This document summarizes a presentation on using artificial neural networks for fault diagnosis of induction motors. It includes an overview of induction motor faults, both electrical and mechanical. An experimental setup is described that uses a machinery fault simulator to introduce various seeded faults to a test motor. Data on vibration and motor current is collected across different motor speeds and fault conditions. An artificial neural network model is trained on 80% of the data and tested on the remaining 20% for fault diagnosis. The model achieves over 88% accuracy in diagnosing faults even when testing data comes from an intermediate speed not in the training data. The conclusions state that the ANN approach can successfully diagnose both mechanical and electrical induction motor faults.