we do at Stable instead is we use software to help charging providers, CPOs, and fleets deploy charging infrastructure at minimal capex and opex. We take into account their fleets' own activity dynamics, their vehicle composition, the geographic parameters of the area to minimize the number of charges-- just sort of the largest number of vehicles at high utilization for those fleets.

And I'll show you what this looks like. If we were going to do the same model that they did in Shenzhen with a giant centralized charging hub full of fast chargers but in this case in San Francisco with an all-electric ride share fleet-- that's 6,500 vehicles-- what you would see is a large number of vehicles flooding out of that centralized hub in the morning. Blue dots are good. These are vehicles that are generating revenue and well utilized.

But as the day goes on, most optimally, more and more of those vehicles turn red. Red is underutilized vehicles, non-revenue-generating miles as they're returning back to that charging hub. What you end up having is this double inefficiency, a very underutilized fleet and a very underutilized hub of chargers.

And so that's very inefficient from a charging perspective. And this is actually using real ride share fleet data over a year-long period simulated in the cloud. What we do at Stable is trying to optimize this.

So if you were going to approach it with a centralized strategy, even doing two to three of these charging hubs in a city like San Francisco for a light-duty fleet, you would need over 1,250 $30,000 chargers. But if you're very clever about where you place them, you place those chargers in places where the zoning is permitted. You place them in places where there is enough electrical capacity or there's more likely to be electrical capacity. The traffic works out. The fleet activity works out.

It turns out, for that same data set, you'll only need around 300 chargers to serve 6,500 vehicles. We've essentially reduced the capex demand by 4x here. The only trade-off here is that you have to distribute those chargers in such smaller numbers, which, from an energy perspective, is a lot easier to do, right? It's a lot easier to run five chargers than it is to run 1,000 at one location.

So at Stable, we have StableNet software. It's network optimalization analysis. It allows us to do four main things-- compute the optimal network size and distribution, the best case network in charging utilization for a given site, the best case fleet vehicle utilization, and the payback period and costing models for that infrastructure. You're going to be spending a lot of money on your infrastructure. Best that you plan where it goes and how much it will cost.

There's an example of what we use. We're working with a CPO for new builds in San Francisco. We computed this analysis to figure out the best location for two commercial fleet customers to deploy their charging infrastructure. And the CPO said, hey, we would love to deploy chargers at that site.

Stable is not a hardware owner or operator. We do not deploy our own charging stations. We work with people who are trying to do it. This is a commercial site we're working on. And the mission, 10 150-kilowatt DC fast chargers for commercial fleet use to be shared across a couple fleets.

How do you work with us? Two main ways. Do you know a fleet looking to electrify? Stable can help calculate the lowest cost of infrastructure for those fleets using our dedicated software tools.

Or do you know of a charging provider looking to serve the fleet dynamics? Of course, we're working with some of the biggest fleets in the Bay Area and California. We're now also starting to bridge into New York City. We know where the infrastructure is most likely going to be needed. We have the fleet data. And we know where to serve these audiences. So I'd love to chat further if you have any more questions. Thank you.

ARIANA: Great. Thanks, Rohan. First question for you-- traffic patterns change all the time. How do you optimize? And you touched on this.

ROHAN PURI: Yeah. So traffic patterns change all the time. Fleet activity doesn't. So what you're trying to do is take a global, one-year view about how things might change week to week, month to month, year to year and where charging infrastructure should most optimally be placed.

The problem with charging infrastructure today is because they're located at retail locations and workplaces, they're off the beaten path and dramatically affect that infrastructure. So we actually simulate many, many different scenarios with variability in between to figure out, what is the common set of infrastructure where it's most likely you're going to get high utilization of charging stations and high utilization of your fleets?

And if you think about it, those two things count on one another, right? If you're charging a lot, you're probably not driving very much. So finding that balance is actually a very tricky thing. But it's worth the planning exercise, which most charging providers aren't really doing. They're just tacking it onto retail partnerships.

ARIANA: How would you say that your solution could help the broader consumer market, as well? Or do you mostly focus on commercial fleets?

ROHAN PURI: We focus on commercial fleets. However, we have had customers come to us saying, hey, we'd like a site that's open for both commercial fleet use and public use. That is a site that we're working on in New York, actually. And so in that case, we actually had to take into account public data about where EVs are most typically operating and merge it with the needs of the commercial fleet data to figure out a common location that might work well for both, taking into account the specific zoning and permitting laws and the electric capacity constraints of that location.

So it's totally possible. It does make it complex, because fleets needs a guarantee of availability. They want to make sure that when they arrive, a charging stall is actually there for them. And if the public has access, it makes it harder to do so. So there's some hardware you need onsite, and that charging provider was willing to do that.

ARIANA: OK. How do you collect the data for charging? And connected to that, how long does it take for you to come up with the optimal charging solution in a new city?

ROHAN PURI: Yeah. So how do we get the data? We have a mix of public and private data sources. So there's beautiful public data sets, because a lot of the cities have mandated the fleets return this data, especially in the ride share segment. So we've seen that a lot.

But we also have private data sets. As you can imagine, many, many fleets are looking to electrify. And they're looking to have infrastructure put in place for them. Most fleets don't want to own and operate the infrastructure themselves. They're willing to give a little bit of data to make sure that that charging infrastructure is in a good place for them. So we actually haven't encountered many issues on that.

How long does it take to onboard a new city? This depends really on the city. In the US, it's pretty straightforward. It would just take a month or two, depending on the data sets that we can acquire. Really, looking at new cities around the world in Europe, Asia, that'll be a little bit harder, because the data is available through different channels. And so we work with our partners-- typically, the companies who are deploying there-- to get whatever data we have. And if we don't, we simulate according to what we know.