Wayve Launches Dedicated Lab for Embodied AI Research
Wayve, the UK autonomous-vehicle software company, has established Wayve Labs, a new research unit focused on robotics and embodied AI. Chief scientist Jamie Shotton leads the lab, which will study how machines understand space, motion, cause and effect, and learn from real-world consequences in unpredictable environments.
The move comes after Wayve raised $1.5 billion in February from investors including Microsoft, Nvidia, Uber, Mercedes-Benz, Nissan, and Stellantis. The funding valued the company at $8.6 billion.
Dozens of Wayve employees already work at the lab. The company plans to hire additional AI researchers and machine learning engineers.
What the research covers
Wayve Labs will focus on core technical problems that affect any system learning to act in physical environments. These include sample-efficient learning from real-world data, transferring models trained in simulation to actual conditions, and reasoning about cause and effect over long sequences of actions.
Research in these areas typically requires advances in world models, hybrid training approaches that combine simulation with logged data and targeted real-world trials, and new evaluation methods centered on physical interactions rather than static perception tasks.
No immediate commercial plans
Wayve does not plan to commercialize the lab's research in the near term. The company is positioning the lab as a longer-term research effort, enabled by the scale of its recent funding.
Large strategic investments from technology and automotive partners can provide runway for non-commercial research projects that might not yield products for years.
What matters for researchers
Practitioners tracking embodied AI research should watch for several signals from Wayve Labs:
- Publication of peer-reviewed papers or release of open-source benchmarks
- Datasets and simulators made available to the research community
- Hiring for core research roles in modeling, control, and sim-to-real transfer
- Announced partnerships for field-testing research in real environments
- Tools or collaborations that connect autonomous-driving datasets with general robotics benchmarks
Whether Wayve Labs publishes its work and releases tools will determine how much impact the research has across the field. Open releases and partnerships would affect reproducibility and adoption among other teams working on similar problems.
For the broader market, more labs focused on embodied AI could accelerate progress on safety, interpretability, and evaluation methods for agents operating in physical domains.
Learn more about AI for Science & Research to deepen your understanding of research directions in this space.
Your membership also unlocks:
