Transcript: Wayve: Scaling Autonomous Vehicles Without Borders 

Executive Summary

In this episode of The Road to Autonomy, Kaity Fischer, Vice President, Commercial and Operations, Wayve discusses the company’s new partnership with Nissan, set to launch in 2027. Throughout the episode, Kaity explains how Wayve’s scalable, end-to-end AI technology enables rapid generalization to new countries and vehicle platforms without relying on HD maps. 

Their “mapless” approach is key to their licensing model, which provides vehicle intelligence directly to global OEMs for personally owned autonomous vehicles.

Key Topics & Timestamps

[00:00] Wayve’s Partnership with Nissan for a 2027 Launch 

Wayve has announced a partnership with Nissan to launch its technology in their vehicles in 2027. This timeline aligns with the release of Nissan’s next-generation ProPILOT technology.

[01:00] The Scalability of Wayve’s End-to-End AI 

Wayve’s technology is designed for global scale, allowing deployment across many countries without requiring HD maps or operating within a geofence. The same vehicle intelligence powers their cars in the UK, US, Canada, Japan, and Germany. This scalability is so efficient that when moving from the UK to the US, the system achieved the same performance level after collecting just 500 hours of data over approximately eight weeks.

[03:00] How the AI Generalizes for Different Countries and Driving Styles 

Because the system uses end-to-end AI, it doesn’t need to be hand-coded for new rules, such as driving on the opposite side of the road. Instead, it learns and adapts by ingesting data, similar to how a human driver adjusts to a new country. This allows the AI to adapt not only to different rules of the road but also to the unique driving cultures in various cities. The generalization process accelerates with more data; it was three times faster to adapt to Germany (the third country) than it was for the US (the second country).

[05:00] Handling “Long Tail” Edge Cases with Learned Behavior 

The end-to-end AI approach excels at handling complex and uncommon “long tail” scenarios that are nearly impossible to account for with hand-coded rules. For instance, a Wayve vehicle seamlessly and safely maneuvered around pedestrians sitting on a curb and in the road outside a pub. The vehicle learns how to interact with the world safely by ingesting data and observing human-like behavior, rather than relying on a pre-programmed list of scenarios.

[07:00] Wayve’s Business Model: Licensing Scalable Software to Global OEMs 

Wayve’s business model is to license its vehicle intelligence software directly to OEMs. This strategy targets automakers who build vehicles for global scale and need a technology partner that can also scale globally. The technology will offer true point-to-point driving, and the software can be upgraded over-the-air, allowing vehicles to advance from L2+ to L3 and eventually L4, depending on the hardware.

[12:00] The Future of L4 Autonomy, Fleet Ownership, and a Partnership with Uber 

Wayve sees pragmatic benefits in L4 autonomy starting with fleet ownership due to economic and maintenance practicalities, though personally owned L4 vehicles are a future possibility. Wayve is actively working with OEM partners and Uber to deploy a safe L4 system in production vehicles fit for a ride-hailing use case. Uber has been a long-standing partner and investor in Wayve.

[15:00] The Strategic Decision to Focus Purely on AI Software 

From its inception, Wayve has maintained a “partner-first” approach, choosing to focus exclusively on its core competency: vehicle intelligence. The company does not plan to build its own hardware or vehicle manufacturing capabilities. This allows Wayve to remain agile and partner with best-in-class companies across the value chain, such as Microsoft for cloud computing, Nvidia for chips, and Uber for market access.

[19:00] Why the Industry is Shifting Toward a “Mapless Future” 

The industry is increasingly recognizing that a mapless approach is required for true scalability. Relying on pre-built HD maps restricts vehicles to a geofence, which is costly and difficult to maintain. A mapless system allows vehicles to traverse new areas “sight unseen,” giving consumers the freedom to drive anywhere and providing fleet operators with a more profitable business model not constrained by geography.

[23:00] Adapting the Wayve Driver to Various Vehicle Platforms and Sensor Suites 

Wayve’s software is decoupled from hardware, allowing it to generalize across different vehicle platforms like sedans, SUVs, or trucks. It takes only about 100 hours of sample data to adapt the AI to a new vehicle model with the same level of performance. The company is also “sensor agnostic,” meaning it can work with an OEM’s existing or preferred sensor layouts, including various combinations of cameras, radar, and lidar.

[34:00] The Future of Wayve: Expanding from Vehicle Intelligence to a Foundational AI Model 

While the first application is on-road vehicle autonomy, Wayve is ultimately building a foundational model with generalized intelligence. In the future, this core intelligence could be adapted for any application where AI meets robotics. Potential use cases include off-road vehicles, warehouse logistics robots, and even humanoid robotics, with the North Star vision of becoming an intelligent, trusted AI partner in many aspects of life.

Watch the Full Episode of The Road to Autonomy

Full Episode Transcript

Grayson Brulte: Kaity, it’s great to have you here. We had Alex Kendall on last year, and he was telling a big story about surfing these big waves, and lo and behold, you surfed a huge wave right into a Nissan deal. Congratulations on the deal. How did it come about? How are you gonna scale it? 

Kaity Fischer: Thank you. Yeah, Grayson. Look, really, really glad to be here. So thrilled to be talking again. Um, and yeah, we recently announced our partnership with Nissan. It has been months in the making, but they’ve been really exceptional. We’re thrilled to be working alongside one of the world’s leading OEMs and to be launching our technology in their vehicles in 2027, which we’ll be here before we know it.

Grayson Brulte: Why 2027? Was it a a development timeline or is that just based on the design of the architecture of what Nissan’s developing? 

Kaity Fischer: Okay, 2027 aligns with the launch of their next Gen Pro pilot technology.

Grayson Brulte: Nissan’s a Japanese company. I’m extremely bullish on the Japanese market. I think it’s one of the most important autonomy markets in the world. Will these vehicles be in Japan? Will they be global, or where do you see them being deployed? 

Kaity Fischer: Yeah, we can’t talk too much about details, but we certainly look at OEM deals generally as one of the benefits of going direct to licensing our technology with an OEM and one of the benefits of our scalable approach to technology. Is that we’ve got the ability for global scale. So by deploying on consumer vehicles, uh, and without requiring HD maps, our vehicles will be able to traverse. Um, you know, many countries, many geographies, all without a geo funds.

Grayson Brulte: You’ve released footage from Germany, Canada, United States, United Kingdom. How is this technology proven to be so scalable? And I say that because you did release a public data set and how when you go to a new country, or my daughter would say, a new land, you’re able to get the driver to drive faster.

Kaity Fischer: I love that you’ve been keeping up with, uh, with our footprint. It’s been a really exciting, I would say year, but gosh, it’s only been six months since we’ve been in all of those new lands, as your daughter would say. Um, so our technology is scalable because we use end-to-end ai. So the same vehicle intelligence. That powers our vehicles in the uk, powers, the ones in the us, in Canada, um, in Japan, in Germany. And, uh, watch this space. ’cause our vehicles in multiple continents are actually currently on a road show. You’ll see them popping up, uh, all over new cities, new states, new countries. Um, but by using end-to-end ai, it means that we don’t operate in a geofence. It means that we can go new places, site unseen. We call that zero shot testing. We can do so incredibly efficiently. So we recently published, sounds like you might have seen that when we took our vehicles from the UK to the us we were able to generalize the technology, so show the same level of performance. Uh, we were able to show the same level of performance in the US with 500 hours worth of data. And that is incredibly efficient compared to more traditional approaches to autonomy. Uh, we did that in about eight weeks. And the eight weeks transparently was a matter of. Fleet size and speed, you could scale that up if you had more vehicles, um, you know, working double time ’cause it’s about 500, um, sampled hours. So the right data that you select from that subset. And, uh, when we went, when we went and took the same fleet to Germany, we actually saw a three x uptick in the performance. Four zero shot. So we were three times faster to generalize our technology in country Number three, compared to country number two. And we’re excited to continue to show how that generalization accelerates as we go into new markets.

Grayson Brulte: Japan you’re testing now are, are you expecting that to accelerate now that, that you’re testing in the Japanese market? 

Kaity Fischer: Yeah, we’ve seen similar results in Japan. Excited to be able to publish those soon. And, um, I mentioned that we’re on a road show. Same goes in Japan. So being able to take our fleet that’s based in Yokohama and Tokyo, um, and take them all across the country.

Grayson Brulte: So the cur, the curiosity factor steps in, so in England you drive on the left hand side of the road, Japan, they drive on the left hand side of the road. America, we drive on the right hand side of the road. How do you train the system to understand the, the nuances in road design and just let’s call it general driving habits of these different markets that you’re expanding to? 

Kaity Fischer: It’s a great question, and notably in the example I gave taking the fleet from the US to the uk that’s of course opposite side of the road. The key thing here is that because we use end-to-end ai, we’re not hand coding or, um, you know, literally training the system on the opposite side of the road, which is what inherently makes it more scalable. So the more data that we ingest, the more intelligent the underlying vehicle intelligence gets. And think of it just like you or I, right? Uh, we’re both American. I know. So we both learned to drive in the United States. You know, you come to the UK and it might take you some adjustment period of, okay, that’s right. I’ve gotta be on the opposite side of the road. Some of these street signs look different. You’re certainly still able to drive. Um, it just takes a bit of adjustment. Same thing with our vehicle intelligence. The same underlying principles apply to the driving task, and then we use that efficient subset of data, um, to sample additional hours that are specific to the market. So that goes from everything from different rules of the road to also different driving cultures that we’re able to adapt to because. We all know that each city is, uh, has unique aspects and features and different driving techniques depending on what drivers you’re encountering.

Grayson Brulte: And I wanna ask you about an edge case. Alex went for a ride with, with Sir Richard Branson and posted a video about to open a Virgin hotel, which I think was a brilliant marketing idea ’cause now you’re part of Virgin Lore. Of Virgin history. But they drove by a pub love pubs and there was individuals sitting on the curb, another individual sitting in the road. And the Wayve vehicle didn’t even hesitate very smoothly just to went around that individual. And Sir Richard made a comment to Alex. Wow. That was impressive. The vehicle didn’t hesitate. How is it able to do that? 

Kaity Fischer: Yeah, and this is where I think our technology, using end-to-end AI really shines. So we think about, again, we don’t have to hand code rules. The vehicle learns through ingestion of data, um, and learning the ways of the world. Again, very human-like behavior. That’s what you or I would’ve done, right? We can seamlessly maneuver through crowded streets, through vulnerable road users, um, or through pedestrians sitting outside of a pub. And so what’s amazing about our technology is that it’s really ideal for solving that complex long tail of scenarios. That you simply, uh, it’s nearly impossible to account for if you had to hand code every scenario. I don’t know how far down on the list, you know, pedestrians outside of a pub would be, but it would be pretty far down there. Yet our vehicle is able to do it because it’s got the intelligence of how to, uh, interact with the world safely.

Grayson Brulte: Your vehicle’s smart. There’s no other way to say it. It’s a better driver than I am. I get distracted and I think about things and, oh, lo and behold. You, you have to take decisive action there. The end-to-end AI approach to me says scalability, that you’re able to scale this, and also to me it says that you’re able to usher in the, the error of personally owned autonomous vehicles. When you’re speaking with OEMs, is that one of the incentives, the things that they’re looking to do? Potentially, 

Kaity Fischer: So with the personally owned piece or the scalability or both? 

Grayson Brulte: Both.

Kaity Fischer: so scalability, absolutely. Our business model, just to be clear, is we’re licensing our vehicle intelligence, our software directly to OEMs, and we all know that OEMs don’t sell a vehicle in one city or one state and or even one country, right when they build. These vehicle platforms, they’re building them for global scale, and that’s how they’re profitable. They have to amortize the, you know, incredible costs of making a safe, performant vehicle. They have to amortize that over global production volumes. So they’re looking for a technology partner that can give their end consumers a really performant solution for autonomy that also scales globally. That means that it can be deployed across many countries to many consumers, and it also means that when the consumers use that vehicle, they’re not limited to where they’re able to use the features. If they’re not limited to mapped stretches of highways or only highway or only urban, when end consumers will use Wayves technology, it’ll be true point to point driving. So you could leave your driveway, um, drive an hour or two to work or run errands. Uh, the vehicle will park at your final destination. And, um, to start, it’ll be in an L two system where you monitor that whole trip and you supervise. But the vehicle intelligence will, you know, do all of the maneuvering from point A to point B. And then our vehicle technology is also, um, able to be upgraded over the air. So depending on the vehicle hardware, we can take that L two plus technology that drives point to point, upgrade it to L three and then all the way to L four as well, um, if the platform is capable. So scalability, um, absolutely something that, uh, is top priorities for our customers.

Grayson Brulte: You need point to point, there’s an OEM that gave me a test vehicle and only worked on the highway ’cause I had to be quote unquote mapped and said, okay, so I’m playing with a thing, trying to turn it on. I’m on the on ramp. I can’t get on on the, on rail. This is useless. I have to get O on the highway over two lanes before I can activate it. And I was going to the market and then I had to take over to get off. It was just, I was like, okay, this is, this is pointless. This is useless. You need to be able to go point to point. To me, that’s, that’s one of the value adds there. So obviously reading your public statements, it’s, it’s, it’s very clear that you’re building a licensing business that you said you’re gonna have multiple OEM partners From an over the air update aspect, how will the software update? Are you gonna have to write special kernels for each OEM or, or how will that work? 

Kaity Fischer: So when we’re working with. OEMs, um, we’re looking at, of course, providing a base technology that’s deployable, quote unquote anywhere, anywhere that they’re selling their vehicle platforms. We work with the OEM to determine, um, you know, what’s the frequency by which the over-the-air updates will be deployed. Of course there are, you know, the roads are always changing. Um, ods are often expanding OEM footprints, of course, they’re gonna wanna make them as, uh, sprawling and diverse as possible. And so the over-the-air upgrades will continue to accelerate both, uh, increased performance for edge cases. And, uh, unusually encountered things. It’ll also increase for a diversity of, uh, footprint and use cases and also the way that roads change, right? Uh, we could drive in a city. Well, I’m sitting here in London at our headquarters, and I can say that, uh, when I traveled to work last week when I was commuting, this traffic certainly looked different and the construction looked different than even it does this week. Um, and so over the upgrades would also account for ever changing obstacles that we see on the road.

Grayson Brulte: Again, you’re scalable and that’s appealing. And if you look at the models of autonomy at emerging, you have the robotaxis market, you have the the licensing market, but then you have a new market emerging. Fleet ownership where I believe that you’re gonna have a Thomas vehicle reached back by A-K-K-R-A, Blackstone or, and Apollo. They’ll loan the asset, they’ll put it on the Uber network, the Lyft Network, or perhaps another third party network. Do you ever get to a point where Wayve licenses the technology to one of these large fleet operators, that they have 10,000 vehicles, 20,000 vehicles? 

Kaity Fischer: Yeah, so it’s, well, the world that you’re talking about, I mean, it exists today, right? They’re just human driven, um, where vehicles are, well, autonomous vehicles would essentially become their own asset class, right? I mean, we saw this in New York City with the taxi medallion market. But for Wayve’s technology specifically, we license directly to the OEMs. That’s our, our end customer. And so we are, you know, supplying that vehicle intelligence. And then OEMs are the ones who would be selling to the end vehicle owners. Now in today’s world, we see a lot of pragmatic benefits about full L four autonomy, starting with fleet ownership. But the beauty of our approach to going to market is that because we start by licensing, you know, really perform an L two plus and L three to consumer owned vehicles. And then we’ll be licensing L four technology as well. Um, from Wayve’s perspective, we’re able to tap into a really incredible market. Everything from personally owned vehicles where people become familiar and comfortable and trusting of autonomy at that L two plus through L three. Um, building the desire for L four, um, all the way through the L four use case. I think that L four, again pragmatically will probably start as fleet o fleet ownership. There’s economic and maintenance, um, practicalities associated with that in the early days, but I don’t think it’s outta the question that L four will eventually also become, uh, a personally owned vehicle reality. I.

Grayson Brulte: It’s gonna come I, I see an emerging market, and I’ve said this very publicly on road autonomy and I harp on it every week pretty much practically on autonomy markets that. At some point, Wayve will get an OEM deal with an OEM and they will use your technology to deploy robo taxis. And this week Andrew McDonald from Uber visited you in London and he put up a post on X, and I’m gonna quote him here. Next, stop working with Wayve and Automakers put self-driving vehicles on the Uber network in multiple markets around the world. It sounds like my inclination might be coming true based on what Mr. McDonald posted on x.

Kaity Fischer: You are very observant. I should should trust that you are always, uh, coming through all of the latest announcements. Look, we were, uh, thrilled to host Andrew this week. We always love having himself or anyone else from Uber coming through our offices, whether that be, you know, in our, with our San Francisco fleet or getting them in the cars here in London or Germany or Tokyo. \ Look, Uber has been an incredible partner and an incredible investor. The chapter of our partnership, because it’s been multifaceted and, and longstanding chapter that we’re currently in, that we’re really excited to bring to fruition is our work alongside OEMs. So being able to deploy an incredibly performant and safe L four system alongside a production vehicle, which would come from an OEM, um, and ensuring that it’s fit for purpose for a ride, hail use case that Uber can put into their, you know, existing fleet. It’s an exciting reality and it’s one that we’re actively working towards and, um, you know, watch this space. I think what’s been really exciting to see over the last couple of years is that there are many automakers who have L four performant vehicles either ready today or ready on a very short time horizon, um, that our production grade coming off assembly lines, which changes the game for scale of deployment. So we’re, we’re looking forward to being able to share more.

Grayson Brulte: I am excited to, to read more, and I’m not gonna associate this with you, but we did get a filing last week. Foxconn is officially in the automotive business. They’re gonna design and manufacture vehicles from Mitsubishi for the New Zealand and Australia markets. Do we ever get to a point where you would consider licensing your technology to a non-traditional OEMI? 

Kaity Fischer: right? I mean, they’ve got a hundred year legacies in many cases of bringing a safety critical technology, validating it and bringing it to mass production scale. These are no small feats, right? And they’re, those, those skills are developed over many years. Um, and really. Depth and breadth of experience, , and we’re able to do the, uh, the piece that, uh, I would say is in really high demand for their consumers. So bringing that end-to-end ai, that vehicle intelligence in complementing their legacy of safe performance systems. Fox kind of always thrilled to see more folks getting into the automotive game and, uh, should they get to the point of having a performant vehicle platform that’s been safety validated. We’d be thrilled to expand our OEM partnerships and uh, and investigate deployment.

Grayson Brulte: I’ll give you a little insight. A lot of the Foxconn folks listen to the podcast. So Foxconn, if you’re smart, call Kaity over at Wayve. They’re, they’re building really great technology. That’s, that’s my pitch to them. The thing I really, really like about Wayve, you don’t have the the legacy I. Overhead, you don’t have to worry about developing a vehicle, building an assembly line. Going through, as Elon says, manufacturing. How, why was the decision made early? Early on to focus on purely building an end-to-end AI driver and not trying to build a full blown automotive manufacturer.

Kaity Fischer: It’s a great question. Um, so maybe a bit of background. So Shockingly Wayve is the third autonomous player that I’ve worked at now, which is crazy ’cause. When I started my career, this career didn’t even exist, so I started in automotive. I got lucky enough to fall into autonomy, and now here I am at Company three. And one of the things that brought me to Wayve beyond the underlying technology being incredible, but is the fact that Wayve has always been firm and their belief of taking a partner first approach. And I think that that is so critically important to being successful and focused in this market. Um, and transparently some of that, you know, some of it’s strategic decision and some of it’s the benefit of timing to market. So being able to stand on the shoulders of others and learn from the experience of, you know, how incredibly difficult and challenging and multifaceted it is to be vertically integrated. Um, and what Wayves, bread and butter is what our core focus is, and how we stay laser focused on accelerating the most performant safe. Autonomous driver is by staying focused on vehicle intelligence. So that means that we don’t plan to build any hardware. We don’t plan to be, you know, the demand capture mechanism. That’s why we partner with folks like Uber and others. Um, and what this allows us to do is be really agile and be really best in class and what, um, you know, what our specialization is. And then we can partner with people who are best in class in every other area of the value chain. So whether that be Microsoft who’s a, you know, a partner and an investor, um, on the cloud side, whether that be Nvidia, who also happens to be a partner and an investor, um, for SOC, um, or whether it be partners like Uber who help us on our path to market and accelerating that we’re able to work with the best of the best and we look at it as all ships rise by not having to do it all yourself in-house. So it’s been an exciting ride and I think it’s one that thus far has proven to be really successful.

Grayson Brulte: You made the right decision. ’cause if you look back at your story career, you worked at an AV developer owned by an OEM. We saw how that went. Thank goodness you left and you worked at another AP developer that that’s developing an antiquated stack and now you’re at an AV developer that is scaling. You have the battle scars. Do you, when you meet with the executive team saying, okay, this is all the stuff I’ve learned from two different approaches, and so that the Wayve does not make the same mistakes that those your previous two companies made.

Kaity Fischer: Another, you know, element of Wayve that I think is unique and is also one of our keys to success is that our team is incredibly diverse but across a couple of axes and it’s a differentiator and a game changer to avoid unoriginal mistakes, to be honest. Um, so we’ve got, of course a really deep talent base that’s world leading and cutting edge on the AI front, ’cause I mentioned that’s our bread and butter. We’ve also got these two other axes or, or groups of talent and it, one is automotive and one is autonomy. And so we’ve brought together teams and experience and scar tissue, um, but also lessons learned and inspiration from these three different industries. And bringing all three of those together, I think is how we again, avoid on original mistakes. Also think differently about the mountain ahead of us. It is both, you know, an incredible challenge. It’s, you know, people talk about as like the space race of a lifetime, but it’s also an incredible opportunity. And so bringing together, uh, those three different types of folks, it makes for incredibly dynamic and interesting conversations in the room. And it keeps us all honest so that we, uh, we don’t repeat history.

Grayson Brulte: It keeps you on your toes.

Kaity Fischer: That is also true. I could not, I can say that happens daily around here.

Grayson Brulte: Which is wonderful, thinking differently and staying ahead of the curve. This week we started getting public statements and we got a post on X from your investor and partner Nvidia, talking about a mapless future. That’s a big pivot for Nvidia because. They bought James Woo’s Deep Map back in the day, and they did a really good things with James and Mark Wheeler there. And now that was, they were all of HD map. HD map when they acquired that great exit for James and Mark, by the way. And now they’re making public statements around map list. Are you influencing that? Are you showing them the technology technological breakthroughs that are possible with a mapless approach? 

Kaity Fischer: I mean, I mentioned NVIDIA’s been an incredible partner. We’re always in admiration of just how innovative and forward thinking, and honestly, they’re constantly reinventing how they think about the market to stay at the cutting edge. Um, so I would say I’d love to see that they’re moving towards a mapless future. I think the industry is largely coalescing around the idea that mapless is truly required in order to scale. I mentioned that OEM customers, they expect a technology that works for their consumers who live drive. Traverse Roads everywhere. mapless is a key piece to unlocking that. It means that, you know, if a consumer’s buying the automated technology, they can take the vehicle everywhere that they need to get. And it means that if it’s a fleet, they have a path to a profitable business. Because they’re not constrained to a subset of their business demand, they can use that fleet, um, and amortize their costs against their most valuable miles because they’re not restricted to a hundred miles or 200 square miles. Um, of terrain. So thrilled to see Nvidia also moving to, to a map list future and, um, you know, we’ll continue on the same track.

Grayson Brulte: Mapless is scalable. Pre-Mapping is not scalable. If you look at it from a cost structure, it, it’s, it’s very expensive. From an operational, it’s very cumbersome, so you have to get to a mapless future for a viewer or a listener that’s might be saying, oh, what’s mapless, how would you describe a, a MAP plus approach to autonomy? 

Kaity Fischer: I think to de describe a mapless approach, you might have to start by describing, uh, a mapped approach. So the way I’d describe it, uh, for those who aren’t familiar. With the mapping industry is with high definition maps or what we call HD maps. Uh, what it requires is that a vehicle that is equipped with very specialized high definition sensors has to traverse every square inch of roads that it wants to be able to drive on in autonomous mode. So it has to traverse those roads multiple times over. That means that you’ve gotta build out vehicles to go acquire those mapping miles. It means that as the roads change, which. I mentioned my commute earlier and changing traffic and changing curbside and, uh, you know, construction has those roads changed. You have to remap. So the maps have to constantly be maintained. You have to pay for the operations to go capture those mapped miles. And also it’s really a challenge from a regulatory point of view because street owners don’t want vehicle miles traveled, um, on the road adding to congestion, um, potentially impacting safety, et cetera. So it’s a challenge from a lot of different angles to require a mapped approach. And the culmination of that is. That the only places that you’re able to operate your technology are the ones that have maps that are updated and accurate, which means you have to operate in a geofence. So now that we see what the mapped world looks like, um, how would I describe a mapless world? If you’re able to take a vehicle and traverse new areas, sight unseen, and still have a performant safe technology, so that means that you can drive your vehicle without boundaries, um, wherever you might need to take it. Whether that be for, um, you know, to the airport on a long haul, commute on a road trip, or just your daily errands around town means that you’re able to leverage your vehicle autonomy to use your vehicle exactly how you see fit.

Grayson Brulte: That’s valuable at the end of the day that that’s truly valuable on the vehicle. Uh, I own a vehicle. You own a vehicle. Other, uh, individuals. A lot of, majority of individuals own vehicles, but they’re all different types. You might have an SUV, I might have a sedan, another neutral might have a, a larger vehicle or, or a crossover or a hybrid. How do you adapt the Wayve driver to various different vehicle platforms? If you look at Nissan, your part Nissan, they make all sorts of different types of vehicles.

Kaity Fischer: Yeah, this is a great question. So we talked a bit about the generalization to new geographies and I mentioned, uh, with 500 hours worth of data we could take our vehicle from the UK to the US and it was driving the same level of performance. Another key differentiator of our technology is we can generalize across vehicle platforms. So when we work with OEM partners, that means that we can use the same vehicle intelligence across, um, you know, a sedan or an SUV or a truck. Um, because our technology is decoupled, our software technology is decoupled from hardware. So we can, uh, work with OEMs to deploy our vehicle intelligence. Across many vehicle platforms and what we’ve shown is that um, it actually takes about a hundred hours of sample data in order to get the same level of performance on vehicle model number two, as vehicle model number one. And when we take data from vehicle model number two, uh, in trainer models with it performance for vehicle model two and vehicle model one, both rise. So that diversity of data is, allows us to accelerate performance across a variety of different vehicle platforms. So generalization isn’t just geographically, it’s also on different hardware, uh, in different vehicles. So when we get back to that point of scalability, OEMs love this because an investment in one technology is able to be deployed against many different models and lines.

Grayson Brulte: I’ll give you another kid’s term happy camper. That’s really, 

Kaity Fischer: Love that.

Grayson Brulte: that’s what it comes down to. How do the sensors work? Does, does Wayve recommend sensors? I know you, you use the existing sensors. Do you recommend. When you’re developing or co-developing certain placements, or how does that come together? 

Kaity Fischer: We are flexible for all different sensor modalities, um, and layouts. So when we partner with an OEM, what we’re able to do is we’re able to work alongside them to say, what sensors do you already have planned on your production vehicles? Here’s the level of autonomy that we’re able to deliver, both relative to the SOC that you have on board, as well as the sensor layout that you have existing. What this enables OEMs to do, I mentioned that we can decouple hardware from the software, is they can choose and specify their hardware layout for their vehicles, both in terms of their preferred sensor suite, whether that be, you know, radar vision, whether that be camera, lidar, radar, um, you know, different levels of performance on their SOC, depending on the price point for the end consumer vehicle. We can work with what they have existing and offer different autonomy solutions. Um, for each level. You can think that performance kind of naturally scales relative to, um, to the level of performance of the SOC. And then when we think about different sensor layouts, um, for level two plus. There’s an incredible amount of diversity that supports a really perform an L two plus system. And then when we get into L three and L four, where we remove the driver as the redundant system and we need to build in more redundancy on the platform, um, that’s where there’s some additional minimum requirements for redundancies. But ultimately, we put that decision and that that power back into the OEM’s hands to make the decision on what’s best for them based on their desired bound cost, um, and also their liability, uh, and safety requirements.

Grayson Brulte: Is it fair to say that Wayve is sensor agnostic? 

Kaity Fischer: We often use that term. Yes, we say sensor agnostic. And um, that means, I mean, even today, uh, what we ingest in terms of data on road data, we often talk about vision. I think it’s the most easy to wrap our minds around as like, oh, vision cameras. Yes, we know that those are equipped on vehicles today. Um, but we also work with our OEM partners to ingest not only vision or camera data from on-road data. Radar data, LIDAR data, these are all inputs that we’re able to ingest in order to increase the performance of performance of our vehicles. Um, and. A differentiation of our approach is that we’re actually able to work with even like non OEM partners in order to ingest data. So we work with third party dash cam providers, uh, to really get the diversity of data at scale. So, uh, let’s call it like, like huge volume of data when we think about getting these edge cases and the diversity, um, from many different geographies. We ingest even, uh, things like the UK Highway code. So that’s, you know, the rules of the road for the uk, that’s a written text and we’re able to ingest it. So because we use end-to-end ai, we have multiple different modalities that can contribute to our vehicle intelligence, including whatever sensors that OEMs have on their vehicles. We 

Grayson Brulte: If Wayve didn’t take an end-to-end approach, would you be able to scale as fast as you are today? 

Kaity Fischer: believe that end-to-end AI is a key unlock to scalability. From a generalization perspective, um, from a commercial and economic perspective and from an ultimate operational perspective, again, removing the requirements around things like geofences, it truly is a game changer compared to the traditional approach to autonomy.

Grayson Brulte: And, you know, you’re onto something when you start seeing marketing messages. And, and I’m gonna name it here because I literally fell outta my chair. Me, I, you call me Sherlock Holmes, or, or, or Clue Holmes Waymo, for the first time last week, used the term generalized AI driver In a public statement, I said, holy wowsers, it validates your whole approach. That Wayve has been pioneering since the beginning. You’re at the top of the Wayve. We’re, we’re getting a really good break here and, and you’re cruising. Just, are you feeling the more momentum in your back as you’re starting to see the terms that Wayve is pioneered and the development techniques that the pioneered is now starting to come into the mainstream of the autonomy market? 

Kaity Fischer: I will call you Sherlock on that one. Grayson. Um, again, points for Observ. So, yeah, so of course it, it feels great that these types of terms are becoming more mainstream in the industry. Um, I think also having clarity and greater understanding of what these terms mean benefits everybody. So we all know what to expect, um, and what’s, what should be desired and what’s important metrics for success when we’re developing autonomy technologies. Right. Um, but ultimately I also, I, I look at. Wind at our back. Right now, I think of for the entire autonomous industry, it feels like we’re at that moment where you’re right at the top of a rollercoaster and the gravity is about to pull you over and you can feel it. The anticipation is so thick, I feel like we’re right at the crest at the top of the rollercoaster, and it’s, it’s drilling and it’s validating to see the industry as a whole, um, be right on the edge of. Through deployment at scale so we can bring these benefits to reality. So that’s the piece I think of when I think of winded our back. And I also, the more people that adopt, you know, similar approaches or that value, the right factors that allow us to democratize access to this technology, the better it is for everybody in all ships rise. So while of course, yes, I’m thrilled that other folks are using the term generalization, um, I think about the impacts to the industry and getting this, this technology on the road, uh, as the ultimate win.

Grayson Brulte: The industry is, it’s, it’s clearly going, you’re right. We’re about to take off like a SpaceX rocket ship. But the other interesting thing that’s happening, I was at dinner last night speaking to an individual they knew all about, they all they knew about avs. I said, what the heck? It, it start, it’s before, oh, what’s this? Oh, you’re the robot guy? No, not the robot guy. They start asking, there’s a high level of interest now from the general public. Around autonomy and being able to experience experience rides since the, the, the game plan is, let’s call spade a spade. You’re gonna work with multiple OEMs, a across the world from a consumer standpoint. How will the consumer know that? Let’s call it Acme. OEM system is powered by Wayve. How are they gonna know that your technology’s embedded in a product that they’re either leasing, buying, or riding it in a robo taxi? 

Kaity Fischer: I mean, our business model is B two, B2C, essentially, right? The OEMs. Are our end customer. Um, however, that consumer acceptance and trust and demand is also something that’s critically important to, uh, you know, uptick or I should even say, uh, for opting in, uh, of the technology. And so one of the things that we’re. Really excited about, and again, one of the many benefits about our approach in go to market of starting with L two plus and traversing rapidly into L four is that we’re able to give consumers exposure to a performant, autonomous system in a way, um, where they still are monitored, monitoring the system where they’re engaged. That allows us to build trust, to build understanding of what they want and need in order to, uh, opt into the technology in order to make sure that they’re, they’re using it at every chance that they can. Those are hugely valuable learnings for Wayve to develop something. That OEMs and consumers want, um, and also for end consumers to get familiar and build trust and know what it is that they want out of a product. I mean, we’ve heard that classic iPhone story, right? Of, you know, uh, someone goes to poll, you know, consumers and they say. Would you ever want a phone without buttons? This is, I’m dating myself here, but that’s like back in the early two thousands. Uh, would you ever want a phone without buttons? And of course everybody says, absolutely not. I need my phone to have buttons. How would I text otherwise? And then of course, the iPhone comes out. And now look, I don’t, I don’t think I know a single person who has a phone with buttons anymore. I think autonomy is similar. There are very, very few proof points in the market. Of what great looks like, and that’s what we’re really excited to deliver, to defining a new market category where consumers know to demand it because it’s the most performance solution out there. So those are the things that, that we look forward to.

Grayson Brulte: And it’s also the safest. I almost got kidnapped last night. I took a ride share V.

Kaity Fischer: Oh, this took a turn.

Grayson Brulte: Yeah, I took a ride. I took a ride share ride home. I go out, I have these two gentlemen, and we go out once a week together and we have one rule that we go out, nobody drives. That’s the rule I. ’cause we like to drink wine. So that’s the world nobody drives. And this, 

Kaity Fischer: Stage advice though. Yes.

Grayson Brulte: it’s very good advice. And all the bartenders know they’re proud of me and I get a little heavier or poor because they know I don’t drive. And the, and I’m not gonna name the name this company outta respect for the company. And the individual gets in, I get in the car and I live on an island. He starts trying. I’m like, whatcha are you doing? Whatcha doing? Whatcha doing? I pull over meat, I get out and I start running. But if I was in an av, that situation wouldn’t happen. So it’s not just the technology breakthroughs. Avs are gonna make our roads safer. They’re also going to make. The experience safer for passengers, such as the incident that I had last night. What we do know, Kaity, is Wayve is building the future. What I’d love to see more than anything is, is a Wayve company shot of everybody riding a big break, either in the Maldives or or somewhere in Australia and, and, and doing it. Right. In your opinion, what is the future of Wayve? 

Kaity Fischer: Future of Wayve. The future of Wayve is where? Maybe lemme take a step back. So, um, I happen to also be six months pregnant with my first child. So we’re very excited. Um, and I think of it as the personal goal for me, or how I relate this back to myself is. I hope my daughter never has to get a driver’s license unless she chooses to. And I hope that, you know, when I’m retired, many, many years from now, and you know, I’m telling my, my child or my children maybe at that point telling them about, you know, what I did in my career. And I say, oh, and I was part of the first deployments of, you know, autonomous technology. And my kids or my grandkids are gonna look at me and they’re gonna say like. Cool grandma. Like the thing that takes us to school every day, like old news and they’re gonna think that I did something really boring for my job and meanwhile that will be the mark of success because it will be so normalized. It’s not noteworthy. And, um, I think of that as, as the ultimate future. Um, but perhaps more specifically, you know, to Wayve’s journey. This vehicle intelligence right now, it’s exactly that vehicle intelligence. But what we’re building is a foundation model that has generalized intelligence. And so this, our first use case in the first market that we think will have huge benefits is on-road autonomy, is putting this in vehicles. The intelligence in, you know, years from now, it can be adapted to anything where AI meets robotics. So it can power, um, off-road vehicles. It could power vehicles that are in a warehouse so that our Amazon two day packages or Amazon one day packages instead, um, it could power humanoid robotics. And so the ultimate, you know, the, the North Star vision, once, once we earn the right by deploying a really perform and safe on-road technology. Is to be that intelligent, trusted AI partner throughout life. Whatever use case that takes.

Grayson Brulte: It’s beautiful and congratulations. Congratulations on your, your pregnancy. I have a young daughter and she went in her first AV at two. And she’s been in every Waymo from the fire fly up, and she’s been in all the shuttles and she’s been in the, the autonomous stuff. Remember all those old vehicles she was in the motional vehicles, the Waymo vehicles. She’s been in every AV that’s on the road and I have a wall in my studio with all these different avs in the years that, that she went in and it’s like, oh, as she grows up, all the vehicles change. And it’s fun because. When you expose children to, to this technology, they talk about it. And she did a presentation, I think it might have been third grade, fourth grade, where she talked about avs. And then daddy got to go in to talk all about avs and, and it sparked its imagination because when you expose children to this technology. I believe you. You’re gonna unlock something, you’re gonna knife something and they’re gonna go on to change the world and they’re gonna build a better, better world. And the more automation we have in the world and I call a dad mode, the better I’m gonna sleep at night if they’re out. ’cause I know okay where you are and if you’re me being a dad of a daughter, oh, you’re gonna your boyfriend’s house, boop, lock doors, turn car, come home. I call that dad mode and that’s gonna be a really great thing. But it’s going to make it safer and I’m not gonna have to go through the anxiety and stress that my parents went. Are you getting in a car with a drunk driver? Because avs don’t drink and drive, and then you’re really truly building the future. I wish you nothing but success at Wayve and I can’t wait to go surfing with you and the Wayve team as we look to wrap up this insightful conversation today. For today, Kaity, what would you like our listeners and viewers to take away with them? 

Kaity Fischer: Oh, just one thing that I hope that. Folks take away is, you know, the era of an AI driven vehicle. It’s, it’s here, it’s now, so Don’t sleep on it.

Grayson Brulte: Don’t sleep on it. End-to-end. AI is the future. mapless is coming. It’s the only way to scale autonomy profitably. The future is bright. the future autonomous, the future is Wayve. Kaity, Thank you so much for coming on The Road to Autonomy today.

Kaity Fischer: Thank you for having me. Loved the conversation. Always glad to chat.

Grayson Brulte: It was fun. Bye.

Key The Road to Autonomy Episode Questions Answered