From 15 hours to one minute: How AI/ML is speeding up GM's development

General Motors (GM) is ushering in a "third epoch" of engineering and design, driven by artificial intelligence and machine learning. This new approach streamlines the development process by integrating diverse functions into a single, probabilistic method, fundamentally changing how GM designs and manufactures its assets. Historically, engineering progressed through empirical, iterative design, followed by virtual tools that improved specific functions but maintained a sequential workflow.

GM's embrace of AI/ML allows for a significant reduction in the time required for complex simulations. For instance, processes that previously took 15 hours can now be completed in a single minute. This acceleration enables engineers to conduct a far greater number of iterations and broader tests than ever before, dramatically speeding up the design cycle and improving efficiency.

The application of these advanced virtualization tools extends beyond early engineering analyses and the traditional domains of aerodynamics or structural design. GM is integrating these capabilities across its various business units, including motorsports, energy and batteries, defense, and even its lunar program. This widespread adoption reflects a strategic shift towards a more integrated and rapid development paradigm.

GM's engineers are now equipped with virtual environments where they can simultaneously optimize hardware and software, and inform design decisions in real-time. This concurrent optimization, combined with the ability to run thousands of design experiments, gives GM a competitive edge. The company also fosters collaboration and technology transfer between its motorsports division and production side, ensuring that the latest advancements are continuously shared and implemented across the organization.