Pod Insights Equity Podcast Serve is betting that food delivery and access to public markets are the keys to scaling robotics About this episode Today on Equity, Rebecca Bellan caught up with Ali Kashani, co-founder and CEO of Serve Robotics, to unpack how Serve is navigating public markets, scaling real-world robotics, and building what it hopes is the future of last-mile delivery.
Listen to the full episode to hear more about:
How Serve went from a lidar-focused startup to a publicly traded company via reverse merger in 2023
What it takes to scale a delivery fleet across cities like L. A., Miami, and Dallas
Why Kashani says Serve’s sidewalk bots collect four times more visual data per day than GPT-4’s vision model
How ground robots and drones might work together to finally crack last-mile logistics
Equity will be back Friday with our weekly news round-up, and special Google I/O coverage from Max. Don’t miss it!
Equity is TechCrunch’s flagship podcast, produced by Theresa Loconsolo, and posts every Wednesday and Friday.
Subscribe to us on Apple Podcasts, Overcast, Spotify and all the casts. You also can follow Equity on X and Threads, at @EquityPod. For the full episode transcript, for those who prefer reading over listening, check out our full archive of episodes here.
Credits: Equity is produced by Theresa Loconsolo with editing by Kell. We’d also like to thank TechCrunch’s audience development team. Thank you so much for listening, and we'll talk to you next time.
Learn more about your ad choices. Visit megaphone.fm/adchoices
Transcript Speaker names marked ? are inferred from speaking time and the show’s known hosts — they are not voice-verified.
0:00 Anthony Ha? This episode of Equity is brought to you by Schneider Electric. Schneider Electric, advancing energy tech.
0:17 Rebecca Bellan? Hello and welcome back to Equity, TechCrunch's flagship podcast about the business of startups. I'm Rebecca Beland and this is the episode where we bring on industry experts to help us explore a trend in the tech world and dive deep. Today, we're joined by Ali Kishani, the co-founder and CEO of Serve Robotics, which is a sidewalk robot delivery company that's backed by Uber and Nvidia, and I believe is the only one that's braving the public markets today. Ali, welcome to the show.
0:40 Speaker 02 Thanks for having me.
0:41 Rebecca Bellan? Yeah, listeners, we actually caught up last night at a really super cool, the most New York robotics salon, a salon of people who came to listen to Ali speak about robotics. What did you think of that? It was a lot of fun. It was in like a, you know, one bedroom apartment packed with about 50 people, bric-a-brac everywhere, books balancing goblets of wine on top of stacks of books. It was like the most New York thing I've ever done.
1:05 Speaker 02 It was a, it was a very cool setup and very New York. I think it's the most New York thing I've ever done.
1:09 Rebecca Bellan? Yeah, yeah, we were talking and outside the window, you could just hear the rush hour traffic honking the horns, trying to get onto the tunnel to New Jersey. So yeah, I hope that wasn't too much for you, coming from somewhat calmer, I think, San Francisco. But
1:25 Speaker 02 It was a lot of fun.
1:26 Rebecca Bellan? So Ali, give us a little bit of your background, all the way from LiDAR and DARPA to where you are now.
1:32 Speaker 02 Yeah, thank you. So I have been, you know, obsessed with robotics one day or another for a while. I did my graduate degrees using LiDAR and cameras, actually putting them together to make a better sensor suite. And basically, the way I got introduced to LiDAR was because I was working on the upper line challenge as an undergrad. So that led to my graduate school. I got my PhD in robotics, basically 10, 12 years ago. And then I started a series of companies. I always wanted to be a founder and be in the startup world. So immediately after graduation, I started my first company. I've been fortunate enough to have a few exits along the way. The last company I started sold Postmates, which if you recall was one of the large food delivery platforms.
2:13 Speaker 02 This was in 2017. We then got acquired by Uber in 2020. And that's when I decided I should spin off my team. So I talked to folks at Uber and we decided early 2021 to make serve an independent company. And three years later, we went public. So here we are.
2:29 Rebecca Bellan? Cool. And you guys went public in a reverse merger.
2:32 Speaker 02 That's correct.
2:32 Rebecca Bellan? Which is not quite a SPAC, but kind of like a SPAC in that it's like a quicker path to market using another company.
2:38 Speaker 02 Yeah, it's kind of faster and more efficient if that's what you're optimizing for. So you merge into public shell. In this case, what's distinct about this versus SPAC is that there's no money in there. There's no one kind of playing lottery with your stock and putting money in a vehicle that doesn't have a company assigned yet. So we basically did the merger in 2023 and then you go through a process of getting your filings and getting everything in place. And then you do a underwritten offering by a bank that allows you to list on NASDAQ, which is what we did just over a year ago.
3:10 Rebecca Bellan? How's it been so far this, your first year on the public markets?
3:14 Speaker 02 You know, maybe it's because I had really low expectations of life as a public CEO, but it's actually been fantastic. When people ask me how it's been, I usually say the only thing worse than being a public company CEO is being a private company CEO right now. It's just the ability to have access to capital is really nice when you're in a robotic space right now. And a venture capital, it's very formal driven. It comes and goes in waves and as a founder, you have to always be ready to deal with that fact. So being in a public markets just gives you stability where you're not worrying about fundraising every six months. You can actually focus on doing the actual work.
3:49 Rebecca Bellan? But it also gives you annoying journalists like me being like, so I saw this on your balance sheet. Or what's with like all of these filings? So you also got to deal with that, but you guys have been pretty good at talking me through it whenever I have questions.
4:01 Speaker 02 Thank you.
4:02 Rebecca Bellan? You're doing pretty good on the balance sheet. It looks like it looks like you have a lot of cash ready to go. And so the plan for this year is to expand from what was around 100 robots. What like late last year to 2,000 this year.
4:17 Speaker 02 That's correct.
4:18 Rebecca Bellan? And so then for our listeners who don't already know, talk us through serves robots. Yeah, what do they look like? What do they do? Where are they?
4:24 Speaker 02 Yeah, they're like if minions and Raleigh had a child. It's a little very cute, say, shopping cart sized robot that moves around sidewalks. If you actually go to LA right now or Miami or Dallas, you can see our robots. We had about 300 today, so we ended last year with 100. We are now over 300. By the end of this year, we're going to have 2,000. You can open Uber Eats and order from 1,500 different restaurants. And one of our robots would bring your food today. And we are expanding that to more markets. Atlanta is coming soon. The robots are mainly operating on the sidewalk. They cross the intersections. When they need to, they can go all the way to the front door of a restaurant. They get loaded, the bin is locked. And then when they arrive at the customer, they can go outside and press a button on the app
5:09 Speaker 02 and unlock the robot to grab their food. So the food is secure. It's very reliable. It's actually more reliable than human couriers already. It's very timely. You know, it takes about 18 minutes for them to pick up to drop off. Yeah, it's growing really fast. I think you probably saw in the earnings, our daily volume went up 75% in the first quarter. I'd be projecting another 60, 70% growth in the second quarter. So as we are deploying more robots, we are seeing the demand increase.
5:34 Rebecca Bellan? And then the Uber Eats partnership, is that exclusive?
5:36 Speaker 02 No, no, the robots can work with, and we can work with any partner, basically. It's just Uber is where we started. They've been really ahead of the curve with autonomy. They started the autonomy API in 2022 with us. So we've been helping them really bring this idea to market. And now they have many other partners and we also are able to partner with others.
5:55 Rebecca Bellan? So Atlanta is one of the next cities. That's about to get quite a lot of a boost of autonomy in Atlanta. Waymo is coming there soon as well, I believe.
6:02 Speaker 02 That's right.
6:02 Rebecca Bellan? Their robot taxis. We've talked about this a little bit last night, but you've said that Serve isn't a food delivery company. Like right now you are delivering food, but that's not the point, right? The point is to solve the problem of robots moving in complex human environment. So can you talk a little bit about why you're choosing food delivery as a way to get to market? And then what other plans you might have for sort of going forward?
6:29 Speaker 02 Yeah, you know, if you think about the vision we've always had for decades now about robots entering our lives, we're getting really close to that reality, especially with the progress with AI right now. You know, in the last couple of years, AI can really understand and comprehend our world and it can communicate with us. So we've really solved two major pieces. Now we're gonna get it to fit in the robots and make it commercially viable in that sense. But there's still this problem of bringing robots to complex environments in the middle of people, teaching it social behavior, basically making sure it's a good citizen and making sure it's safe. And there is already a lot of robots on environments like warehouses that's already scaled and doing quite well.
7:09 Speaker 02 But those are really controlled environments and you you really tell people how to behave next to the robot. And then there's, of course, a lot of focus on the streets, which is self-driving cars. A lot of money has been invested in there, of course. I feel like there's a really big opportunity in between those two spaces, which is places where you don't have control, where you have to tell the robot how to behave next to people, but you're not also moving at fifty miles an hour with a two-turn vehicle. And that's been our focus from the beginning. The reason we picked food delivery and maybe more broadly last mile is because it's one of the fastest-growing areas. It hasn't seen a ton of productivity gain in the last couple hundred years. The way, for example, the long-distance shipping has become so much more efficient and cost-effective. This has not been the case in last mile.
7:53 Speaker 02 Our food delivery within last mile is the most expensive because you need your stuff in thirty minutes. You can't batch them in a truck. It just is a very good place to get started, to solve a really big pain point. But the same technology that works on a sidewalk can work in any other environment where, again, people and robots need to coexist.
8:09 Rebecca Bellan? You said before that you don't need a large vehicle to carry a burrito, and so that solves that problem. But my question is, your sidewalk delivery robots operate in dense environments. Are people delivering burritos with cars? Like, I'm in New York, no one would do that. Like, it's all delivery systems on e-bikes that are charging through the bike lane.
8:26 Speaker 02 Yeah, it's a very New York question, actually, that I get because pretty much anywhere outside of New York, cars are the more popular mean for this kind of orders. But we were born inside Postmates, as I shared earlier, so we had the data. Ninety percent of deliveries that happen in the US are happening in cars. And that last 10% would be bikes and on foot and, you know, motorcycles, all the other basically methods.
8:49 Rebecca Bellan? But what percent of the last mile delivery, like literally those denser ones are happening with cars?
8:54 Speaker 02 You know, off the top of my head, I don't remember that, but I can tell you over half of deliveries are only two and a half miles or less. So half of deliveries are short distance, 90% are being done in cars. So even if you kind of assume that all the 10% is in the last mile portion, that's still another 40% plus of deliveries that are done in cars that shouldn't be.
9:14 Rebecca Bellan? So then you're operating in big cities, Miami, Los Angeles. We spoke a little bit about how there's the difference between, you know, special purpose robotics versus general purpose robotics. And I do want to talk a little bit about that because as a question that you got a lot last night as well. Everyone's kind of really hyped on humanoids. But before I ask that question, do you think that the way that you're approaching robotics, like in the streets, operating with humans? Are you thinking that that will help get robotics closer to general purpose?
9:46 Speaker 02 Yeah, I think so. Look, the navigation problem is always going to be there. The idea of walking among people, all the robots that we talked about, the humanoids, et cetera. At one point, when the form factor, the physical problem is solved, they're going to have to actually come in and coexist with us. So teaching robots how to do that is pretty important. It's a very important step. And not just in a research lab, but actually in cities. Sidewalks are one of the most kind of unexpected environments in terms of the amount of randomness and the long tail of edge cases that you're going to have. Like our robots have dealt with pretty much anything, including like goats, like a random goat on a sidewalk. Like, you know, you just never know what you're gonna see today. If you can do that, I think you have the foundation that you can now take that as a platform and apply to a lot of other spaces.
10:30 Speaker 02 And because of how fast this is growing and because of how expensive it's been, it's just there's a ton of demand for it. It unlocks a lot of opportunity.
10:37 Rebecca Bellan? So they're handling goats, they're handling edge cases. How much of this is fully L4 at this point? Where, how much of the remote operator is getting involved? Do they still need to help them cross the street? Like, where are we at?
10:49 Speaker 02 Yeah, our approach from the beginning was L4 is the holy grail. And L4 is basically a mix. So what it means is the robots should be able to do a lot of things by themselves. And in fact, they shouldn't really rely on people for safety. But because the world is unexpected and you're always going to have things that you haven't trained for, it's really helpful if someone can step in and help the robot when it needs help. So whenever we leave our operating design domains, ODD, then a human can help the robot. Because again, cities have a lot of side alleys and places that you may not have gone, but this one delivery is going to take you to. And then if we are stuck, if we don't know, if the robot doesn't know what to do, it can ask for help too.
11:30 Speaker 02 And then there's certain places like crossing a sidewalk or crossing an intersection right now, where we actually want someone to be in charge of the robot at that time. Mainly because that's where the risk is elevated. And it's not because of the robots. The robots are actually much more harmless. But the problem is the cars. Cars have 3,000 times more kinetic energy than one of our robots. So robots, even if they're coming into contact with someone, they're unlikely to hurt someone. But when you're at an intersection, the car can pose a risk to people. So we, right now, are playing it really safe and having someone be involved. And then over time, I expect we would make that more and more autonomous as we collect more data and become more confident.
12:08 Rebecca Bellan? Is there anywhere that you're doing that fully remotely? And not, like, maybe if you've been in L.A. longest, are you comfortable letting them cross the street without a remote operator there?
12:16 Speaker 02 We're not crossing streets without an operator yet, but it's definitely on the roadmap.
12:20 Rebecca Bellan? All right, now, taking it back to humanoids. Everyone is frothing at the mouth about humanoids because probably Elon Musk is just out here selling his vision and people forget that he makes timelines that he can't stick to. I think, I don't know, maybe they're just excited. And then, you know, we've got Figure, which is out here with a $40 billion, or attempting to get a $40 billion valuation, I should say. What thoughts?
12:45 Speaker 02 Yeah, a few. If you think about this sort of thing in a very short timeframe, I think we are definitely in the hype cycle right now, at, you know, the hypey part of the hype cycle. But if you think about it long term, I think it's definitely a real thing and it's coming. So, you know, if I was to be conservative, I would say in twenty years, these things are gonna be everywhere. Tomorrow or next year, they won't. The reality is somewhere in between and the, you know, people can have different opinions. How soon? Mine is a bit more conservative. I think we are probably at least ten years away from that form factor being commercially and economically deployed, and those are the keywords. And part of it is because for robots to be cost effective, and generally in every technology, we've seen the same cycle, which is you start with narrow use cases,
13:29 Speaker 02 where you build a custom-made vehicle for a purpose or a custom-made product. And then at some point, once you scale that and you bring the cost of components down, you know, you can get to the economics of it faster by focusing on a narrow application. So, once you scale that, you are then able to make it more general purpose. It usually doesn't go the other way around, where you start with a general purpose first. It always starts with the narrow applications. And as an example for this, Elon's estimate for the cost of a Tesla about is twenty-five grand, and that's at scale. And that's an Elon estimate, as you kind of mentioned. So, we will have to see how soon they're gonna reach that target. Our robots are already cheaper than that, and, you know, not by a little.
14:10 Speaker 02 So, that actually matters a lot when you're trying to solve a low-revenue kind of business, which is last mile delivery. So, you have to start with this kind of narrow applications. In five, ten years, when these are more scaled, and again, it helps bring the cost of everything down and develop the technology further, then everybody's focus would shift into, okay, how can I make my robots now do more things? And that's what humanoids ultimately solve, and it's kind of the next step in reducing cost and bringing more efficiency. So, I think we are, again, about ten years away from humanoids becoming real. But in these ten years, people are investing. I am very supportive of those investments. Thank you to all those folks that are betting their money into this kind of research, because it's gonna bring that to reality sooner than if it doesn't.
14:51 Rebecca Bellan? So you say narrow applications can lead to larger scale, which can lead to the general purpose. But like, I feel like I'm missing a few steps in between that. Is it because it's just financially not viable to try to get to gen purpose without the scale? Or is it because you need the scale to get more data to be able to reach that? What I've heard from people, the problem with general purpose robotics is that we haven't gotten that internet level scale of data that allowed the chat GPT aha moment.
15:18 Speaker 02 Yeah, it's actually both. I'll speak to both of them. So on the cost sides, one thing that I really focused on from the beginning with SERV was getting to build a profitable autonomous mobility company as soon as possible. And in a space where it's a very high capital cost type of development, that's really important in building a lasting business is finding a way to get there. The self-driving car companies, for example, have spent about a hundred times the capital than we have. But we are now among the largest fleets out there operating. Now, if you obsess over the actual financial numbers and unit economics, you wanna really cost optimize your hardware. So for example, why does a delivery robot need a head? For humanoids, it's an important part of the form factor, but for a delivery robot, it's completely irrelevant.
16:02 Speaker 02 So when you're building a generalized form factor, you are adding things that are not necessary and removing those parts and making hardware that's just really good. And by the way, a lot of these humanoids have like two hours of battery. I need to have a robot that goes out all day on a single charge and can just keep delivering. It should not be sitting there charging or coming back home to swap batteries. Like now that makes sense, it really hits your economics. So you make a custom vehicle for that application because it's a very high volume application. You can get the cost down much further than you can do the general purpose. The second piece of it is the data. So just to give you an interesting estimate, when we deploy our 2000 robots by the end of this year,
16:43 Speaker 02 our robots would collect more images every single day than GPT four's vision model was trained on. In fact, not just more, four times more images every single day. And those images would also be paired with LiDAR data, which is really valuable if you're trying to train a model to understand 3D geometry. It would be paired with Actions of the robot. So what did the robot do that led to this? For example, it steered to the right and it got this new image. So you can actually use it for training models that way. So that is going to be super valuable as well. So there's a reason why I think historically there's always been this sequence of narrow scale, then generalized, and we're kind of seeing the same thing here.
17:20 Rebecca Bellan? Well, if we're talking about the data, just to play devil's advocate, Tesla does have a ton of driving data, right? I mean, they have millions of cars with autopilot, with FSD, that have been learning how to interact in environments. Like, does that give them an edge when it comes to reaching general purpose for the robots?
17:37 Speaker 02 Yeah, absolutely. I think it does give them an edge. Basically, there are few companies right now in a position to have fleets out there where they can access that data. Tesla is definitely one of them. Google has their fleets. Obviously, we are scaling, so it definitely gives you the advantage. It still doesn't solve the hardware cost problem, but it does solve the data problem. And then for the broader markets, to be able to reach the same point as Tesla's of the world, you need more companies that are operating fleets, hopefully profitably, at scale, so that that data becomes more commoditized.
18:07 Rebecca Bellan? Yeah. Are there any efforts in the robotics space to pool and share all of the data that everyone has so that everyone can try to reach the general purpose robotics problem together?
18:18 Speaker 02 You know, I don't know of any effort I would consider right now, but we've always looked at it as the data itself is an opportunity to advance the space. So we are talking to partners about different ways of using the data. Some other folks are much more protective of their data. I assume they wouldn't want to share that. But I think, again, with more players out there, this becomes more likely that the data becomes accessible. And it's really just a matter of time at this point, because everybody has initially focused on GPUs and solving that problem. Now, a lot of the focus is in securing access to data. So it's less collaborative, more competitive right now. But there's a lot of work being done right now that I think is going to unlock new data sources, especially with robots.
19:05 Anthony Ha? What moves our world forward? Energy. History has been shaped by it, revolutions won with it, and progress dependent on it. For 190 years, Schneider Electric has been a leading force in this story, creating the technologies that harness energy and transform the competitive landscape. And today, in the dawning era of energy and industrial intelligence, Schneider is helping customers build more efficient, resilient, and sustainable operations across every sector. From data centers to homes, from buildings to industry. Schneider Electric, your energy technology partner. Discover more at se.com.
19:43 Rebecca Bellan? So you mentioned that your robots, you've grown this fleet, it's out here collecting data about the real world environment, about how it interacts with humans all the time. That also, I think, leaves it open for risk, right? And by that, I mean cybersecurity risk, I mean risk to people's privacy. I think the media made a meal of it, maybe us too. I actually don't know if we covered that. You know, one of your robots was being stolen. And you worked with police to provide camera data of the people who were stealing your robot. But I mean, this is something like Waymo has been asked by the police to provide video data of potential crimes and things. So, you know, you can always get that warrant. And as we're moving into a world where there's a lot that you could potentially be doing that's illegal today.
20:26 Rebecca Bellan? I mean, Trump said that vandals of Teslas are domestic terrorists. And really, what does a vandal mean? Couldn't it mean a protester? Like, these things can get a little fluid. So it's like, how do you, are you anonymizing that data? Like, you don't want to get to a point where the police are asking for data and then you have to give it, or maybe you do, maybe you want to help. I don't know. How are you thinking about it?
20:47 Speaker 02 Yeah, you know, it's definitely a real issue that we need to address as an industry and, you know, work with regulators, work with cities. One thing that I don't want to do is be sitting there and dictating these things for the rest of society. I think that's really not something that I should be doing, or anyone in my position should be. So what I think is the right approach to this is work with the cities. Maybe people in different cities have different... Expectations. Maybe they actually in some city, maybe they want more protection in some cities, they want more privacy. Right now, our policy is pretty straightforward. We don't share unless it's, you know, a safety risk or damage to our property and our staff. Or it's a warrant. So that's, you know, following the law. So otherwise, we don't share. It's pretty straightforward.
21:31 Speaker 02 In the future, if you have robots everywhere, I think that the cities should tell us or that people, basically, through a democratic process should decide. Here's an asset. This data exists. It can be turned away. It can be used for applications to make people feel safer or otherwise. I really, I mean, what I'm saying here can be taken out of context and what I really mean is it should not be our decision. I just don't think that that we are in a position to make that call. For now, we are going to default towards privacy, which is don't share unless it's required by the law. Or it's again a safety or a property damage to our property, which I think everybody would say is, you know, a fair thing to do for us to use to protect that.
22:11 Rebecca Bellan? Would you consider anonymizing facial data or just kind of like scrambling it? I mean, I can imagine a world, for example, I spoke to Reddit CEO recently and they were talking about why it's really important for Reddit to not know who their users are and not be able to identify them because it's a place where you come to share things. You don't know when something could become illegal and it's like, that puts you in a sticky situation, right? Would you rather just not be able to access that kind of data?
22:36 Speaker 02 Yeah, I think that's a very, that's one elegant solution if it comes to you. You can't protect the data in any other way and you avoid the but it being accessed because of, I don't know, overused by law enforcement, for example, then you can kind of take that approach to it. Technically, it hasn't been been any issue yet, but we are at a smaller scale. So I could expect that as we scale, the problem set kind of changes and evolves. So if we feel like we are not in a position to protect the data, then we have to take additional measures. And yeah, I would definitely be be open to that sort of thing.
23:06 Rebecca Bellan? You're building autonomous tech. Obviously, this is becoming a platform for laying up other technology or working with automakers. I believe you said you're working with a top tier automaker in Europe. You work with Magna to produce your robots, but also to help produce other vehicles. Have you been approached by the military? You know, you worked with DARPA, you know, anything autonomous, the military comes knocking, right?
23:25 Speaker 02 Yeah, not directly, but we've had people approach us who want to go address those problems. So it's definitely come up more than once.
23:32 Rebecca Bellan? Are you open to it? I mean, this seems like defense tech is huge right now. It seems to be a big pot of money for a lot of companies and they're really embracing the dual use.
23:40 Speaker 02 Yeah, you know, it's been definitely a discussion topic and we are developing our principles around it right now. But there's things that are obvious. We don't like to be involved in weaponry or things like that. We're also, honestly, it's not that nobody should do it. It's just we're too small to really understand how to do that sort of thing and do it right and in a way that we feel like it's morally right. And then there are certain applications that I think were obvious, like if someone asks us for help in Ukraine, and I'm not sure how, but as an example, but I'm not going to say no to that. That to me seems like a really obvious situation where the line is very clear.
24:14 Rebecca Bellan? Taking it back to the autonomous delivery space, you have some competitors. You've got AV Ride, they're also opening up in cities. There's Starship, there's Veu, there's KiwiBot. I'm noticing a lot of the sidewalk delivery robot companies are operating mainly at college campuses, but you don't seem to be doing that. Is there, is there a reason?
24:32 Speaker 02 Yeah, we kind of deliberately decided to focus on the cities. It's a bigger opportunity.
24:38 Rebecca Bellan? Less vandalism, more vandalism?
24:41 Speaker 02 You know, it's, it could go either way. I don't know. It's hard to guess. College students, I mean, on one hand, it's a more controlled environment, you could say. On the other hand, it's college students. So it's, it's hard to tell. I don't know actually if that's the reason. But a couple of reasons was, one, the market opportunity, colleges are smaller subsets. You know, solving the city sidewalk is a harder problem. So if you solve that platform thing that I mentioned earlier is the playbook for us, it becomes more relevant. And the other side is because again, we're inside Postmates, we have access to the data, the revenue you can generate on college campuses was less per robot because the basket size, as we call that, is smaller. So when people order, they're ordering fewer things. A lot of times it's say coffee. That means the revenue opportunity per transaction is less. Also, at night, a lot of people go home. So college campuses are less busy.
25:26 Speaker 02 So a lot of pickiness around lunch, not as much over dinner. Summers are quieter. So net, net, you have this seasonal picky demand volume, which makes it harder. To build a robust business with solid kind of revenue generated for the robots. So by going to the city and having a more steady type of revenue opportunity there, we also could make our robots more advanced. We can have more GPUs and better sensors, et cetera. So that's been our theory of the case. And there are a lot of folks that are on to campuses and they're seeing success there. So, and I have a lot of respect for them. I know a lot of the CEOs in those companies. Our approach was different.
26:00 Rebecca Bellan? And then you've also done some different things than other companies. Are you still piloting the the serve wing situation with the? So this was a sidewalk delivery robot passing the baton to a drone, a wing drone. Is that still going on?
26:15 Speaker 02 It is.
26:15 Rebecca Bellan? What was the deal there and what was the thinking aside from this is cool to see?
26:18 Speaker 02 Yeah, no, being is is an alphabet company. And yes, we are currently working with them on this and would have more to share soon. But the idea is very simple. So drones are fantastic for delivering food over longer distances. Like if you're going to go five miles in a minute, they can actually do that, which is really awesome. But the one challenge for drones and actually self-driving cars is getting to the front of the restaurant where they need to be loaded. Restaurants, as you would expect, are usually in busy business districts where real estate is harder to get. They don't have parking spots or even often they don't own their own real estate. And there's a lot of people around, so you don't want drones buzzing by or. So what's nice about our robots is we can go to any restaurant right now. We don't need any infrastructure. We don't need anything to change.
27:03 Speaker 02 We're doing that today. So when the delivery is only a couple of miles, we can just do it ourselves. But many deliveries aren't. Half of the deliveries that are more than a couple of miles, we can actually pick up the order and then hand it over to a drone that could be basically in a handoff location a couple of blocks away in, let's say, a dedicated parking space for this purpose. And we can do that completely automatically, no human involved. So the robot basically would use this little arm thing that hands over the package to a piece of physical infrastructure in place in that location. And then the drone grabs it from it. And you can look up the video of this that we posted. But it's really, really cool. And I think putting drones and robots together, it really makes this a comprehensive solution for last mile, which is very exciting.
27:44 Rebecca Bellan? But more importantly, you hinted at something last night and I really want to hear if you can give us a quick story of how that handshake happened because you said something about being architecting a situation in which you were maybe sitting next to the CEO of Wing.
28:01 Speaker 02 So yeah, I had this idea because again, you know, one thing when you're born inside the industry, you have this unique perspective. When you're working with all these restaurants, you know what works for them, what doesn't. So I had this apathy about, hey, we should really get the robot drone thing to work together. And I really love what Wing, folks at Alphabets, kind of Wing subsidiary, have been working on. So I found myself at a dinner with Adam, the CEO of Wing, you know, at CES. And the next day, I noticed that he was at the airports catching the same flight back to San Francisco. So I may have engineered my way to kind of sitting together on the plane. It was Southwest where you just can't sit anywhere. So I found a way to kind of steer conversation and make sure we sit together. And then I had a captive audience
28:44 Speaker 02 for a couple of hours. Then I pitched him the idea.
28:47 Rebecca Bellan? Nice. So smart.
28:48 Speaker 02 You got a hustle.
28:49 Rebecca Bellan? Yes, this is exactly what you got to do as an entrepreneur. You got to push yourself into those situations. It's inspirational.
28:54 Speaker 02 I don't think I've admitted this to him, by the way. So he might learn this for the first time.
28:58 Rebecca Bellan? Okay, well, I think he'll be flattered, I hope. I think it's cool. It's something I would do. That means half the reason I went to this, you know, party with you last night. I'm trying to meet people that I could talk to and bring on the show. And yeah, it's a good way to do it. So this has been really fun. Where can our listeners connect with you online?
29:13 Speaker 02 I'm on Twitter, I'm on LinkedIn. I think our robot has some social media presence as well. If you watch John Mulaney on Wednesday nights, our robot is usually there. It's a sidebar for John. So yeah, there's a lot of ways to catch a robot or myself.
29:27 Rebecca Bellan? Amazing. Okay, cool. Well, thanks so much for joining us again. And to our listeners, you can find me on Blue Sky and X. You can find Equity on the same places, at EquityPod. Talk to you next time.
29:38 Rebecca Bellan? Equity is produced by Teresa Locan Solo with editing by Kel. We'd also like to thank TechCrunch's audience development team. Thank you so much for listening and we'll talk to you next time.
Previous $1 Billion a lot of money these days? Next OpenAI goes all in with Jony Ive as Google plays AI catchup Page updated October 12, 2026 · transcript and summary are machine-generated
