UK-based autonomous vehicle startup Wave started life as a software platform loaded into a tiny electric “car” called Renault Twizy. Complete with cameras, the company’s co-founders and PhD graduates, Alex Kendall and Amar Shah, fine-tuned the deep learning algorithms that power the car’s autonomous systems until they can navigate a medieval city without assistance.
No fancy cameras or radars were needed. Suddenly they knew they had something.
Fast forward to today and Wayve, now an AI modeling company, has raised a $1.05 billion Series C funding round led by SoftBank, Nvidia and Microsoft. This makes this the UK’s biggest AI fundraiser to date and among the top 20 AI fundraisers worldwide. Even Meta’s head of AI, Yann LeCun, invested in the company when it was young.
Wayve now plans to sell its self-driving model to various car OEMs as well as manufacturers of new autonomous robots.
In an exclusive interview, I spoke with Alex Kendall, co-founder and CEO of Wayve, about how the company is training the model, new fundraising, licensing plans and the broader self-driving market.
(Note: The following interview has been edited for length and clarity)
TechCrunch: What drove the balance to achieve this level of funding?
Kendall: Seven years ago, we started the company to build an embedded AI. We deal with building technology with our heads down […] What happened last year was that everything really started to work […] All the elements required to make this product a reality [came together]and, in particular, the first opportunity to develop embedded AI at scale.
Now production vehicles are rolling out with GPUs, surround cameras, radar and, of course, the appetite to now bring AI into and enable, an accelerated journey from assisted to automated driving. So this fundraiser is a validation of our technology approach and gives us the capital to turn this technology into a product and bring it to market.
Very soon you will be able to buy a new car and it will have Wayve’s AI […] Then this goes into enabling all kinds of embedded artificial intelligence, not just cars, but other forms of robotics. I think what we want to achieve here is to go far beyond where artificial intelligence is today with language models and chatbots. To really enable a future where we can trust intelligent machines that we can delegate tasks to and, of course, can improve our lives. Self-driving will be the first example of this.
How did you train the self-driving model over the past two years?
We partnered with Adsa and Ocado to collect data for the autonomy of the tests. That was a great way to get this technology off the ground, and it continues to be a really important part of our development story.
What is the plan for licensing AI to OEMs, car manufacturers? What will be the benefits?
We want to enable all car manufacturers around the world to work with our AI, of course, in a wide variety of sources. More importantly, we will receive different data from different cars and markets, and this will produce the most intelligent and capable embedded AI.
What automakers have you sold it to? Who did you land?
We work with a number of the world’s top 10 car manufacturers. We are not ready to announce who they are today.
What moved the needle for SoftBank and other investors in terms of your technology? Was it because you are essentially platform independent and every car will now have cameras around it?
This is largely correct. SoftBank has publicly commented on its focus on artificial intelligence and robotics and self-driving [tech] it’s just the intersection of that. What we’ve seen so far with AV 1.0 approaches is where they throw all the infrastructure, HD maps, etc., into a very limited environment to prove that technology. But it’s a very long journey from there to something that can be developed at scale.
We found that — and this is where SoftBank and Wayve are fully aligned in the vision of creating autonomy at scale — by deploying this software and a diverse set of vehicles around the world, millions of vehicles, not only can we build a sustainable business, we can also to receive various data from around the world to train and validate the safety case so that we can develop AV at scale through driving around the world.
This architecture works with on-board intelligence to make its own decisions. It’s video-trained as well as language-trained, and we also bring reasoning and general-purpose knowledge to the system. So it can deal with the distant, unexpected events you see on the road. This is the road we are on.
Where do you see yourself in the landscape right now in terms of what’s already developed out there?
There’s been a bunch of really exciting proof points, but self-driving has largely taken off in three years, and there’s been a lot of consolidation in the AV space. What this technology represents, what AI represents, is a complete game changer. It allows us to drive without the cost and expense of lidar and HD. This allows us to have the built-in intelligence to operate. It can handle the complexities of fuzzy lane markings, cyclists and pedestrians, and it’s smart enough to predict how others will move so it can negotiate and operate in very tight spaces. This makes it possible to deploy technology in a city without causing stress or anger around you and drive in a way that conforms to the driving culture.
You did your first experiments in those years, enriching the Renault Twizy with cameras. What will happen when car manufacturers put a lot of cameras around their cars?
Car manufacturers are already building vehicles that make this possible. I won’t name brands, but pick your favorite brand, and especially with the higher-end vehicles, they have surround cameras, surround radar and integrated GPU. All of these are what make it possible. They’ve also now implemented Software Defined Vehicles so we can do over-the-air updates and get data from the vehicles.
What was your “book”?
We have built a company that has all the pillars needed to build it. Our playbook was AI, talent, data and computing. On the talent front, we’ve built a brand that sits at the intersection of AI and robotics, and we’ve been fortunate to bring some of the best minds around the world to work on this problem. Microsoft is a long-standing partner of ours, and the amount of GPU compute they give us in Azure will allow us to train a model at the scale of something we’ve never seen before. A really massive, embedded AI model that can actually generate the safe and intelligent behavior we need for this problem. And then Nvidia, of course. Their chips are the best in their class on the market today and make the development of this technology possible.
Will all the training data you get from the brands you work with be mixed into your model?
Correctly. This is exactly the model we were able to prove. No car manufacturer is going to produce a model that is safe enough on its own. Being able to train an AI on data from many different car manufacturers will be safer and more efficient than just one. It will come from more markets.
So you will essentially be the owner of possibly the largest amount of driving training data in the world?
This is certainly our ambition. But we want to make sure that this AI goes beyond driving — like a true embedded AI. It is the first visual language action model that can drive a car. It is not only trained on driving data, but also on text and other Internet-scale sources. We even train our model on the PDF documents from the UK government that tell you the highway code. We go to different data sources.
So it’s not just cars, but robots too?
Exactly. We build the embedded AI foundation model as a general-purpose system that trains on very diverse data. Consider home robotics. The data [from that] it is varied. It’s not some confined environment like construction.
How do you plan to scale the company?
We continue to grow our AI, engineering and product teams both here [in the U.K.] and in Silicon Valley, and we just started a small group in Vancouver as well. We’re not going to “scale” the company, but we’re going to use disciplined, purposeful growth. The headquarters will remain in the UK
Where do you think the centers of talent and innovation are in Europe for artificial intelligence?
It’s very hard to find 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 top five or six hubs in the world. London has been a great place for us so far. We started from academic innovation in Cambridge. Where we are now in the next chapter is a somewhat less traveled road. But in terms of where we are now, it’s been a brilliant ecosystem [in the U.K.].
There are many good things to be said about corporate, legal and tax. In terms of regulation, we have been working with the government for the past five years on the new autonomous driving legislation in the UK. It has passed the House of Lords, it has almost passed the House of Commons and it will soon come into force and make all this legal in the UK The ability of the government to lean on this to work with us […] We have really worked in the weeds on this and had over 15 ministerial visits. It’s been a really great partnership so far, and we’ve definitely felt the government’s support.
Do you have any comments on the EU’s approach to self-driving?
Self-driving is not part of the AI act. It is a specific industry and should be regulated with subject matter experts and as a specific industry. It’s not uncoordinated, and I’m glad about that. It’s not the fastest way to innovate in certain industries. I believe we can do this responsibly by working with specific automotive regulators who understand the problem. So domain specific setup is really important. I am glad that the EU has adopted this approach to self-driving.
