Rivian buys UK team to digitize 400,000 daily fleet clips

Rivian has acquired Third Dimension, a 15-person technology firm based in the United Kingdom, to enhance its autonomous driving capabilities. The purchase, first reported by TechCrunch, was not accompanied by a disclosed price tag. The move integrates a specialized team into Rivian's existing autonomy and AI division, which is currently led by James Philbin.

Third Dimension was established in 2024 by Tolga Kart, a former Tesla Autopilot simulation engineer, and Piotr Sokólski, who spent four years at Wayve developing neural simulators. The startup previously raised a $7 million seed round led by Felicis in October 2024. At that time, the company identified autonomous vehicle manufacturers, game studios, and the US military as potential clients for its generative AI tools that create 3D environments for game engines.

The acquisition leverages data from Rivian's existing vehicle fleet, including R1S, R1T, R2 models, and commercial vans. Owners who opt into camera data sharing contribute to a dataset that generates more than 400,000 driving sequences daily, according to Philbin. While the majority of this footage captures routine driving conditions, autonomous systems often struggle with rare, high-stakes events. Examples of such edge cases include a mattress blocking a fast lane, a police officer overriding a red light, or a child entering the roadway from between parked cars.

Rather than waiting months for these specific events to occur naturally in the wild, engineers at Rivian will use the acquired technology to reconstruct real-world scenes in three dimensions. This process allows developers to modify variables and test their software against modified versions of actual incidents. Philbin stated that simulation is central to the training and validation of autonomous systems, noting that Third Dimension has created technology that converts real-world data into detailed 3D environments.

The deal also expands Rivian's physical footprint in London, as the majority of Third Dimension's staff are based there. This addition to the UK workforce complements Rivian's growing presence in the region. The integration of this team signals a shift toward processing large volumes of fleet video to generate synthetic testing grounds, a strategy shared by competitors such as Waymo and Tesla. Waymo recently unveiled a world model built on Google DeepMind's Genie 3, while Tesla has demonstrated a neural world simulator for its Full Self-Driving software.

Financial milestones will drive the deployment of Rivian's self-driving vehicles. In March, Uber agreed to invest up to $1.25 billion in Rivian for the rollout of R2 robotaxis. Of this total, $300 million is paid upfront, with the remaining balance contingent on achieving specific autonomy milestones. The first major target is placing 10,000 driverless R2 vehicles on Uber's network in San Francisco and Miami by 2028.

Consumer-facing timelines have already undergone adjustments. The point-to-point driving feature is scheduled to reach a small group of invited Autonomy+ owners later this year, though testers must sign nondisclosure agreements. CEO RJ Scaringe originally promised eyes-off driving capability for 2026 in January 2025, but that target has shifted to 2027. Level 4 capability is planned for 2028 on the Gen 3 hardware revealed at the company's AI & Autonomy Day.