LONDON, May 7, 2024 — Wayve closed a $1.05 billion Series C led by SoftBank Group, with participation from new investor NVIDIA and existing investor Microsoft. The financing will support Embodied AI products for production vehicles, research infrastructure and expansion into additional markets.

The round gives Wayve capital to develop a driving system intended to learn from varied real-world environments instead of relying on a separate high-definition map and hand-written rules for every location. In the company's Series C announcement, Wayve described its approach as mapless, hardware-agnostic and designed for automakers and fleet owners.

A large round backs a different route to automation

Wayve calls its approach Embodied AI because the model learns how a physical system interacts with people, vehicles and changing surroundings. It uses driving data, simulation and evaluation to improve an end-to-end model that connects perception with driving decisions.

This differs from systems built around detailed maps and extensive location-specific rules. A more general model could make it easier to adapt software across vehicles and regions, but generalisation is a technical objective rather than a proven guarantee. Safety-critical behaviour still requires structured validation, monitoring and regulatory approval.

Investment priorities announced by Wayve

  • developing Embodied AI products intended for production vehicles;
  • scaling foundation models for assisted and automated driving;
  • advancing fleet learning, simulation, measurement and active-learning tools;
  • working with automakers and fleet owners across different hardware platforms;
  • expanding operations, partnerships, geographic data and global recruitment.

SoftBank joins the board

SoftBank led the round and joined Wayve's board as part of the transaction. NVIDIA participated as a new investor, while Microsoft added to its existing backing. The three companies bring different resources: investment capacity, automotive computing and cloud infrastructure.

Investor participation does not establish production readiness or a deployment date. It provides Wayve with financial and technical relationships while the company develops models, tools and commercial partnerships. Vehicle manufacturers remain responsible for integration, validation and compliance within each intended operating environment.

A generic unbranded crossover uses translucent perception fields while moving from a rainy city past pedestrians, a stopped bus and roadworks toward an unfamiliar coastal road
The changing road illustrates the generalisation challenge: recognising people, temporary works, traffic and lane geometry without a location-specific visual script.

The product path spans several automation levels

Wayve said its software could allow manufacturers to upgrade vehicles from L2+ driver assistance toward L4 automation as the models progress. These levels carry materially different responsibilities. Assisted driving requires an attentive driver, while higher automation depends on a defined operational domain and much stronger evidence.

The funding announcement expressed an ambition to develop and launch the first Embodied AI products for production vehicles. It did not mean that an unsupervised L4 consumer service was available in May 2024. Product claims should be tied to the capability, vehicle, geography and supervision requirements actually approved at deployment.

Road trials provide data, not automatic proof

Founded in 2017, Wayve said it had tested on UK public roads since 2018 and worked with delivery fleets operated by Asda and Ocado Group. A Financial Times demonstration before the announcement included ordinary traffic, temporary roadworks and one safety-driver intervention during a difficult merge.

Trials help expose a model to varied situations and show engineers where it fails. They are not equivalent to a statistically complete safety case. Useful assessment needs intervention rates, scenario coverage, severity measures, independent review and comparison with relevant human and automated baselines.

Data diversity is central to geographic expansion

Wayve plans to build geographically diverse data assets as it enters new markets. Road markings, vehicle types, weather, pedestrian behaviour and local rules vary significantly. A hardware-agnostic design must also account for differences in cameras, computing platforms, vehicle dynamics and maintenance.

Data collection brings governance obligations. Faces, number plates, locations and behavioural information may require protection, minimisation and controlled retention. Automakers and technology suppliers need clear responsibility for training data, software updates, incident reporting and cybersecurity.

Research concepts remain distinct from released features

Wayve cited its LINGO and GAIA research into language-responsive and generative models. The company linked this work to possible future interfaces, personalised driving and simulation. These research directions should not be read as a list of functions already included in a production vehicle.

The Series C gives the company time and computing resources to close that gap. Progress will be demonstrated through validated vehicle programmes, repeatable safety evidence and durable manufacturing partnerships rather than the size of the financing alone.

About Wayve

Wayve develops Embodied AI software and foundation models for assisted and automated driving. Its mapless, hardware-agnostic approach is intended for automakers and fleet operators across different vehicles and environments.

Media contact
Wayve
Tilly Pielichaty
Email: press@wayve.ai
230-238 York Way, London, N7 9AG, United Kingdom
Phone: +44 20 4574 5971
URL: https://wayve.ai/press/series-c/