MIT’s mechanical engineering department now offers a course that doesn’t just teach theory; it has students design real products using AI and machine learning. Course 2.155/156 “AI and Machine Learning for Engineering Design” is led by Associate Professor Faez Ahmed and gives students tools to tackle everyday design challenges, from bike frames to city infrastructure, with an AI-assisted boost, tells MIT News.
The benefits are immediate. AI tools speed up simulations, sharpen design accuracy, cut development costs, and even help flag maintenance issues before they pop up. The course grew into one of the most popular electives, drawing students from engineering, computer science, management, and even cross-registered students from Harvard.
Students work in teams on “design contests” with starter code and live leaderboards that encourage continuous iteration, making solutions better with each pass. Projects range from predicting ground forces in runners (“markered motion capture data”) to altering the architecture of 3D printers or redesigning furniture like cat trees. These projects aren’t just classroom exercises; they often turn into award-winning papers or published research.
What’s distinctive here is how AI is placed front and center. “When you see machine learning out there, it’s often abstract,” says graduate student Ilan Moyer. “This class pulls back the curtain on how it actually serves physical design.”
MIT is no longer teaching AI as a side note. It’s embedded directly in design education, showing future engineers how to collaborate with algorithms to create better products, faster.