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

In 2024, Sterling Anderson made a significant career shift from the world of self-driving cars to a legacy automaker, stepping into the role of chief product officer at General Motors. Having co-founded Aurora, a self-driving startup, in 2016 after his tenure at Tesla, Anderson has witnessed firsthand the evolution of engineering practices at GM. Anderson describes GM's journey into what he refers to as the 'third epoch' of engineering and design. He reflects on the origins of engineering, tracing back to when humans drew inspiration from nature, such as birds, to create functional designs. "There was a time when humans looked at birds and were like, ‘OK, those wings seem to work pretty well. Let’s go and design something that looks like them,’" he explained. The early era of engineering was characterized by a slow and empirical approach, where inventors built prototypes based on observations and iterated through trial and error. Anderson noted that this process was time-consuming, often leading to marginally functional outcomes. With the advent of powerful computers, the second age of engineering emerged. This phase saw the introduction of advanced virtual development tools that enhanced the efficiency of the design process. Anderson pointed out that technologies like computational fluid dynamics (CFD) and finite element analysis (FEA) began to aid engineers in their work, reducing reliance on physical prototypes. Despite these advancements, Anderson highlighted that the developmental process remained linear, with design teams passing responsibilities from one discipline to another, often reverting to previous teams when issues arose. Now, as GM harnesses the power of AI and machine learning, the automaker is poised to dramatically reduce development times, potentially transforming the entire landscape of automotive engineering.