AI Supremacy

Share this post
Wayve Partners with Microsoft to Scale Autonomous Vehicles
aisupremacy.substack.com
Prospectus

Wayve Partners with Microsoft to Scale Autonomous Vehicles

Wayve.AI it is the first to deploy autonomous vehicles on public roads with end-to-end deep learning.

Michael Spencer
May 22
3
1
Share this post
Wayve Partners with Microsoft to Scale Autonomous Vehicles
aisupremacy.substack.com

Hey Guys,

I follow autonomous driving advances pretty closely as it relates to artificial intelligence at scale. One of the startups that has always been on my radar is London based Wayve.ai.

Their AV2.0 technology completely rethinks how to solve self-driving by taking a pure machine learning approach. The result is a next generation AV technology that can scale and adapt to any driving condition, anywhere in the world: unlocking the promise of a better way forward. A.I. can power an autonomous Cargo ship at sea for 500 miles, what Robo-Taxis will do on the roads soon will be incredible.

As of the May, 2022 announcement, Wayve, the London-based company pioneering autonomous vehicles that the Virgin Group has invested in, has announced it is working with Microsoft to leverage the supercomputing infrastructure it needs to support the development of AI-based models for self-driving vehicles on a global scale.

The prospect of autonomous vehicles in our cities is good, even as it has taken a long-time for the last one percent to work itself out. Gross Ton (in Japan) vessel became the world's first commercial cargo ship to be navigated entirely by artificial intelligence (A.I), Electrek reported. The state of autonomous vehicle scalability is still very tricky.

Recently, Baidu and Pony.ai have been given permission to operate their autonomous vehicles without safety drivers in Beijing, a first for robotaxi services in China, mind you last time I checked Beijing was in a shadow semi lock-down due to Covid-19, so I’m not exactly sure what exactly is going to happen.

Wayve.AI Leans into Microsoft Supercomputer Infrastructure

I’ve always been fascinated by how different the approach of Wayve.ai has been. They have raised $456 million to date. Backers include D1 Capital, Baillie Gifford and Microsoft. Now Microsoft is partnering and together to join Wayve’s industry-leading expertise using deep neural networks and vast quantities of data to train AI models, with Microsoft’s engineering excellence in powering large-scale AI systems, including access to its Supercomputer. When designed around Wayve’s unique approach, the power of Microsoft’s supercomputing infrastructure will make it possible to bring self-driving technology to more places and customers sooner.

When designed around Wayve’s unique approach, the power of Microsoft’s supercomputing infrastructure will make it possible to bring self-driving technology to more places and customers sooner.

With all the hype around self-driving vehicles, it’s taken industry at least ten more years than many of us expected for this to finally go mainstream due to the difficulty of the problems. Wayve was founded following world-class research in deep learning from the University of Cambridge, building their first robot in a house-office garage. Kendall himself has a background in AI, with a Ph.D. in deep learning.

Microsoft has a way of supporting startups with massive supercomputing infrastructure while having exclusive rights to commercialize their products as we saw with OpenAI and the $1 Billion in funding help.

AV2.0

Wayve’s approach, which the company dubs AV2.0 (Autonomous Vehicles 2.0) as opposed to AV1.0, the term Wayve uses to refer to “classical autonomous drivers”. It’s truly more deep learning centric than most AV labs and self-driving tech startups.

The solution Wayve is pursuing is a holistically learned driver. In other words, an end-to-end deep learning-based approach to autonomous driving. When asked to make a point-to-point comparison between the AV2.0 and AV1.0 architectures, Kendall responded that AV2.0 has just one component, and therefore a comparison is not plausible. The company has made some pretty outlandish claims over the years, in the tradition of A.I. labs making claims to boost hype.

Wayve.ai might be the most adaptable AI system for robo-taxis. This is because of its deep-learning approach. For instance the company claims that much like a human driver, their AV technology can learn how to drive in one city and then apply that knowledge to new places it’s never been to. While other AV technologies work only on specific mapped streets or testing areas, our system can go from driving in London to driving in another city without the need to be re-engineered.

While this may not be the most elegant description, it’s a passable metaphor for their unique approach. As argued by the Wayve team in an Arxiv publication, AVs today are designed around the same deliberative robotics architecture, which is an expansion of the sense-plan-act paradigm. The problem is broken down in a few key areas: sensing, scene representation, planning and control.

Wayve’s team believes that the majority of these are sufficiently mature for driving based on the success of respective benchmarks.  While further gains may be had, none of these areas will offer a step change to unlock scalable driving. What’s needed to achieve an autonomous future according to Wayve is decomposition: Solving driving with data.

As such I’m happy to see that Microsoft is supporting the company go mainstream as I see them as having an outside chance to scale better than other AI startups trying to solve these same problems including Google’s Waymo One, GM’s Cruise and so many others.

Wayve.AI appear to prioritize a deep learning approach with scalability. This has always been a winning approach. Wayve is committed to making our AV hardware as accessible, universal, and scalable as possible. This has led us to a camera-first approach with a low unit cost that can be used by any customer fleet, mobility service or vehicle platform. Alex Kendall was on one of the ARK Invest podcasts back in the day, it’s worth a listen here.

How Smart Robotics are Born

Wayve focuses on leveraging video collected via cameras in real-time, with radar data having a complementary role. To train its deep learning models, Wayve collects more than one terabyte of data per minute, Kendall claimed. The company has been working with Microsoft and its Azure cloud since 2020. So it’s not surprising to see in 2022 that it is deepening its partnership with Microsoft. Just as Tesla is using its own specialized supercomputer infrastructure, there seems to be a trend here starting.

If Tesla has aspirations to build smart robots with how supercomputers help boost self-driving cars, Microsoft could eventually share those ambitions so this story is more important in the future of technology than the PR at first appears. If Tesla is planning its Optimus general purpose robot, if I were Microsoft I’d be planning to do the same.

The Wayve.AI component of Azure?

Mark Russinovich, CTO, Azure at Microsoft said: “Supercomputing capabilities are key to processing the immense amount of data required for the simulation, validation, and training of AI models that enable safe and secure autonomous driving. Wayve is combining its expertise in deep learning-based autonomous vehicle systems with Microsoft Azure computing power to bring self-driving transportation experiences to more people and organizations faster.”

Alex Kendall, CEO of Wayve said: “Joining forces with Microsoft to design the supercomputing infrastructure needed to accelerate deep learning for autonomous mobility is an opportunity that we are honored to lead. Deep learning systems thrive on data, and we’ve put an immense amount of effort into understanding what it takes to get these systems on the road. We are excited by the opportunities that this collaboration will create as we push deep learning to new levels of scale.”

Wayve.AI claims that in the past few years it has seen an absolute acceleration in performance at a higher scale of training, Kendall said: more data, more compute, more parameters in  machine learning models. This, he went on to add, is really starting to push the boundaries of what is possible for any commercial cloud offering today. 

Check out Wayve.AI’s video here.

If you enjoy my articles, help support the channel by sharing it with someone you think might be interested in it as well.

Thank you for reading AI Supremacy . This post is public so feel free to share it.

Share

Founded in 2017, Wayve is on a mission to reimagine autonomous mobility through embodied intelligence. 

Wayve.AI appears to be going places now just five years later in 2022. I did not see any financial disclosures of the deal or what the partnerships entails in terms of ownership. Though the promises were lofty:

“If you think about a lot of the supercomputing technologies that are developed today, a lot of them are around large-scale text or natural language processing. But moving from kilobytes of text data to petabytes or exabytes of video data is really what’s required to make mobile robotics or autonomous driving work at scale with machine learning,” Kendall explained. ” that’s what Wayve and Microsoft are setting out to build.”

Wayve.AI’s driving intelligence is built on a Fleet Learning Platform, where it learns to drive with massive-scale driving data. As they drive, the A.I. continually improves and compound this rich data curriculum. They claim that not only does their system get smarter as it encounters more difficult driving, but it can also work across a range of use cases from grocery delivery to ride-hail to other forms of transportation.

Microsoft might want to use that approach to more products and case studies. Ultimately I think this means Microsoft should develop its own robotics lab more seriously. Time will tell if my intuition is right on this, as Microsoft has made some pretty wild bets on the Industrial Metaverse and even Gaming, it would bode well for them in the future if they did so.

According to Wayve.AI it is the first to deploy autonomous vehicles on public roads with end-to-end deep learning. Made up of a global team of experts in machine learning and robotics from top organizations around the world, the Wayve team is headquartered in London with their fleet of vehicles testing in cities across the UK. Wayve aims to be the first to deploy autonomy in 100 cities. 

Good luck to Wayve.AI and its Microsoft collaboration.

Join 69 other paying subscribers if you want to support my channel. The money goes to my basic needs (like food, rent, utility bills) as I pivot to the Creator Economy in a slow grind of material poverty.

You can also start a free-trial. I’ll give you an added week from the normal one Substack offers.

Get 14 day free trial

1
Share this post
Wayve Partners with Microsoft to Scale Autonomous Vehicles
aisupremacy.substack.com
1 Comment

Create your profile

0 subscriptions will be displayed on your profile (edit)

Skip for now

Only paid subscribers can comment on this post

Already a paid subscriber? Sign in

Check your email

For your security, we need to re-authenticate you.

Click the link we sent to , or click here to sign in.

author
Michael Spencer
May 22Author

To read Wayve AI's own announcement on this go here: https://wayve.ai/blog/wayve-to-scale-deep-learning-for-autonomous-vehicles-with-microsoft-supercomputing-technologies/

Expand full comment
ReplyCollapse
TopNewCommunity

No posts

Ready for more?

© 2022 Michael Spencer
Privacy ∙ Terms ∙ Collection notice
Publish on Substack Get the app
Substack is the home for great writing