German automaker BMW Group has announced a partnership with French artificial intelligence company Mistral AI to expand the use of AI technologies in crash simulation testing. The collaboration aims to improve quality and accuracy while accelerating the execution of complex engineering tasks.
This partnership marks the first step toward broadening the use of specialized AI across additional areas of vehicle development and throughout BMW Group’s value chain.
BMW conducts thousands of virtual crash simulations every week, generating enormous volumes of engineering data. Over the years, the company has built a database exceeding one petabyte of crash simulation data, containing detailed information on vehicle structures and material behavior. This extensive dataset provides the foundation for training a specialized AI model tailored for industrial applications.
The partnership combines BMW’s engineering expertise and data resources with Mistral AI’s capabilities in developing and training advanced AI models. Together, the companies aim to create specialized AI systems capable of supporting highly complex development tasks.
Marjorie Janiewicz, Chief Revenue Officer at Mistral AI, stated that industry-specific AI models can help address complex engineering challenges such as crash test simulations.
BMW is focusing on what it calls Large Industry Models (LIMs), AI systems trained using specialized engineering and simulation data derived from vehicle development and safety testing processes.
Unlike general-purpose AI systems, these models incorporate deep domain expertise directly into the model itself. They leverage industrial data, engineering knowledge, and technical environments that enable AI systems to learn directly from BMW’s development operations.
The partnership highlights the growing importance of industrial data in advancing AI technologies used for vehicle development and safety testing, reinforcing BMW’s commitment to integrating artificial intelligence into its future engineering processes.
