The archive · Space, Robots, Defence · Technical decision · 2017–2026
Wayve's end-to-end driving bet: no maps, no rules — $1.05B, then $1.5B at $8.6B
Cambridge-born Wayve bet one self-learning neural network could drive cars without HD maps or hand-coded rules; Series C $1.05B, then $1.2B at $8.6B.
Wayve
What the business is
London-based embodied-AI company building an end-to-end autonomous driving model ('AI Driver') that automakers and mobility platforms license; the same platform extends to robotics.
Starting capital：$200M Series B (Jan 2022, Eclipse-led); $1.05B Series C (May 2024, SoftBank-led); $1.2B Series D (Feb 2026) at $8.6B post-money, with $1.5B secured in the round; $2.8B total raised across 4 rounds
How it started
Wayve was founded in Cambridge in 2017 by Alex Kendall and fellow computer-vision researchers. Their thesis: driving is a perception problem, so a single neural network trained end-to-end on real driving data could learn behaviors no engineer can hand-code and generalize without HD maps. The first tests ran on the tiny streets of Cambridge in an electric Renault Twizy.
What happened
Eclipse led a $200M Series B in January 2022, with Microsoft and Ocado (which invested $13.6M) among backers, as Wayve trained its model on delivery vehicles. In May 2024 SoftBank led a $1.05B Series C with NVIDIA and Microsoft — the UK's largest AI fundraise ever — and Wayve began pitching its mapless, hardware-agnostic 'AI Driver' to automakers and robotics companies. It started US testing in San Francisco in October 2024. In February 2026 a $1.2B Series D at an $8.6B valuation brought in Uber, Mercedes-Benz, Nissan and Stellantis, with Uber committing additional milestone capital tied to robotaxi rollouts planned to start in London in 2026.
How it ended up
Scaling toward revenue: as of February 2026 Wayve plans commercial robotaxi trials with Uber in 2026 and L2+ 'hands-off' software in consumer vehicles from 2027, having demonstrated zero-shot driving in 500+ cities in a single year.
Background
Wayve was founded in Cambridge in 2017 by computer-vision PhD Alex Kendall and colleagues on a contrarian thesis: instead of HD maps, lidar and hand-coded rules, a single end-to-end neural network could learn to drive from raw camera data, the way humans do. The first tests used an electric Renault Twizy on Cambridge's narrow streets.
The approach stayed contrarian through years of research funding: Eclipse led a $200M Series B in January 2022, with Microsoft and Ocado investing, as Wayve trained on delivery vehicles. In May 2024 SoftBank led a $1.05B Series C with NVIDIA and Microsoft — the UK's largest AI fundraise ever — backing Wayve's plan to license a mapless, hardware-agnostic 'AI Driver' to many automakers rather than build its own fleet.
By February 2026 the industry had converged on its approach: a $1.2B Series D at an $8.6B valuation brought in Uber, Mercedes-Benz, Nissan and Stellantis, with Uber committing milestone capital to deploy Wayve-powered robotaxis in 10+ markets starting with London in 2026, and L2+ 'hands-off' driving planned for consumer cars from 2027. Wayve says it became the first AV company to drive zero-shot in 500+ cities in a single year.
What has to be true
- Maps and rules do not scale: every city, road and edge case costs human engineering, while an end-to-end model trained on diverse data generalizes — attacking the bottleneck rivals were stuck on.
- Licensing to automakers instead of building a robotaxi fleet means every equipped car becomes a data source, compounding Wayve's model advantage without its own capex.
- Strategic investors (SoftBank, Microsoft, NVIDIA, then Uber and three automakers) validated the bet and brought distribution, converting a research thesis into a platform.
- Timing mattered: the industry's shift to camera-plus-GPU vehicles gave Wayve the training data and deployment path its approach needed.
What can be applied
Bet on the scaling bottleneck: rivals mapped and ruled every road; Wayve's end-to-end model generalized without city-specific engineering, turning a contrarian bet into a platform automakers license.
Aftermath
As of February 2026 Wayve is scaling toward commercial deployment: $1.5B secured in the Series D round (including Uber's milestone capital), $2.8B raised in total, and an $8.6B valuation. It plans robotaxi trials with Uber in London in 2026 expanding to 10+ markets, supervised L2+ autonomy in consumer vehicles from 2027, and continues to pitch the same embodied-AI platform to robotics companies beyond cars.
Sources
- Wayve raises $1B to take its Tesla-like technology for self-driving to many carmakers
- Wayve starts testing its buzzy end-to-end self-driving tech in the US
- Wayve secures $1.5B to deploy its global autonomy platform
- Wayve secures $1.5B to deploy its global autonomy platform
spotted an error? The archive wants to know.
Your turn
You just read one. Describe what you are building, and see who is betting on the same thing.
Free account · 3 free questions · no card