AI & Jet Engines: MIT Challenge Tests Next-Gen Engineering

Massachusetts Institute of Technology undergraduates recently participated in the JARVIS Challenge, a competition designed to test the integration of artificial intelligence into the process of designing, building, and testing a small gas turbine aero engine. Seven teams of students, many with limited prior experience in turbomachinery, used AI tools, accessed through MIT’s Parley platform, to accelerate their work.

The challenge revealed that while AI excels at information gathering and organization, it currently struggles with nuanced design and physical understanding. Teams found AI helpful for tasks like summarizing textbooks and sourcing vendors, but limitations emerged in detailed CAD design and navigating manufacturing constraints.

Ultimately, Team 811 Crew, relying more on fundamental engineering principles, successfully built and tested a functioning engine, while Fast and Fractured, a team that heavily utilized AI, experienced a setback during testing due to rotor issues. Professor Zolti Spakovszky of the MIT Gas Turbine Laboratory emphasized that AI's role is to augment, not replace, engineering judgment. The challenge underscored the continuing importance of hands-on experience, fundamental knowledge, and strong vendor relationships in successful engineering projects.