SoftBank leads $1bn funding for UK self-driving startup Wayve

British self-driving technology startup Wayve said on Tuesday it has raised $1.05bn (R19.46bn) in a funding round led by SoftBank Group to accelerate the development and launch of production model vehicles of its Embodied AI technology that can learn from and adapt to human behaviour.

Nvidia also contributed in the Series C funding round as a new investor, as did existing investor Microsoft.

The latest funding brings Wayve's total funds raised to just over $1.3bn (R24.10bn) and marks the largest investment in a British startup focused on artificial intelligence technology.

British Prime Minister Rishi Sunak hailed the funding round as "a testament to our leadership in this industry, and that our plan for the (UK) economy is working".

Founded in 2017, Wayve's autonomous driving technology uses AI that the startup said will enable vehicles to "navigate situations that do not follow strict patterns or rules, such as unexpected actions by drivers, pedestrians, or environmental elements".

"This will enable carmakers and fleets to accelerate their transition from assisted to autonomous driving," Wayve CEO Alex Kendall told Reuters.

The startup's technology is integrated into six different vehicle platforms including electric vehicles such as the Jaguar I-Pace and Ford Mustang Mach-E as part of advanced driver assistance systems, Kendall said. As the self-driving technology advances, Wayve's AI will be upgraded using over-the-air software updates.

The problem faced by robotaxi startups and other self-driving companies is that developing vehicles that can truly drive themselves has proven more difficult than originally imagined.

Among the main challenges is that self-driving software systems have lacked humans' ability to predict and assess risk quickly, specially when encountering unexpected incidents or "edge cases".

As the scale of that challenge has become clear, major investments in autonomous startups like Wayve are increasingly rare.

Wayve president Erez Dagan told Reuters the company's technology is "built to generalise its driving knowledge from one scenario to another because it is nearly impossible to imagine every situation a self-driving car needs to reliably handle.

"By leveraging the raw power of AI, we can build an embodied AI system that's learned from real-world and synthetic data how to handle edge cases at a rate that surpasses human programming."