GM's AI-Powered Push to Halve Vehicle Development Time

For decades, automakers could afford to take their time, with some models staying on the market for over a decade before a full redesign. But the competitive landscape has shifted dramatically, largely due to China’s rapid EV development cycles, which can bring a new car from concept to showroom in two years or less. In response, General Motors is leveraging AI and simulation to dramatically accelerate its own product development.

Sterling Anderson, GM’s chief product officer and a former Tesla and Aurora Innovation executive, is spearheading this effort. Along with Jason Fischer, GM’s executive director of virtual integration engineering, Anderson is implementing a new approach that collapses the traditional, siloed design process into a single virtual environment. This allows engineers to assess how a design change affects a finished vehicle in about a minute, compared to the 15 hours it previously took.

The results are already tangible. The electric GMC Hummer went from initial design to showroom in just two years, half the typical four-to-five-year cycle. Front-end crash simulations, which once required 15 hours of computing, now take less than a minute. This speed enables engineers to explore thousands of scenarios, from snow and rain to different suspension setups, and to simulate human behavior in the cabin. GM is even applying this technology to develop tires for NASA’s lunar rover, allowing engineers in Michigan to adjust gravity in software.

This AI-driven approach is not just about speed; it’s about creating more robust and refined vehicles. By uncovering weak points early in the design phase, GM ensures that physical testing begins with a stronger product. The company aims to apply this fast-track development to all its vehicle and technology programs, fundamentally changing the way cars are designed and tested.