Rethinking Autonomous Vehicles: A Conversation with Wayve's Kaity Fischer

In a recent episode of Conversations in the Park, host Timothy Papandreou sat down with Kaity Fischer from Wayve to discuss how the London-based company is revolutionising autonomous vehicle technology with their unique approach. Here's their conversation about the future of self-driving cars, the power of embodied AI, and why sometimes the obvious solution isn't the right one.

Timothy: Kaity, you've had quite a journey from Detroit to London. How did you end up in the autonomous vehicle space?

Kaity: Growing up in Detroit, there was no such career as working in autonomous vehicles, so it wasn't exactly on my hit list. I started at a tier one automotive supplier and was fortunate enough that when they acquired two autonomous vehicle startups - one from 美国卡内基梅隆大学 and one from 美国麻省理工学院 - they tapped me to handle the business side. That first company is now known as Motional. Three autonomous companies later, here I am at Wayve, which brought me to London three years ago.

Timothy: Wayve's approach to autonomous vehicles has been described as contrarian. Can you explain what makes it different?

Kaity: The traditional approach, what we call AV 1.0, is a robotics-based system that relies on hand-coded rules. Every time the vehicle encounters something, an engineer needs to code a specific rule for that scenario. Instead, we use end-to-end AI, where our vehicles learn just like humans do - through experience and processing data from sensors. We don't need HD maps or geofencing, and we're sensor and vehicle agnostic. This means we can work with whatever sensor suite automakers are already integrating into their vehicles.

Timothy: That's fascinating. How does this translate into real-world applications?

Kaity: One of my favourite examples is watching our vehicle handle unprotected turns. Remember, nothing is hand-coded in our system. Our vehicle can recognise when another driver flashes their lights or waves their hand - very human behaviours that have nothing to do with the rules of the road. These are the nuances you'd never be able to solve through traditional hand-coding.

Timothy: Let's talk about data. How does Wayve approach data collection differently?

Kaity: We've developed partnerships with fleet owners like Asda , Ocado Group , and DPD UK . This allows us to collect fleet-scale data across regions or countries, not just one city or market, without needing to own and operate our own massive fleet. It could be dash cam video that's already being used for driver monitoring, or we might deploy our own camera solution. This gives us incredibly diverse, global-scale data, including corner cases and different sensor types, all while being incredibly cost-effective.

Timothy: You recently announced some major partnerships and funding. What can you tell us about that?

Kaity: Earlier this year, we announced our Series C, bringing in $1 billion with key partners like SoftBank Group Corp. , 英伟达 , and 微软 . We've also announced a partnership with Uber to accelerate L4 technology development. But what I'm particularly proud of is our work with the UK Government on passing the AV bill into an AV act - it's a world-first for federal-level legislation paving the way for L4 technology deployment.

Timothy: Your business model has evolved. Can you talk about the shift from L4-first to a more graduated approach?

Kaity: About a year and a half ago, we transitioned to an L2+ and L3 first approach, with the ability to upgrade to L4. While L4 remains our North Star, we realized we needed to create value for consumers and customers today. This approach gives us access to data that accelerates our technology improvement, helps us build relationships with OEMs, and allows us to grow alongside consumer understanding and acceptance of the technology.

Timothy: Tell us about Lingo - it seems like a game-changer for user interaction.

Kaity: Lingo allows users to interact with our vehicles using natural language, either typed or spoken. You can ask questions like "Why did you stop here?" or "What are you seeing right now?" It helps build trust by making AI more explainable. You can even give simple directions like "pull over behind the red bus." It's about creating a collaborative experience between human and vehicle, rather than just being a passenger.

Timothy: What advice would you give to people looking to enter the autonomous vehicle industry?

Kaity: The most important things are curiosity, resilience, and humility. We're working in an industry where the technology is developing in real-time, the policy scene is still forming, and no one has yet developed a financially viable business model. You need to be able to thrive in ambiguity and understand that getting things wrong is how we learn. We sometimes joke that if you've been to more than two AV companies, you might be a masochist!

Timothy: Final thoughts on the future of autonomous vehicles?

Kaity: We're at an incredible moment in this industry. The technology, policy, and business models are all reaching maturity. What excites me most is the potential to fundamentally improve people's lives, save lives, and change how we think about mobility and access to opportunity. But it's crucial to stay grounded - that's why we keep our teams co-located with our vehicles. We never want to lose touch with the fact that we're developing safety-critical systems that will have real impacts on people's lives.

Through this conversation, it becomes clear that Wayve's approach represents more than just a technical innovation - it's a fundamental rethinking of how autonomous vehicles should learn and operate. As the industry continues to evolve, their focus on adaptability, scalability, and human-centered design could well prove to be the key to making self-driving cars a practical reality for everyone.

????Catch the full interview on YouTube, Spotify, or Apple Podcasts????