The proposed work develops a cost-effective IoT and artificial-intelligence-based predictive maintenance system for electric motor fault detection. A BO geared DC motor is monitored using an ESP32 microcontroller integrated with DHT11, INA219, and MPU9250 sensors to measure temperature, current, voltage, and vibration parameters. The collected sensor data are processed using time-domain statistical features and transmitted through Wi-Fi to a local server for analysis. Machine learning models, including Decision Tree, Random Forest, and XGBoost, are employed and combined using a soft-voting approach to improve fault classification. The system categorizes the motor condition into normal, highcurrent, and maintenance-required states. In addition, vibration, current, and temperature trends are analyzed to provide early warnings before critical thresholds are reached. Blynk IoT is used for remote monitoring, while automatic email notifications are generated when abnormal conditions persist. The study demonstrates the feasibility of implementing an affordable real-time predictive maintenance solution suitable for laboratories, small industries, and academic applications.

Harshitha S, Keerthana B, Kruthika B Y, Prof. Ravi Y C, "AI-Based Predictive Maintenance System for Electric Motor Fault Detection" International Journal of Scientific Research and Engineering Development, V9(5): Page(1103-1109) September - October 2026. ISSN: 2581-7175. www.ijsred.com. Published by Scientific and Academic Research Publishing.