AWS, Aumovio expand AI-driven development of self-driving vehicles

By Abhirup Roy and Akash Sriram

LAS VEGAS, Jan 6 (Reuters) – Amazon’s cloud unit has partnered with German automotive hardware supplier Aumovio to support the commercial rollout of self-driving vehicles, starting with Aurora’s autonomous trucks, the companies said on Tuesday.

As part of an expansion of a long-running partnership, Amazon Web Services will become Aumovio’s preferred cloud provider for autonomous driving development powered by artificial intelligence tools.

The new tools are set to be used for the first time for autonomous freight company Aurora’s planned deployment of driverless trucks at scale from 2027. Its shares were up more than 8% on Tuesday.

Automakers worldwide have poured billions into AI systems that power long-awaited self-driving technologies, which have faced several technical challenges.

The collaboration reflects a broader shift in autonomous driving from research to commercial deployment, particularly in freight. Aurora has already launched limited driverless operations in the U.S.

“The big accelerant in the industry has been the use of engineering AI, because it allows development and validation with significantly fewer resources,” Ozgur Tohumcu, general manager for automotive and manufacturing at Amazon Web Services, told Reuters.

Using AWS’s cloud systems, Aumovio’s engineers can sift through vast amounts of driving data with generative and agentic AI to detect rare situations such as road debris and pedestrians in traffic lanes and speed up training and validation of autonomous systems.

“When you validate a Level 4 system, you’re trying to prove it behaves correctly in extremely rare situations that are very hard to find in the real world,” Jeremy McClain, head of system and software at Aumovio’s autonomous mobility unit, told Reuters. “Without AI, finding those edge cases in massive data sets would be very difficult.” 

Aumovio, spun off from German tire maker Continental last year, supplies the hardware platform for Aurora’s self-driving system and a separate “fallback” system designed to bring a truck safely to a stop if the primary autonomous driver fails, company executives said in an interview.

(Reporting by Akash Sriram in Bengaluru and Abhirup Roy in Las Vegas; Editing by Sahal Muhammed)