Wayve expands its footprint through OEM partnerships
Wayve secured $1.5 billion in Series D funding to scale its global autonomy platform at an $8.6 billion valuation. The company intends to grow by licensing its vehicle-agnostic AI Driver to automakers and mobility operators. This commercial strategy focuses on providing a software layer for any vehicle, anywhere, rather than building a dedicated fleet. Alex Kendall, the CEO of Wayve, stated that autonomy scales through a trusted platform that automakers and fleets can deploy globally. Nissan plans to adopt the Wayve AI Driver for its ProPILOT system starting in FY2027, following a media driving event in Tokyo.
Stellantis aims to integrate the technology into its STLA AutoDrive platform for an L2++ rollout in 2028. Wayve also works with Uber to conduct supervised robotaxi trials in 10 cities, beginning with London this year. The UK Secretary of State for Transport, Heidi Alexander, said the investment will cement the UK as a powerhouse for the next generation of transport by supporting firms like Wayve. While Tesla builds vehicles to increase the value of its own AI, Wayve seeks to spread its intelligence across the 90 million cars produced annually by various manufacturers through global partnerships.
Tesla transitions to end-to-end neural networks
Tesla’s Full Self-Driving software now utilizes end-to-end neural networks to control driving through vision alone. The v12.5.6.3 update provides end-to-end capabilities for highways, city streets, and parking lots for AI4 hardware owners. Tesla does not provide this highway capability to AI3 hardware owners, who receive only improved city driving models. This shift replaces a modular stack that combined the HydraNet for perception with a neural network and Monte-Carlo Tree Search for planning. The old architecture required manual rules to score trajectories based on collision probability and human-likeness. You might wonder if the hardware limitations of older Tesla models will eventually halt their software progress.
In 2021, Tesla’s planning used a Monte-Carlo Tree Search algorithm enhanced with a neural network to reduce expansion nodes from 44,000 to less than 300. The 2022 Occupancy Network converted image space into voxels to predict occupancy volume and flow.
| System | Deployment Model | Hardware Requirement |
|---|---|---|
| Tesla Model Y | Vertical Integration | AI4 / HW4 |
| Tesla Cybercab | Vertical Integration | AI4 / HW4 |
| Wayve AI Driver | Software Licensing | Any OEM |
| Waymo Driver | ADS-centric | Partner OEMs |
Tesla’s Cybercab functions as a dedicated robotaxi without steering wheels or pedals. The company uses its FSD data, which exceeds 5.5 billion miles, to underpin these developments. This data helps Tesla make crashes ten times rarer than human driving.
The industry splits into specialized layers
The autonomous driving industry shows a clear divergence in how companies capture profit and manage risk. Tesla manages the vehicle, the AI, the compute, and the robotaxi service. Wayve operates closer to the OS side, providing intelligence to various manufacturers. May Mobility provides autonomy as a service, leaving fleet ownership and maintenance to partners like Toyota. Waymo retains the ADS while using partners like Uber for ride-hailing.
The market splits into industrial layers where companies excel in specific areas. Waymo reported 635,867.9 miles of testing between 2015 and 2016, while Cruise submitted 590,000 miles of testing in 2022. While Waymo uses its ADS to mitigate collisions, Cruise recorded 15 collisions during its San Francisco testing period. Wayve’s approach allows it to avoid the capital intensity of owning a fleet. In San Francisco, protestors used traffic cones on the bonnets of autonomous vehicles to render them inoperable. Waymo reported 424,000 miles of autonomous testing in 2015, during which its safety drivers disengaged the system 341 times.
| Model Type | Business Focus | Scale Strategy |
|---|---|---|
| Tesla-type | Vertical Integration | Owns vehicle and service |
| Waymo-type | ADS-centric | Uses partners for fleet |
| Wayve-type | Software Licensing | Licenses to OEMs |
| May Mobility-type | Autonomy-as-a-Service | Separates fleet and ADS |
Wayve’s strategy aims to reach the vast majority of the market by avoiding the need to retrofit its own hardware onto existing vehicles produced by various manufacturers. While Tesla uses AI to maximize the value of its own products, Wayve uses AI to spread intelligence across all vehicle types. This difference in industrial structure defines the competition between the two leaders.
