Rivian has acquired Third Dimension AI, a two-year-old startup led by a former head of simulation for Tesla’s Autopilot program that turns real-world sensor data into editable 3D environments, to speed up the training and validation of its driver-assistance and self-driving software.
The company did not disclose financial terms and announced the deal after the US market closed on Thursday.
Third Dimension’s leadership and technical staff will join Rivian’s autonomy research and engineering teams.
The startup’s 15-person team, most of it based in the UK, will be absorbed into the autonomy and AI group led by James Philbin, Senior VP of Autonomy and AI, TechCrunch reported in an exclusive interview with Philbin.
“Simulation is central to how we train and validate our autonomous driving systems,” Philbin said in the statement.
“Their technology and expertise will strengthen and accelerate autonomy across our consumer and robotaxi roadmaps,” he added.
From Video Games to Driving Data
Third Dimension was founded in 2024 by Chief Executive Officer Tolga Kart and Piotr Sokólski.
Kart was Head of Technical Program Management and Simulation for Autopilot at Tesla from April 2018 to October 2020, according to his LinkedIn profile.
That stint spanned Tesla’s April 2019 Autonomy Day, when the company unveiled its in-house self-driving computer.
After Tesla, he was Vice President of Engineering at Parallel Domain, a company that generates simulated sensor data for autonomous-vehicle developers, from May 2021 to March 2023.
He started Third Dimension in January 2024, nine months before its funding was announced, and lists himself as based in the San Francisco Bay Area.
Before moving into autonomy, Kart spent more than a decade in video games.
He was Senior Director of Development and Production at Activision’s Sledgehammer Games from January 2013 to March 2018, where he worked on Call of Duty: WWII and, according to his profile, led a 180-person multidisciplinary organization.
Earlier he spent seven years at Electronic Arts, including as Development Director on Dead Space 3 in Montreal.
Sokólski spent four years at Wayve, the London-based self-driving startup, building photorealistic neural simulators for autonomous cars, and more than three years at Google.
The company came out of stealth in October 2024 with a $7 million seed round “led by Felicis, with participation from Abstract Ventures, MVP, Soma Capital, Solari Capital, and other investors,” according to its announcement at the time.
It then described itself as “based in California,” with “locations in the UK and Turkey.”
At the time it pitched a one-click tool to generate 3D worlds for video games, film, the US military and autonomous vehicle companies. Kart said then that “3D generation is the next frontier for generative content.”
The pitch has since narrowed to robotics and autonomy. The company’s website now markets a product called SuperSim as “high-fidelity spatial generation so autonomy teams can validate, train, and deploy with confidence.”
Its homepage demo is a scene reconstructed from WayveScenes101, a public driving dataset released by Wayve.
The company’s team page lists 13 people, including Kart, Sokólski as Chief Technology Officer and Rupprecht, and displays logos of former employers including Wayve, Tesla, Google, Luma AI, Nvidia, Niantic, Amazon and Unity.
Its UK arm, Third Dimension AI Limited, was incorporated in England on April 10, 2024, with Kart as a director, according to Companies House records. Sokólski was appointed a director on September 10, 2024, and the unit’s latest accounts run to December 31, 2025.
What Rivian Is Buying
Most autonomous-driving simulators have been built on video-game rendering engines and synthetic data, which can look real but behave differently from the physical world, Kart told TechCrunch.
Third Dimension instead reconstructs scenes from real sensor recordings and then lets engineers alter them.
“These are not synthetic kind of video game styles, they’re real things that happen in the world, but now we get that extra dial of being able to move the cameras around, shift other agents, have maybe different lighting or weather conditions,” Philbin told TechCrunch.
Rivian said the platform lets it “simulate complex driving conditions without waiting for rare real-world edge cases to occur.”
Philbin said the main value lies in validation, the step that decides whether software is safe to release.
“You really want to have a good sense of this sim-to-real gap,” he told TechCrunch. “Otherwise, you’re measuring something that’s not representative of how it will actually perform in the real world.”
Rivian had done “a pretty extensive kind of evaluation of a lot of different providers” before choosing Third Dimension, he said.
The company receives more than 400,000 driving sequences a day from customer vehicles whose owners have enabled camera data sharing, according to Philbin.
Christian Rupprecht, Third Dimension’s Chief Research Officer, wrote on the company blog in August that the startup starts “from high-fidelity, multimodal reconstruction of the real world” and uses it as a foundation for “unlimited, programmable generation.”
A Competitive Process
“It was a competitive process, but we ultimately chose Rivian for four reasons,” Kart wrote on X.
He cited Rivian’s product, its ambition, and the chance to stay “100% focused on our original thesis: how do we use simulation to validate autonomy at scale?”
Kart also praised the “no nonsense, first principles team” built by Philbin, and said he had long admired Chief Executive Officer RJ Scaringe’s “grit and determination as a founder.”
He did not say who else was involved. Neither company has said whether the deal includes retention packages or earn-outs.
The deal is Rivian’s first publicly announced acquisition since it bought Iternio, the Swedish developer of the A Better Routeplanner EV navigation app, in June 2023.
“The team is excited to be able to focus on one problem,” Kart told TechCrunch, adding that serving “many, many different masters” had been hard for a small company.
The deal also gives Rivian a chance to expand the AI hub it opened in London in July 2025 to recruit engineers for its Palo Alto autonomy team, Philbin said.
A Crowded Autonomy Calendar
The acquisition lands weeks before a series of autonomy milestones.
Rivian has opened applications for an invite-only early access program for Point-to-Point, a supervised feature that drives the full route from origin to destination, starting with owners in California and Florida.
It will hold its second Autonomy & AI Day in San Francisco on December 9, promising “brand new demos.”
Under its agreement with Uber, Uber or its fleet partners are expected to buy 10,000 fully driverless R2 robotaxis, with service planned in San Francisco and Miami from 2028, and an option to buy up to 40,000 more from 2030.
The companies plan to reach 25 cities across the US, Canada and Europe by the end of 2031.
Uber’s investment is tied to milestones, and Rivian expects a second $250 million tranche in the fourth quarter on “achieving certain technical milestones,” Vice President of Investor Relations Chip Newcom said in September.
Rivian also plans to fit LiDAR to R2 vehicles, with the first going to employees this year and to customers in 2027.
Leadership Turnover
The deal comes as Rivian’s Chief Strategy Officer, Dagan Mishoulam, leaves after nearly 11 years to join Arm. Arm designed the compute architecture behind Rivian’s in-house autonomy processor.
Rivian’s Chief Financial Officer is also preparing to leave by the end of this month.
Rivian shares closed down 3.2% at $12.06 on Thursday before the announcement and were up 0.5% at $12.12 in early premarket trading on Friday, according to Nasdaq data.
