BMW expands its deal with Amazon Web Services to include autonomous vehicle data

BMW says it will use Amazon Web Services (AWS) cloud software to manage data from its autonomous vehicles.

The company has a preexisting deal with AWS to handle data from its connected cars. Now, the German automaker is expanding that deal to include data from its next-generation advanced driver-assist system (ADAS), which is partly powered by Qualcomm’s Snapdragon Ride Vision platform.

BMW says AWS will help provide cloud storage from data related to “generative artificial intelligence, Internet of Things, machine learning, and storage capabilities to help accelerate the delivery of highly automated BMW vehicles.”

The AWS software will be integrated into BMW’s “Neue Klasse” platform for its future lineup of electric vehicles, due out in 2025. This platform will process roughly triple the volume of vehicle data compared to the current generation of BMW models, the company says.

Driver-assist features, like adaptive cruise control, parking assist, and partially automated highway driving, create tons of data from engineering teams to process and analyze for future updates and improvements.

For example, the platform provides the framework needed to process, catalog, and store millions of miles of real-time driving data in Amazon Simple Storage Service (Amazon S3). Engineers and data scientists can then search, identify, and visualize relevant driving scenes to develop and train models using Amazon SageMaker, AWS’s service for building, training, and deploying machine learning models in the cloud and on the edge. 

BMW’s Neue Klasse vehicles will have significantly expanded brain power over existing models. The onboard computer will be able to process 20 to 30 times the data volume of previous models — or around double the amount of data that was previously possible. This will enable a greater fusion of the vehicle’s sensors, which will help enable higher levels of autonomous driving. The company’s iDrive driver-assist feature is designed to support both Level 2 and Level 3 autonomous driving systems.

Advanced driver-assistance systems, defined as Level 2 by the Society of Automotive Engineers, include lane keeping, blind spot detection, automatic emergency braking, and adaptive cruise control. Most major automakers include some version of advanced driver assistance in their vehicles today. Level 3 refers to highly automated driving, also called conditional automation, where the driver still needs to be able to take over the vehicle upon request.