Uber's electric bet on electric vehicles. What does the rise of EVs and autonomous vehicles mean for the future of mobility?

At Uber, we almost think about the AV ecosystem as existential for us, right? So if we think about where the future is going, we think the future will be electric, autonomous, and shared. If you look at London as an example, 40% plus of the miles are EV. It just becomes our core business. What we're seeing actually is like, in markets where we have AVs, earnings for drivers have actually increased. And in markets where we have AVs, the overall growth rate for those markets have also increased. This episode is brought to you by ACOR, the non-partisan, non-profit organization, uniquely operating at the intersection of energy affordability, reliability, and clean energy deployment. ACOR is focused on strengthening the electric grid and driving clean energy investment that delivers for the American people. ACOR's membership includes industry leaders across the clean energy economy. Utility-scale clean energy investment has been booming in the US. Nearly 80% of it was financed, developed, owned, equipped, or contracted by ACOR members. Visit www.acor.org to learn more about ACOR's work and upcoming events, like the ACOR Finance Forum on May the 12th to the 13th in New York City. Hello and welcome to The Energy Gang, a discussion show from Wood Mackenzie about the fast-changing world of energy. I'm Ed Crooks and on today's show we're going to be talking about EVs and ride sharing. With fuel costs soaring because of the Iran war, the case for electric vehicles in terms of national security and economic resilience is very much back on the agenda. But of course, the industry still faces many challenges. And in some click, in some countries, including the United States, sales have been faltering and things look really very difficult for EVs. So the question is, what can be done to accelerate EV deployment? And in particular, we're going to be talking about autonomous vehicles, AVs, and asking whether they are really the key that is going to be unlocking the future of electric transport. To talk about that, I'm joined by Amy Myers-Jaffe. Amy is the Director of the Energy, Climate Justice and Sustainability Lab at New York University. Hi, Amy. How are you? I am good, Ed. And you know, this is my favorite, favorite topic. So really excited for the listeners to hear who our guests are. Yeah, absolutely. Really interested to get into it with you. And also, as you say, with our two guests, because it's a pleasure to welcome two guests from Uber. Andrew Cornelia is Uber's Global Head of Electrification and Sustainability. And Samarth Kedrawal is the Global Head of Fleet and Autonomous Vehicles, also with Uber. Hello, both of you. Welcome to the show. Hello, guys. Thanks for having us. Thanks very much for coming on. So look, something we always like to do when we have new people on the show is get them to talk a little bit about their careers, how they got into energy, how they got to the positions they now hold. So perhaps we could just do that briefly with both of you. I mean, Andrew, maybe to ask you to start off, what's your story? How did you get into working on electrification at Uber? Well, so I can tell you that I started my career at Tesla back when it was a one car company. I always joke and say that this was before people knew how to spell EV. But I worked across all areas of the business. I worked on the vehicle business. I worked on the energy business. And I gravitated towards the global charging business, thinking that that was one of the major challenges to solve for EV adoption writ large. My curiosity took me further down the rabbit hole of charging. I joined a Series B company called Volta Charging. Over a couple of years, we took that public. And then after that, I joined as the founding CEO, a joint venture between Mercedes-Benz and Emanate Energy to build a high power charging network in North America, bringing quality charging to all drivers. And then I got a call from Uber and it's hard to turn down an offer to join a platform of our scale and global reach and innovation. So I've been with the team about four months now leading the global electrification and sustainability organization. Right. Fantastic. And very interested to hear more about, as you say, Uber strategy in just a moment. Before we do, though, Samarth, what's your story? How did you get to where you are today? Yeah. So my current role is I lead our autonomous fleets team. And so the team's mandate is basically the day-to-day operation of the actual AV fleet deployment on the ground. So the team working at the depots and the pitstops, but then also a big part of the mandate is all the infrastructure work we're doing to actually stand up the infrastructure required to deploy AVs. Before this, I've been doing my current role for about two years. Before that, I had the pleasure of running strategy for Mobility at Uber. So working very close to Andrew McDonald, who's our CEO. And then before that, I did a variety of ops roles at Uber for the last, my first four or five years at Uber. So this will be my 10th year at Uber, which is longer than I thought I would be here. And then before that, I was in investment banking, which I did not like very much. But very glad to have escaped it. Yeah. We're glad you're here in AVs because it ensures that I'm going to be right because Ed is always was skeptical of AVs and I'm like a big no, they're coming. So I have confidence in you, Samar, so glad to hear it. Indeed, indeed. Very true. Thanks very much for joining us today. And so we're going to be talking to you really more about the AV side of things, though. Before we get onto that, I want to talk about EVs. And Andrew, perhaps pick up first of all with you on that. Let's start at the beginning, maybe in terms of Uber's approach to EVs. Why do you have an electrification strategy? Why is it important to Uber as a company? Yeah, Ed, I would start by saying, you know, we have a firm conviction that the future will be electric and autonomous. And in a large part, autonomous means electric. Why we believe that? we see a benefit for our earners, which are our drivers, as well as our riders, and long-term for the business itself. And it really fits into two buckets. One is we believe that the electric experience is a higher quality experience. They're newer vehicles, they're quieter vehicles, they're more technology forward. So that benefits our riders. Our riders are actually valuing our electric product at one of the sort of top levels of all of our products across our portfolio. And the second reason is because it will cost less over time and the economics will be better for the organization. So in markets today where the total cost of ownership works, our drivers are earning more or saving more. And we see that trend line continuing to progress globally. Right. And so to be clear about that, then that's just because total cost of ownership is lower for an EV when you add in everything, fuel, maintenance, and so on. Cost of ownership is lower and riders are typically making what about the same i mean i feel like when i uh look at a ride on the uber app it's kind of if i get an uber x or an uber electric it's about the same price right so their their revenues are about the same but their costs are lower is that the point that's correct yeah what we're seeing in select markets today is that the total cost of ownership so everything from fuel to maintenance uh to the actual cost of the vehicle over the life of the vehicle, especially the way that our drivers ride and use, will be lower compared to an alternative. Right. Andrew, and let me ask you, because I've discussed this with you in the past, you've surveyed riders, so you have a sense of why riders are preferring EVs when they make that selection, if they do make that selection. Can you just share that with us? I mean, what are the drivers there? Yeah, I'll give you a real-life example. I was recently in our UK office and we have what are called green light hubs where we aggregate some of our drivers in real life to solve problems like registration or licensing. And I sat down with a number of those drivers and we asked that exact question, Amy. And the feedback we get is, you know, they're owners of their own P&L. So, you know, they will comment on the maintenance being much lower. They'll comment on people who have access to home charging, being able to save close to $8,000 a year compared to the public charging. And then people who are using public charging, benefiting from some of the programs we have, like significant promotions and discounts compared to some of the public rates. So, you know, we today have one of the largest EV platforms across all rideshare, and there's a reason that we do. And let me ask you, and then how about on the passenger side? Have you pulled the passenger side? We have. I mean, I think data and insights is kind of Uber's secret weapon here. We have many years of collecting these types of insights. And what we consistently see is that the electric product, which is an actual product within our product portfolio, and you can select that in many markets as well as electric comfort, is top three within our rider satisfaction scores. And it goes back, Amy, to what I said before. It's quieter, it's smoother, it's more sustainably oriented. So we're very much seeing that in the data play out. Yeah, I have to say, I can totally believe that. I will very often choose an EV, not particularly because I want to be green, although obviously that's important and useful. And depending on where you are in the world also, obviously there's a big debate about how much greener actually using the EV is, if the power comes from gold, whatever. Anyway, putting all of that to one side, as you say, just as a rider experience in terms of the smoothness of the ride, the quietness of it and so on. As you say, I would often choose an EV just on those grounds. What about in terms of what you do to incentivize drivers to use EVs? Because am I right in thinking, I mean, do you make payments or used to make payments? We don't anymore. Or there's a there's a kind of a lump sum that people can get in some places. I mean, how does it actually get set up? Yeah, I would say that our general ambitions to help accelerate the adoption of electric vehicles has not changed. Our strategy has. Historically, we very much incentivized drivers through direct funding to make the switch. So today we have a program in the U.S. called Go Electric. It provides monetary incentives to a select cohort of drivers to help them reduce their upfront payment of vehicles. Going forward, the change is really thinking about what are the challenges for EV adoption and where can Uber, with both our catalystic involvement as well as strategic investment, really help tackle some of these challenges and barriers. So I'll give you one example. We have multiple partnerships with OEMs around the world to offer unique and special discounts for our drivers that are only available through our platform. So that would be one example of how our interventions help make it easier for drivers to get into an EV. Right. I mean, for instance, I saw there were some headlines. You announced this deal with Rivian the other day, in which case then, actually, that's a good opportunity. It's a good segue to talk to you, Samarth, about, again, sort of Uber's motivations in exploring AVs and thinking about that. Why is this something you're investing in and supporting? What do you think is the payoff to you from developing AV capabilities? Yeah, I think at Uber, we almost think about the AV ecosystem as existential for us. So if we think about where the future is going, if you think of how Andrew started his answer, we think the future will be electric, autonomous, and shared. And in many ways, we're thinking about a meaningful percentage of miles currently being driven by humans, which will eventually be driven by AVs. And we do think that, similar to how Uber started, where Uber came in and the category grew, we feel like something similar will happen with AVs where AVs will come in, the price per mile will go down, the overall gap or the overall market will increase and still be more human drivers, but also a lot more AVs on the road, right? So as we think about the next five, 10, 15 years, you know, we definitely think AVs are going to play a much, much bigger bar. And you're seeing that already, right? Like, you know, you go to San Francisco, you see Waymo's all around, you see the Zookses. We actually launched Dubai this morning with WeRide. So we're seeing that in Dubai and Abu Dhabi as well. So we're seeing it on the ground now. And you'll see that trend just increase with every passing month. I mean, it's kind of an interesting thing because talking to both of you about the strategies in the context of a oil shock war, where we don't know how long this shock is going to last or what the long-term ramifications are going to be. you know, I see kind of like ride sharing and especially AVs as being kind of like where e-commerce was kind of important and interesting. And you could see that there was a long term trend. But then all of a sudden we had COVID and e-commerce took off like a storm. And remote working as well, being another example of that, right, where the technology all existed, but people didn't use it nearly as much as they use it now and are continuing to use it, even though COVID is well behind us. Yeah. So if I don't want to deal with gasoline prices, you know, and Uber's fleet, whether that's an AV fleet in a city or whether that's a driver fleet, you know, in other different kinds of all different kinds of communities, if there's some EV driver that can pick me up and that means I don't have to go buy gasoline. I'm thinking that that could be a good thing in the next couple of months. You know, I can answer maybe just from a human powered vehicle side. I think, look, Amy, you know this better than anyone. And, you know, the stability of electricity prices, it's not completely stable compared to what's going on in the world today, but it's much more stable. So, you know, if you have a home energy rate, you know, that is a rate tariff that you're locked into for some time. So the best place to charge, the cheapest place to fuel compared to any types of vehicles on the road is at home. If you can plug in your vehicle and the second place right now, at least compared to how fuel prices have increased at the pump, So fuel prices at the EV charging stations are relatively stable over the last month. So yes, it has been more stable compared to what's going on in the world today. And what about with AVs, Samoth? Is that going to be something that is going to be also accelerated, do you think, adoption of AVs by oil price shocks like the one we're seeing at the moment? I think it will be a longer term adoption, Ed. I think there'll obviously be some near-term shocks. but the reality, like we don't have that many vehicles on the road for it to actually be, you know, meaningful behavior just because of the oil shocks. I think from the AV side, I think the more direct impact you'll see is like running an AV, like the total cost of operation, what we call the TCO, it's just way better, right? Like there is a reason why all the AV companies are electric and just the ongoing cost of running these cars 16, 18, 20 hours a day, it just makes a lot more sense to be electric right and again these cars are doing a lot more cars than the average human driver right like we're doing three shifts a day um you know it's it's it's it's high high intensity and and running an ev just makes it a lot more economical that's really interesting i haven't thought about it because of course in principle and i think people have actually done it you could have a gasoline-engined av they do exist right i think there are there are some out there certainly there have been prototypes But I hadn thought about that point about the utilization which is if you using it all the time those maintenance cost issues that you get with an EV many fewer moving parts all the rest of it all those kind of things just really make the electric vehicle overwhelmingly the compelling solution for an EV So and I think the other thing that just really, really interesting and, you know, thinking about things that Uber's been saying recently about their strategies is you have all this algorithmic information and programming and platform. So if I'm an Uber driver and I'm in an EV, tell us a little share with us a little bit what you can tell me. You can tell me right where and when I want to charge. You can tell me the rider. you could send me a driver that can definitely make it to the airport in traffic? Like, tell us a little bit about the sort of algorithmic part of the equation. Yeah, Amy, I'll go back to what I said about data and insights. One of the insights we've gotten from our drivers is that EV charging, surprise, surprise, for the last 15 years this has been the case, is one of the barriers for adoption. So we are taking a big stance to solve that problem for our earners. We're doing that in a couple of ways. We're doing that first through technology and we're building in our app a very native and seamless flow such that a driver can find where to charge and also be routed where to charge. And also through very sophisticated recommendations now that are live in five markets, be told when to charge. And that's really important because it's not just about sort of where and when, but it's also the time and the opportunity cost. So we're helping optimize some of that inefficiency in what exists today. If there is not charging available, and this is our recent announcement in the New York Times, we're also getting very sophisticated about our ability to influence demand and have confidence in underwriting net new infrastructure. We call them our backstop agreements. So we've actually signed LOIs with five charge point operators, both in Europe as well as in the US. And what we're doing is we're creating a financial commitment to help bring new charging infrastructure to very much underserved areas, which is predominantly urban centers where our drivers spend the majority of their time and actually where a fair number of them live. I think the last thing I'll say, Amy, is this is critical for our driver population, because if you think about our drivers compared to the general population, we are adopting EVs far quicker, five times faster. However, the challenge is significantly larger around charging because where the normal population will charge at home 80% of the time, it's almost an inverse equation for our drivers. It's not that high, but a fair number and the vast majority of our drivers are doing most of their charging in the public. So if we can help solve that and reduce charging by 30, 40 percent through our promotional agreements, help them find charging easier and then also provide new charging where there isn't. This is a meaningful step forward. You know, it's interesting you say that I have an EV and it came with, you know, free charging. And so I used to drive to the mall to charge because after all it was free. Why put it on my electricity bill at home? but almost everybody else that would sit and chat with me while I was charging my car were all uber drivers and and they had high sensitivity to if chargers were broken at the stop because I used to Ed you remember I used to complain bitterly all the time about how these chargers were always down yeah and has that got better is the availability of those chargers improved since as you say whenever it was a couple of years ago when it used to be a huge problem I can't even tell you because I got so disgusted that we put in a home charger, just like Andrew's saying. But maybe Uber has some insights. I mean, are your drivers finding better reliability at the charging stations that exist? Do you know anything about that? Well, I would say broadly, and I've been in the industry focused on this problem long enough to see the trend line actually move in the right direction. And, you know, the gold star metric is first time charge success where you plug in and it works. And that is significantly increasing year over year. But what we're doing to further promote available and, you know, sort of high uptime charging is that we are partnering with a select number of certified charging companies to make sure that we're only promoting the highest quality charging networks within our map and within our app. Right, because I have to say, so I'm going to do the lazy journalist thing now of reporting on conversations with my Uber drivers. But I feel like in this context, it's kind of, it's forgivable. But so I go to London quite a lot. When I go to London, I very often take an Uber. When I do that, it's very often an EV. And I think, what's your number? It's about 40% of Uber drivers on the platform have an EV, I think. So, you know, that makes sense. That's right. Yeah. 40% of miles in London are EV miles. OK, gotcha. Yeah, thanks. But yeah, so let's say 50% ish, a bit less. And so every time I'm in an EV, I always get talking to the driver. Do you like the car? And generally it gets a positive response, except that people complain about the charging. And the two things I hear about the charging, one is the wait time. And people say there's just not enough charging points, not enough fast charging. and the guy will say you know i get get to the fast charger and there's kind of one you know which is convenient to me and there's someone on there he's going to be there for an hour and there's two other people waiting so i'm going to wait kind of four hours before i can get fully charged up again that's a real problem that people see and then actually just recently the other thing that people have started uh mentioning is the price of fast charging and i was talking to someone the other day who was saying actually when i run the numbers for me um it's i'm not really saving a lot i don't think on if i had a gasoline vehicle although he would have said petrol this being in the uk he said i don't think i'm saving a lot uh on the petrol engine car because the fast charging price is very high said if i charged at home that would be a lot lower cost then the economics would be very compelling and it would really make sense but forget the fast charger it's more expensive and so i'm not sure really what i'm saving and and then he was saying what so in fact the next time I get a car, I may drop the EV and go back to a petrol engine car again, just because, as I'm not sure the economics really work out for me. So what do you do about that? I mean, do you recognize those? And as you say, you talk to a lot of drivers, you kind of hear about all these things. Do you recognize those as issues that people are raising to you? They're huge issues. So I'll sort of repeat myself in terms of, you know, what we're trying to solve. There's, you know, where and when, which is we're doing that through a technology solution in the app. There's, you know, when there isn't charging, we're investing in net new infrastructure. But the how much, Ed, is crucial here, right? You know, so 30 to 40% of your monthly total cost of ownership is actually from EV charging, especially if it's in the public domain. So if we can reduce that by 50%, you're reducing TCO by about 15 to 20%. So we're doing that in two ways. And I'll use London as an example. So ideally, you should charge at home. So the barrier to charging at home is either you don't have access to a home, that's significant, but also people don't like spending the upfront cost of the installation and the hardware. We recently launched a partnership with a company called Podpoint, where we're actually taking that upfront payment and spreading it into a pretty nominal monthly lease payment. And we're actually converting a significant amount of people who weren't using home charging into home charging users. And when you look at what they were doing before, which was public charging, we're saving them about $8,000 per month, which this is significant dollars as we think about it. On the public charging side, for those people not lucky enough to have home charging, we are signing promotional agreements with a select number of preferred charge point operators. And in exchange for really our promotional engine, which is influencing our demand towards their stations, we are getting anywhere between a 30% to 40% discount, which is meaningful for our drivers when they start thinking about their wallets and their ability to earn. So I want to turn to Samarth for a minute because I'm understanding what Andrew's saying about the human drivers. And as Ed and I are saying, we've had our experiences. I have my experience with an Uber driver who was taking me to a faraway airport and he had kind of a low battery left. and I was nervous, like if they're 18-wheeler overturns, will we make it to the airport? And you're telling me now the algorithm wouldn't have paired me with that guy if he was really low on and so all good. But Samarth, how does it work out with AVs? Like, how do you decide where you're going to charge these AVs? Are you going to do that based on electricity price of time a day, right? Because you're going to have this algorithm where you know it's a robot and you can tell it when to charge. Like, how's that going to work out? So, Amy, I think in the business of AVs, uptime is the most important, right? Like, these cars are incredibly expensive. And the number of humans it takes to run these cars, at the moment at least, is meaningful, right? So every mile or every minute this car is not on the road, it's a problem. And so to answer your question directly, the way we think about it is the cars come in, let's say using our numbers, the cars come in once they're closer to about 25% or 20% charge, and then we charge them up to 80%, right? And that's to make sure that the battery life is elongated, et cetera. If the charge is about 25%, Amy, we don't dispatch it. So if there's a risk that Amy doesn't get to the airport, we don't let it take a trip, right? And we also have incorporated technology that we have trips that as the car is coming back to the depot to be charged, we will only send trips to the AV, which are on along the way, right? So we're trying to make sure we're maximizing every hour, but making sure that charge is maximized. So that's one thing we do. The other thing we do is we think about when is the best time to charge, depending on not only prices, but also what we're seeing in the marketplace. We need the cars to be near demand by 4, 4.30pm, which means we will likely do things like charging and cleaning around 2pm. And all of this is a tightly sequenced marketplace algorithm, which we run, which you call the supply curve, which takes into account how many charges to have available, how many depot operators do we have, when does marketplace come online, what are the earnings per hour. So all of that is part of the models that Uber runs to figure out when to actually charge and when to dispatch of vehicles. And how are you deciding about where you need to have charging for those AVs? Because you're running most of them in big cities. So are you going to have your own depots in big cities? I mean, how is that going to work? Yeah, it's a two-part answer. So part one in how do we decide? So I do think one of the full best secret sauce in this industry is we know exactly where the demand is. We have a decade of data now on knowing exactly where people are going, how they're going, etc. So we can use all of that data to figure out what is the optimal location to introduce deadhead miles, i.e. empty miles that an EV or an EV might be doing. So we use that data to figure out where we want to put our depots, which is maintenance and some charging, as well as what we call our pit stops, which is our charging-only stations. Just think about a line of chargers and charging is the main objective. So we use our data to figure out the optimal locations to do stated miles. So that's part number one. part number two is, and to answer your question anyway, is Uber taking a more direct control? Short one says yes, right? So if I take a step back and if I think about, you know, the installments of vehicle value chain, you have the hardware providers, right? Like the actual base vehicle platform, you have the software kit, you have fleet management, and then you have distribution, the Uber app. That third layer of fleet management is where we're taking on a lot more control of our destiny, right? And doing that is actually taking control of the underlying infrastructure. So you will see us be more aggressive on this stance. We think that the underlying real estate and sites with the right zoning, right location, and sites which can get power will be hugely valuable. And that also details a little bit about our recent announcements, you know, where in the Bay Area, in Dallas, et cetera, we have indeed taken control of our own sites which are already for AVs. Right. So is that, from what you're saying then, is that kind of a new thing then? That is going to be what happens as AV adoption grows and you build out your network here is what you will actually be acquiring and owning your own sites, releasing your own sites, whatever it is, and having your own, as you say, you talked about two types of facilities. So you have a depot and you have a pit stop. And while the pit stop is for the quick charge up and the depot is for, as you say, kind of turnover, feel clean and wear. are those are those going to be um centrally located in cities typically or do you kind of put them on the outskirts and then drive you know and also i mean very basic question when do the vehicles go to these places then do they have an automatic trigger that says oh well i'm down to 25 charge i better go to the pit stop now or do you need a human intervention to kind of look at the you know the indicator and say oh hang on this one is running low it needs to go back in How does all that work? Yeah. So many questions, Ed. So I'll take them one by one. Yeah, yeah, I'm sorry. Sorry. Yeah, yeah, no, no. All good. All good questions. Yeah. So I think, you know, in your question about location. So think about the depot being in the more industrial area of town, right? Think about maybe 15, 20 miles away. Oh, sorry, 15, 30 minutes away. You know, mileage can depend based on the city. Think about larger warehouses where we have higher ceilings, where we have maintenance lifts and so on. And what's very important, there's a zoning, right? So we need zoning and we need high power, which usually means it's a bit further away. And that's where we do our time-based inspections. So the cars go back for any monthly inspections. That's where if you get into a fender bender, that's where the car is fixed, right? So think about that as a depot. Think about it as a large warehouse. Pit stop, think about maybe in an open-air parking lot, very close to the bot, right? So open-air parking lot, which we re-stripe, we put in, depending on the site size, anywhere from, let's say, 10 to 30 chargers. There's a lot of Wi-Fi, right? All these cars need a lot of Wi-Fi infrastructure to make sure data is uploaded and offloaded. So that's a huge part, and obviously security and so on. So that will be in the city center. And so the way to think about network design is we could have maybe one depot, which takes care of maintenance, and we could have a distributed set of pit stops, which is close to demand, to reduce dead end miles. So that's like the general archetype we go after. And then at the second part of your question is how does all of it work? Depending on the partner, it's different. So for example, if I take the Middle East to take a non-US lens, Uber decides and Uber works with our tech partners such as WeRide Baidu and so on to figure out when the car needs to be dispatched back to the depot The lion share of cases is charging related right Like you charge, you go to a pit stop, you go to a depot. However, there might be instances where, let's say, Ed was riding and he forgot his phone in the car. If Ed forgets his phone in the car, we will manually say, listen, car, there's a valuable device left behind. Come back to the depot so that we can secure, we keep it so Ed can get his phone back, right? It varies, but for the most part, coming back is largely charging related. And I'm afraid I have to confess that's not actually a hypothetical example. That is exactly something I have done in Anuba in my time. So, yeah, it's good to know there's a procedure for dealing with that. Got it. So something else that I wanted to ask then is, given, as you say, the way you're now looking at putting this infrastructure in, both for the AVs and for the human-driven vehicles, how much of a constraint is it on you to get a grid connection and to find available power? I mean, just thinking about this. Oh, man, I was thinking I was going to ask that question. So you took it right off the top of my mind. Well, I'm sorry, but yeah, I mean, it is the obvious question, right? Because, I mean, just doing the math, right? So, I mean, a fast charger can be what? Up to 100 kilowatts or so, can it? I mean, it's kind of 30, 50, 100. So if you're getting that much power through, as you say, you've got one of your pit stops, 10 to 30 vehicles may be charging at one time. So that's a megawatt to three megawatts on just that one location. You're trying to proliferate those all over a city. That feels like something that's going to be a very significant new challenge for the grid to manage. And so we talked endlessly on this show about data centers and the issues that they create. The question is, maybe you can't put your depot near a data center. And exactly, maybe, yeah. And I guess maybe, and we'll come on to that point in a moment. I mean, there's much smaller. I mean, the scale is much smaller. But then when you add them up, right, if you aggregate them, even if each individual one is much smaller than a data center, So if you get enough of them, and if these vehicles, EVs and AVEVs really catch on, then it's not similar. It does, you know, it becomes certainly material. It's hard to get distribution line hookup as it is to do, you know, larger transmission lines. So you still have the problem of a distribution line and transformers. Is there a big enough transformer in that area? Great point, yeah. So, yeah, we're all ears. So it's definitely a problem and it's something that's top of mind for us. So to add to your point, many of these sites are three, four, five megawatts. And some of our sites, we're asking for even more, like outputs of eight megawatts. And it does take time. So a lot of our time is actually spent working and developing close relationships with utilities all over the world. We're working hand in hand with these utilities to make sure that they are planning appropriately. And whenever we go through a site review, we're working hand in hand with them to make sure that can we actually get the power we need? Because if we don't get the power, the site is somewhat rendered useless, at least from an EV standpoint. From an EV standpoint, we can use lower power sources. So that's a huge part and huge focus of the team. The other thing we're doing, by the way, is also working with the OEMs. We're trying to figure out how do we work with the OEMs to figure out charging throughput. right can we charge at 200 kilowatt per hour and then will that mean faster throughput will that mean less charges per vehicles like the charge each vehicle ratio can go down so we're also doing a bunch of work directly with the manufacturers to think about their battery competition to think about how do we optimize vehicle performance there right so we're doing both those things and then Amy to your point you know there are there are a few sites where let's say when we when we started, we have about one to two megawatts, but then we've put in requests for additional power. And that means adding transformers. And so we've done it. We've had a few sites already where we had to add new transformers put in the site just to make sure that we're getting the energy that we require. And I will say that, you know, in credit to all the utilities, they've been incredible, right? They've been fantastic partners. I think on their end, they're also seeing the change, right? They're seeing the advent or the oncoming of EVs, and they're just trying to manage all the requests that we have, along with all the data sent requests, right? Because it's many times like the same tech company is asking for both. Right. So, okay, because you've answered the question that I was just about to ask, but I just want to get to ask it anyway, which is you talk about the utilities being great partners. I have to say, this is my kind of journey on this one, is when I first heard about EVs really growing, I thought, wow, this is going to be fantastic for the utilities. I thought all the power companies will love this. This is a great new customer base for them. They're going to replace the oil companies and the gas stations as the new suppliers of transport fuel. It's going to be wonderful. Then they're going to embrace this with both hands. Turns out that's not always the case, right? And actually, very often, additional demand is a headache and a pain for utilities. It's another challenge, something that's difficult for them to manage. And so I was going to ask whether you do, in fact, get a kind of a welcoming and constructive response from those companies, or whether you find them trying to put barriers in your way and to kind of obstruct you and say, no, this can't be done. It's too difficult for the grid. Sorry, Amy, what's your question on this? So I'm going to leap in here because if I'm a data center and I'm talking about 24-7, then I'm really a burden, right? But if I'm some little group that might want to charge at noon when I have a surplus of solar energy or at night when I have a surplus, if I'm in Texas, a surplus of wind energy, you might actually be interesting to me. especially if you'd start making your algorithm come only at times that are not peak demand times, then I might love you to death, right? So I don't know. Like, are you guys very time of day with these utilities or they're not sophisticated enough time of day to have that conversation with you? Yeah, I think two responses. I think one add to your first question. I think it has been more welcoming than confrontational is the short answer. I think there have been some raised eyebrows. And what I mean by that is we're asking for what, the equivalent of 4,000 or 5,000 households worth of power on one small square parking lot, right? So it does lead to like, what are you guys doing? Like, why are you asking for so much? But I think once we tell them, I think there's strong appreciation and they understand. And again, we're also giving them plenty of lead time, right? One of the reasons we're starting now at this small scale is we didn't want to go and make unreasonable asks. We're doing it very much in partnership with them, which I think is well respected and well appreciated. So they have the adequate time. They can ask all the questions. We can go through all the processes the right way. So I think that there's mutual appreciation on that. And then, Amy, to your question, I think if I think about the human-driven and Andrew should chime in here, then we can do things like slow charging overnight. Then we can do things like between shifts, the car can charge. On AVs, however, we are running the entire time, right? So it is quite steady. Like similar to data center, it is quite steady. So if I give you like, you know, a well utilized public charging site and like a really well utilized public charging site, you will see maybe 20 to 25% utilization. In our AV sites, you're seeing upwards of 65% every day, right? If anything, my current headache is we're getting congestion at the depots because there's not enough open chargers for the cars to charge when they come back. So that's a bigger problem right now. So maybe piggybacking off of what Samarth said, I think when we're self-performing, timelines are everything and coordination with utilities is key. When we're partnering with existing public charging companies, what's more important is actually the financial commitment and the confidence that they can go spend a lot of money to build these urban locations. to use Samarth's point. I mean, if we're building, you know, three, four, five megawatts, these charging companies historically have been reticent to put that type of capital in the ground without having some backstop or essentially, you know, mitigation to merchant risk. And what we're doing, and we're changing the dynamic in terms of how to actually gain confidence in building this type of asset class is we're coming in and providing that guarantee and that confidence. So that's one on the cost component. Amy, to answer your question on time of use, the answer is yes. We are working very closely with our charging company partners, namely around providing very sophisticated time of use pricing because exactly back to what we discussed before, our drivers are very price sensitive. So if we can influence when they charge throughout the day to help change the tariff or demand or cost profile of our charging partners by providing discounted or different rate structures, we will do so. And we're seeing meaningful change in behavior when we do it. But what you're saying, to make sure I'm understanding, what you're saying is on the AV side, you're like a mini data center in the sense that you want 24-7 for the AVs on these depots. For the moment, yes. I think as time goes on, as we have more capacity, I think then we can start getting into time of day and optimizing. I think right now, we need as much capacity as we can to make sure that we actually have enough charging for the cars. In a future state, I think we'll get into that and we'll get to the point where Andrew is, where we've optimized it for time of day and earnings. So I'm the one who always takes us to science fiction, right? And I've had other people tell me this will never happen, but Samarth, I have you right here in front of me, so I have to ask. Like, could you imagine a time when, is it that you make so much money optimizing this expensive vehicle on having a human being sitting in it that you'd never take a hundred of them or a thousand of them or 10,000 of them and plug it in somewhere to help with the grid flex alert problem? I mean, can you foresee a time when the batteries in all these AVs could be a flex alert? A VPP, basically. I mean, that's what we're talking about. A VPP, yeah. Yeah, like an asset? Like an electricity asset? Short answer is yes. I think, are we close to that? No, but eventually, do we see that happening? Absolutely. Right, in a world where we have thousands, you know, we definitely could make this work. And would it be something that you would, since we've all been talking about emergencies, right? Is it something that would just be like a community service that Uber could provide? We're bringing a bunch of AVs over to a hospital to make sure that they have electricity in a flex alert crisis, right, or after a storm or something like that? Or do you think there'd ever be a time when the amount of money you could make by having them sell back the electricity in their battery could be higher than the money you make having them move around? Interesting question. So a few thoughts. I think one thing that Andrew and I have spoken about is, is there a world where we get to the scale and optimization where having control of infrastructure basically exposes us to a new asset class where we can basically give back to the grid. And does the value of that land become a lot bigger because of that capability? So that is something we're thinking about. Again, this is a few years, if not decades down the line. To your point, in emergencies, I think the thing that I keep in mind is, in emergencies, we have to be very careful about autonomy capability. We want to make sure that we aren't coming in the way of first responders. We want to make sure that the AVs actually has the actual ability to maneuver past, you know, police over whatever it might be. Right. So I think, again, even the point, even the AV capability set, I think we'll have to go a few ways before your situation, you know, comes to life. And then I think the third thing in terms of like whether we can make more money, I don't know is the honest answer. I think we have enough questions to figure out, can we make money in a normal classifications? Yeah, just do it in the regular business. Exactly. So I hope so, but I don't know. Well, I'll chime in and maybe contribute one or two things, because I think this is both an AV, but also a hybrid conversation, right? So if we're thinking about what the fleet looks like over the next five to 10 years, it will always be hybrid. It's what Dara has said consistently. And I think what is underpinning our charging strategy on both sides, AV and human, is our insights and data and our ability to essentially orchestrate demand. So Amy, I think it actually goes both ways. One of our core goals as an organization is continue to provide multiple opportunities for our earners to earn, meaning that there could be opportunities for the human drivers as well in the future to do something else. And if that is to plug in their car to earn some money, I think we'll have to figure out whether or not, Samara's point, the math, maths, and it makes more sense to plug in versus go pick up a driver. I think that's still a long way out to figure out. But to play into your sci-fi scenario, we are working on the core competencies to have that type of discussion productively. And what about being a power supplier yourself? Is that something, I mean, would you put your own solar panels in these depots, have stationary storage, build a little gas-fired plant? There's obviously a lot of things you could do. And if you look again over at the data center world, all the things that those companies are thinking about very actively at the moment because of their challenges in getting a grid connection. They're being pushed, probably actually having a grid connection is the optimal solution for them, but because they're finding it hard to get connected on the timetables, they want to get connected, they're looking at bringing their own generation. Is that something, do you think, you might get pushed towards as well? So Swartcon says yes. I think you'll see a lot of similarities between what the data collect companies have gone through and what we will go through. I think particularly given the long lead times on getting these two, three megawatts that, you know, we are using on a temporary basis, you know, liquid natural gas just to make sure we, for the first six months, while we have the connection to come through, right? Our goal is to make sure it's grid tied, it's clean energy, but we will be taking steps in the interim just to close those gaps. But again, we work very closely with the utility and our tech partners to make sure that if we're doing short-term measures like that, we're appropriately buying the credits or anything else we need to do to make sure we're offsetting those emissions, to make sure we get in the right way. Right. But as you say, then, that's a very straight analogy with what's often happening with data centers, which, as you say, they have their local power initially just so they can get online quickly. And then they progress to a grid connection after that as soon as they can. I just wanted to go back some more, something that I mentioned earlier a while back, and you were talking about OEMs this Rivian announcement that hit the headlines a couple of weeks ago What the significance of that Yeah So again if I go back to my value chain where we have the vehicle manufacturer the software provider, fleet management, and distribution, which is Uber, in some ways, Tesla is going left to right. They started the vehicle platform, then they built on the self-driving, then they're doing some of their fleet management, and then they have an app. In many ways, we are going right to left, right? We started with a consumer app, and now we're trying to figure out how do we take control of our destiny on the other bridges, right? So we are doing a bunch of work with OEMs. You mentioned Rivian. We did a similar deal with Lucid, where we combined Lucid and Neuro, which is a self-driving stack. And working with companies like Lucid and Rivian enable us to get that supply where we can get thousands of cars who are totally enabled, who can work with multiple software providers and we can deploy them on Uber, right? So for us, it's, you know, supply, like Uber's always been a supply-led business, right? Like, you know, from the very get-go, supply is what makes Uber possible. Like you need a strong supply base to make sure that head and AME get good ETAs to make sure that it's a high quality car. And similarly, we're thinking about securing supply in the AV world as well. And the only way we secure supply is by doing deals like the one we did with Rubio. Right. And this brings me to, Andrew, something you were saying earlier about markets around the world where you operate. Obviously, Uber, a global business. And you said something about EVs make sense in the markets where they make sense. And my immediate assumption based on that, and correct me if I'm wrong, is that is in particular markets where drivers have access to low cost EVs. And in particular, I guess that means, therefore, Chinese EVs, not just BYD, but the other very high quality, very low cost Chinese EVs that are now available, increasingly being exported to the world, except to those countries where there are massive trade barriers and they aren't really allowed in, such as the United States. um am i right to think that is that a crucial factor then in terms of where ev adoption is really progressing the fastest that it's those countries that are allowing imports and in particular imports of chinese technology that are really uh seeing the greatest competitive advantage for evs and as you say just in terms of total cost of ownership are those the places where that TCO is lowest? You know, internally, we have a scorecard where we look at essentially TCO today and then TCO through Uber's interventions and where we can move the needle. Today in Europe is where essentially the scorecard is green. If you look at, you know, London, Paris, some of our markets in Portugal, I mean, we have upwards, as I said before, 40% EV miles on the road, some other big cities, 30%. But across Europe, 20% of our miles are EV miles. In the U.S., as a comparison, it's closer to about 10%. Obviously, California is much higher. But a lot of that, as you said, Ed, is vehicle supply or vehicle mix. We are looking at markets where we can get low-cost vehicles, both through intervention. So I mentioned some of our OEM partnerships. We are striking deals with companies like Kia to bring discounted vehicles to our drivers, especially in the U.S. markets. But I will tell you, and I'm actually headed down there in a couple of weeks, Latin America is a big growth opportunity for us. Brazil is one of our largest markets by ride volume at the whole company. When I looked at our list of top 20 cities by ride volumes, I was surprised to see a fair number of them were all in Brazil. So Latin America, where we can get Chinese vehicles, both in cities like Sao Paulo, but also Mexico City, will be a core focus for us as we think about interventions and ways to move the needle. So I always say that this sort of fallacy of trying to block Chinese cars from coming to the U.S. is not really what's going to affect the fate of Detroit, because it's not about the couple million, 10 million, whatever plus cars that get sold every year in the U.S. Now that one in four cars sold in California is an EV, right? And that's the biggest car market in the U.S. It's about what international markets. You know, Mexico was a primary destination for the sale of American made cars and even American used cars. So I just want to share for our listeners, but also for you, Samartha and Andrew, a quote I saw in the media yesterday from Paul Jacobson, who is the CFO of GM. And he was asked this question about whether they were going to change their strategy of canceling their plans for EVs because, you know, look what's happening. You know, I have some statistic from Octopus Energy saying that their EV charging unit sales are up 20 percent already. Right. We've got numbers coming out of, like you say, out of Europe and Asia, quite significant already. in fact it's a funny you know trivia point there's an electric rickshaw maker out of india called tesmo and they said that they already sold their march inventory of electrified rickshaws so anyway to the point here's his quote quote it takes four to six months of sustained high gasoline prices before people start to change their car buying preferences he says i don't think we're seeing that. And so he's saying that GM is not going to make any adjustments in light of the war. That seems super short-sighted to me. I thought the original canceling of the strategy was short-sighted because are they just saying they're not going to try to compete to sell cars globally? Like, what are they saying exactly? And I was keynoting a conference in Houston. So I made this point, you know, on stage, and then you could hear a pin drop And at the coffee break, everyone was coming up to me and say that was an astounding thing to say, because I talked about when I lived in Houston, I had a suburban because, you know, you have to drive 50 kids around. You're going to soccer. It's flooding. You need like a really tall car. And so I did. I mean, I also had a hybrid vehicle, but I had a truck for when I really need to have a truck. and um and that truck which was very affordable at that time i think i might have paid twenty seven thousand dollars for that truck that same truck is like eighty to a hundred and ten thousand dollars today depending on loading it with leather or you know features and you know you could buy a chinese ev for that same twenty seven thousand dollars that i bought uh my old truck with. So I said, if you're some person in Mexico and you have a choice between this snappy $27,000 Chinese EV with all kinds of bells and whistles, including like maybe a refrigerator or TV in the back for your children, right? Or a $100,000 American made passenger truck. what are you going to buy what dream are you having that people in mexico or brazil or other places are going to buy these giant cars that we use here in the united states for big money when they can get a really cheap chinese ev and you could hear a pin drop in there because that was you know clearly like a perspective you know i don't know you want to comment uh because i know you are working with some OEMs, but what do you think? Yeah. And just while you're thinking about it, and as a subsidiary thought on that, I wonder if you see that particularly sharply at Uber, because what you say is people who are driving for a business, they're not driving for ego gratification or to look flash or for pleasure. The economics really matter to them. And the way that stacks up in terms of costs and revenues is important to them more than anything else. But Igor, what do you think? Yeah, I think, Amy, you kind of hit the nail on the head. I mean, I do think that trade restrictions have prevented, obviously, some of these vehicles from being available in U.S. markets or other markets. I think, you know, some of these foreign markets have access to these lower cost vehicles. There's been a perception, I think, issue in the past around these. I think that's changing, especially with younger demographics. For our drivers, the vehicle is a tool, and they're looking for a high-quality, low-cost tool that they can use for a long period of time. So what we're seeing in some of these markets where these low-cost, high-quality tools are available is that they're great options for our earners, and they're meaningfully moving the needle on some of our other corporate goals like EV adoption and pushing that forward. So we'll continue to lean into that. so we ought to wrap up very soon but before we do then i want to go back to amy's point about the the science fiction uh aspect of things and imagining where this is going to be i mean if we talk to you have you back on the show in 10 years time hopefully see you again before then but uh if you were back here in 10 years time where would evs and avs have got to by then do you think I mean, I don't necessarily need you to put numbers on it in terms of percentage of rides, but just in terms of the position they will have in Uber's strategy. Andrew, what do you think? I think EVs will continue to be a focus point. And what you'll see is that a majority of our markets will flip from this being a minority or small segment to just the market. If you look at London as an example, a big focus for me is figuring out how we create incrementality around a relatively smaller product portfolio. But when you look at London where 40% plus of the miles are EV, it just becomes our core business. So that's the goal is how do we lean into EVs being part of our core business, building economically durable models and market opportunities for us. And really, as I said before, driving the dollar per mile down so that you can increase surplus value for Uber. I think actually that's a good segue probably to Samarth as well, because honestly, the same strategy across both platforms, EV as well as human. Yeah, Samarth, what do you think? Where are AVs going to be? I think Andrew said it very well. I think our North Star is how do we reduce the cost per mile to give back that surplus value to earners, to consumers? and in 10 years, Eddie, what you'll see or what we hope we will see and what our strategy is, is an expansion of the overall category size because we've reduced that cost per mile and you'll see a lot more human drivers, but you'll also see a lot more EVs on the road and ideally we're doing it in a way where we're causing less congestion, we're enabling further access to charging deserts and we're overall increasing mobility, you know, more than just the urban downtown geos. So that's sort of been Uber's story and vision since day one. And I don't think that'll change. Amy, where do you think it's going to be? You know, I have to say, I think urbanly, AVs are going to wind up becoming more and more popular. You know, some of the thing with new technology, you know, people say people got microwave ovens and nobody had a problem with them. So they were adopted very fast. Whereas nuclear power, when it started out, the second there was Three Mile Island, it slowed everything down. So, you know, what's your climate resilience plan for AVs, I think, is going to be very critical to the business success of them. But if you overcome that anxiety that, you know, people might have, I think it's going to be a compelling technology. and I think its versatility in terms of what you can do with it is going to be like huge. And so therefore, I just think all the people who are, you know, going out with baseball bats and all the things that people do today because they want human drivers to have an earnings, I just think the utility of it is going to become so huge that I imagine them, I feel Jetsons. I just think they're going to take off big time over time. What we're seeing actually is like in markets where we have AVs, earnings for drivers have actually increased. And in markets where we have AVs, the overall growth rate for those markets have also increased. So if I just think about the U.S., Austin and Atlanta, right, where we partnered with Waymo, like those two markets are growing almost twice the average growth rate of our other markets in the U.S., right? And we're seeing much better trends in driver earnings as well. So I hope those people at Baseball Bats are hearing because the drivers are doing well, too. Actually, that makes sense, Samarth, and I'll tell you why. If I can reliably rely on, I hit my app on my phone and I get a driver, whether that's a human driver or a robot driver, right? the more reliable it is and the more available it is the more people will say i don't need a car and the more people say i don't need the car the more opportunity there is for both human drivers and robotic drivers yeah very interesting no that is that is a fascinating thought and i have i mean just my final thought on this is that having been skeptical about both evs in general and avs in particular, as you know, Amy, and we have discussed this a lot. We caught every show. On every show. On every show we disagree about this. But I do think I have come round to your position, Amy. I think you have persuaded me that although it's clearly a bumpy road and there are delays and setbacks and we're very much in a setback in terms of the EV market in the United States right now, for instance, or certainly we were a couple of months ago. But I think depending on how the Iran war and the energy crisis related to that plays out, it's still going to be a tough market for EVs, I think, whatever happens. But as you say, globally, in terms of the direction of travel and the arguments that people are thinking about in terms of where they want to place their bets on technology and how they want to manage their exposure to energy price risk, EVs are just going to be increasingly compelling. And EVs along with that. Yeah, I do think that's right. So unfortunately, we do have to leave it there, but it has been fantastic talking to you. Thanks very much, Andrew. Thank you, guys. Thank you, Samarth. Thank you very much. Thank you very much, Amy. Look forward to talking again very soon. Many thanks to our producers, Molly Mowen and Toby Biggins-Gilchrist. And above all, as ever, many thanks to all of you for listening. We really value your feedback, so please do keep that coming. And we'll be back soon with all the latest news and views on the future of energy. Until then, goodbye.