Loading...
Loading...
GM’s new chief product officer Sterling Anderson says the automaker is entering a “third epoch” of engineering where AI/ML collapses traditional siloed design, simulation and manufacturing workflows into fast, probabilistic virtual environments. By training models to approximate heavy simulations like FEA and CFD, runs that once took ~15 hours can now complete in about one minute, enabling far more iterations and broader testing. GM applies these tools across vehicles, motorsport, energy and def
GM's shift shows AI/ML can transform engineering workflows by replacing slow physics runs with fast probabilistic models, enabling far more design iterations and faster time to market. Tech professionals should prepare for cross-disciplinary toolchains and tighter integration of ML models with CAD, simulation and manufacturing pipelines.
Dossier last updated: 2026-06-02 03:41:30
GM’s chief product officer Sterling Anderson says the automaker is entering a “third epoch” of engineering where AI/ML collapses formerly discrete design, simulation and manufacturing functions into a probabilistic, unified workflow. By replacing slow, high-cost simulations with AI-accelerated virtualization, tasks like finite element analysis that used to take ~15 hours now run in about one minute. That speed enables vastly more iterations, broader testing, and earlier use of virtual environments across GM’s businesses—motorsport, energy and batteries, defense and even its lunar program—shifting simulation from a verification step to a core design tool. The change promises faster development cycles, lower costs and tighter integration between teams.
General Motors is collapsing traditional engineering silos by using AI/ML-driven virtualization to accelerate design and testing, cutting formerly 15-hour finite element analysis (FEA) runs to about one minute, GM Chief Product Officer Sterling Anderson says. By embedding probabilistic, broadly informed models into workflows, GM can run far more iterations in parallel across aerodynamics, structures, batteries, motorsports, defense and lunar programs. The shift—from empirical prototyping to simulation to AI-augmented virtual testing—lets teams explore broader design spaces faster, reduce physical testing, and shorten development cycles, potentially lowering costs and speeding product launches in automotive and adjacent tech domains. The approach also echoes similar work from IBM and Dallara.
GM’s new chief product officer Sterling Anderson says the automaker is entering a “third epoch” of engineering where AI/ML collapses traditional siloed design, simulation and manufacturing workflows into fast, probabilistic virtual environments. By training models to approximate heavy simulations like FEA and CFD, runs that once took ~15 hours can now complete in about one minute, enabling far more iterations and broader testing. GM applies these tools across vehicles, motorsport, energy and defense — even lunar programs — shifting virtual tools from verification to active design drivers. The change shortens development cycles, reduces physical testing, and scales engineering creativity and risk exploration.
Cognition’s Scott Wu says AI coding agents shouldn’t replace humans