UK-based autonomous vehicle startup Wayve started as a software platform integrated into a tiny electric car called the Renault Twizy. Decked out in cameras, the company’s co-founders and graduate students Alex Kendall and Amar Shah tweaked the deep learning algorithms that power the car’s autonomous systems until they got it to drive around the medieval city unassisted.
No fancy lidar cameras or radars were required. Suddenly they realized that they were on to something.
Fast forward to today: Wayve, now an AI flagship company, has now raised a $1.05 billion Series C funding round led by SoftBank, NVIDIA and Microsoft. This makes this the largest AI fundraiser in the UK to date and one of the top 20 largest AI fundraisers worldwide. Even Meta’s AI boss Yann LeCun invested in the company when it was still young.
Wayve now plans to sell its autonomous driving model to various automobile manufacturers as well as manufacturers of new autonomous robots.
Alex Kendall CEO, Wayve
In an exclusive long-form interview, I spoke with co-founder Alex Kendall, co-founder and CEO of Wayve, about how the company trains the model, the new fundraising, licensing plans, and the broader self-driving vehicle market.
(Note: The following interview has been edited for length and clarity.)
TechCrunch: What was the decisive factor in reaching this donation amount?
Kendall: We founded the company seven years ago to develop embodied AI. We have dealt intensively with building technology […] What happened last year was that everything really started to click […] All the elements needed to make this dream product a reality [came together]and in particular, the first opportunity to deploy embodied AI at scale.”
“Now their production vehicles are coming out with GPUs, ambient cameras, radar and of course the desire to now bring AI to market and enable an accelerated path from assisted to automated driving.” So this fundraiser is a validation of our technological approach and gives us that Capital to turn that technology into a product and bring that product to market.”
“Very soon you’ll be able to buy a new car that has Wayve’s AI…Then it’s about enabling all kinds of embodied AI, not just cars but other forms of robotics.” I think the last thing we What we want to achieve here is to go far beyond the current state of AI with language models and chatbots. But really enabling a future where we can trust intelligent machines that we can delegate tasks to, and of course that can improve our lives, and self-driving will be the first example of that.”
TC: How have you trained your self-driving model over the past few years?
Kendall: We worked with Tesco and Ocado to collect data on test autonomy. This was a great opportunity for us to bring this technology to market and it continues to be a really important part of our growth story.
TC: What is the plan for licensing AI to OEMs and car manufacturers? What will be the benefits?
Kendall: We want to enable all car manufacturers around the world to work with our AI, of course across a wide range of sources. More importantly, we will receive diverse data from different cars and markets and this will produce the smartest and most powerful embodied AI.
TC: Which car manufacturers did you sell it to? Who did you end up with?
Kendall: We work with a number of the ten largest automobile manufacturers in the world. We are not ready to announce who they are today.
TC: What motivated Softbank and the other investors regarding your technology? Is it because they are virtually platform independent and every car now has cameras?
Kendall: That is largely true. SoftBank has been public about its focus on AI and robotics and self-driving [tech] is just the intersection of it. What we’ve seen so far with the AV 1.0 approaches is that they bring all the infrastructure, HD cards, etc. into a very limited environment to test this technology. However, there is still a long way to go from there to something that can be used on a large scale.
We found that by using this software and a diverse set of vehicles around the world, millions of vehicles, we could not only build a sustainable business – and this is where SoftBank and Wayve are aligned in their vision to create autonomy at scale , completely agree In the company, we can also get diverse data from around the world to train and validate the safety case to be able to deploy AV on a large scale by driving around the world “hands-free and eyes-free”.
This architecture works with the intelligence on board to make its own decisions. It is trained on both video and voice, and we also bring general reasoning and knowledge to the system. So it can handle unexpected long-tail events that you see on the road. This is the path we are on.
TC: Where do you currently see yourself in the landscape in terms of what is already being used there?
Kendall: There’s been a lot of really exciting evidence, but autonomous driving has largely plateaued for three years, and there’s been a lot of consolidation in the AV space. What this technology, what AI represents, is that it is completely groundbreaking. It allows us to drive without the cost and expense of lidar and HD. This means we have the necessary on-board intelligence for operations. It can handle the complexities of unclear road markings, cyclists and pedestrians, and is intelligent enough to predict how others will move, allowing it to negotiate and work in very tight spaces. This makes it possible to use technology in a city without causing fear or disruption on the road, and to drive in a way that is consistent with driving culture.
TC: Back then you did your first experiments and packed the Renault Tizzy with cameras. What happens when car manufacturers put lots of cameras around their cars?
Kendall: Car manufacturers are already building vehicles that make this possible. I wouldn’t name any brands, but choose your favorite brand and especially on higher end vehicles they have surround cameras, surround radar and an integrated GPU. All of this makes this possible. Also, they have now introduced Software Defined Vehicles so we can do over-the-air updates and pull data from the vehicles.
TC: What was your “playbook”?
Kendall: We have built a company that has all the pillars necessary to build it. Our playbook was AI, talent, data and computing power. In terms of talent, we have built a brand that represents the intersection of AI and robotics, and we have been fortunate to attract some of the best minds from around the world to work on this problem. Microsoft has been a long-time partner of ours, and the amount of GPU computing power they are making available to us in Azure will allow us to train a model at a scale we have never seen before.
A truly enormous, embodied AI model that can actually develop the safe and intelligent behavior that we need for this problem. And then of course NVIDIA. Their chips are the best in class on the market today and enable the use of this technology.
TC: Is all the training data you receive from the brands you work with merged into your model?
Kendall: That’s right. We were able to prove exactly this model. No single car manufacturer will produce a safe enough model on its own. The ability to train an AI using data from many different car manufacturers will be safer and more powerful than just one. It will come from more markets.
TC: So you actually have probably the largest amount of driving training data in the world?
Kendall: That is certainly our ambition. But we want to make sure this AI goes beyond driving – like real embodied AI. It is the first vision-voice-action model capable of driving a car. It is trained not only on driving data, but also on Internet-scale texts and other sources. We even train our model using the UK Government PDF documents that tell you the Highway Code. We use various data sources.
TC: So it’s not just cars, but also robots?
Kendall: Exactly. We build the embodied AI foundation model as a general-purpose system trained on very diverse data. Think about home robotics. The data [from that] is diverse. It is not a restricted environment like manufacturing.
TC: How do you plan to scale the company?
Kendall: We continue to expand our AI, engineering and product teams here [in the U.K.] and in Silicon Valley, and we also just founded a small team in Vancouver. We will not scale the company “at lightning speed”, but rather pursue disciplined, targeted growth. Head office remains in the UK
TC: Where do you think the AI talent and innovation centers are in Europe?
Kendall: It’s quite difficult to search anywhere outside of London. I think London is by far the dominant place in Europe. We’re based in London, Silicon Valley and Vancouver – probably the five or six largest centers in the world. London has been a great place for us so far. We originally emerged from academic innovation in Cambridge. Where we are now with the next chapter is a slightly less well-trodden path. But given our current situation, it’s a brilliant ecosystem [in the UK].
I think there is a lot of good things to say in the areas of cooperation, law and tax. On the regulatory front, we have been working with the government for five years on new autonomous driving laws in the UK. They have been passed by the House of Lords, have almost passed the House of Commons and should soon come into force to make this all legal in the UK. The opportunity for the government to get involved and work with us […] We worked really hard for this and had over 15 ministers visit. It’s been a really great partnership so far and we’ve definitely felt the support of the government.
TC: Do you have a comment on the EU’s approach to autonomous driving?
Kendall: Self-driving is not part of the AI law. It is a specific industry and should be regulated with subject matter experts and as a specific industry. It’s not an uncoordinated catch-all term, and I’m glad about that. It is not the fastest way to innovate in certain industries. I think we can do this responsibly by working with specific automotive regulators who truly understand the problem area. Therefore, sector-specific regulation is really important. I am pleased that the EU has chosen this approach for autonomous driving.