The centre, which is located in Yokohama, will enable Wayve to accelerate the development of AI-powered driving software in collaboration with automakers that are looking for more flexible and safer assisted and automated driving technologies.
Wayve’s entry into Japan follows the recent announcement of its planned collaboration with Nissan, as talks continue with other international automotive groups. The expansion also comes after its launch earlier this year in Germany, with the company expanding its international footprint after last year’s $1.05 billion Series C funding round, led by SoftBank Group with participation from NVIDIA, Microsoft, and Uber.
Japan has a strong track record in the development of next-generation automotive technologies and as a global hub for vehicle production, it was seen as an ideal location for Wayve to refine and scale its AI Driver as it looks to commercialise its technology.
By incorporating training data from Japan’s complex road environments, Wayve will be able to strengthen the generalisation of its foundation model and enhance adaptability across global markets. The new Yokohama centre will support testing and development in Tokyo and surrounding regions and help to deepen collaboration with Japanese OEMs.
Commenting on the announcement Alex Kendall, Co-founder and CEO of Wayve, said, “Japan is a global leader in automotive innovation, defined by its legacy of trust and engineering excellence. By establishing our presence here, we’re embedding those principles into the development of our embodied AI. We’re here to collaborate with local partners to deliver advanced AI software that strengthens the global competitiveness of Japanese automakers and accelerates the rollout of safer and smarter mobility technology.”
Wayve’s Embodied AI software learns from experience and driving data, rather than relying on hand-coded rules, HD maps, or expensive sensors. This allows the system to adapt rapidly across markets and vehicle types.
In a recent deployment, Wayve’s end-to-end AI reached near-parity performance in the US after just a few weeks of training data, demonstrating its potential for fast, scalable rollout and in the process leapfrogging traditional development cycles.
