Industrial electric motors are the backbone of modern manufacturing systems, including semiconductor fabs, HVAC systems, robotics, self-driving electric vehicles and smart factories. Unexpected motor failures can cause major downtime, production loss, and maintenance costs. Traditional condition-based maintenance (CBM) methods rely heavily on vibration and temperature monitoring, which often detect faults only after mechanical degradation has already progressed.
Recent advances in ultra-sensitive micro search-coil magnetic sensors provide a new opportunity for early electromagnetic fault detection before major mechanical symptoms appear. By detecting weak stray magnetic flux signatures generated by rotor faults, stator winding degradation, eccentricity, torque ripple, and load anomalies, these sensors can potentially provide 3–5 weeks earlier warning compared with conventional vibration-based monitoring.
This UROP project focuses on the AI-copilot–enhanced design, simulation, and fabrication-oriented development of low-cost ultra-sensitive PCB micro search-coil magnetic sensor module for predictive maintenance of industrial electric motors. Students will learn how first-principles electromagnetics, MEMS engineering, edge AI, and multiphysics modeling can be combined to build next-generation predictive maintenance systems.
In parallel, students will learn an “AI-copilot” engineering workflow inspired by Gabriel Petersson’s project-first + recursive gap-filling + teach-back validation method. Instead of passively studying theory for months, students will rapidly build simplified working models, identify knowledge gaps, use AI tools to accelerate learning and engineering iteration, and validate understanding through quantitative analysis and technical explanations.