Rivian Buys UK Startup Third Dimension for Self-Driving Work

Rivian has acquired Third Dimension, a startup whose team is mostly based in the U.K. Its technology uses AI to turn real-world data into 3D worlds. Terms of the deal were not disclosed.

Rivian is not moving into gaming. It wants to use the technology to turn real-world data into 3D scenes that speed up development of its advanced driver-assistance features and of the self-driving system for the robotaxis it plans to supply to Uber.

The 15-person team, most of it based in the U.K., will join Rivian's autonomy and AI group led by James Philbin. Philbin said in an interview that the purchase also lets Rivian expand its AI hub in London.

From game development to driving

Tolga Kart and Piotr Sokólski founded Third Dimension in 2024 and raised $7 million in seed funding from Abstract Ventures, MVP, Soma Capital and Solari Capital. The product was built to speed up video game production. The founders also saw uses in defense and autonomous vehicles, and they will now concentrate on automated driving.

Rivian currently sells "Universal Hands-Free" software, which lets drivers take their hands off the wheel on highways while they keep watching the road. By the end of the year the company plans to release a version that handles a whole trip over surface streets and highways. The driver will still have to supervise and be ready to take over. Full autonomy is planned for robotaxi versions of the R2 SUV, which are due on the Uber app in 2028.

Kart said a small company has to serve many different masters, and that the team is glad to focus on one problem at Rivian's scale.

Simulators for self-driving systems have traditionally been built on rendering or video game engines and often rely on synthetic data. Kart said a gap remains between such environments and the real world. Third Dimension's approach is to match the physical world instead.

Rivian collects aggregated, anonymized driving data from tens of thousands of R1T pickups, R1S SUVs, delivery vans and R2 SUVs whose owners have enabled external camera sharing. According to Philbin, more than 400,000 sequences arrive from customer fleets every day.

"For us, it's about taking real scenarios, but making them simulatable," Philbin said. Engineers will be able to move the cameras, shift other road users and change the lighting or weather. Rare edge cases can be simulated without waiting for them to happen on the road, which Philbin tied to validation, the step that decides whether a system is good to ship.

Source: TechCrunch