ShawnHymel/tinyml-example-anomaly-detection

This project is an example demonstrating how to use Python to train two different machine learning models to detect anomalies in an electric motor. The first model relies on the classic machine learning technique of Mahalanobis distance. The second model is an autoencoder neural network created with TensorFlow and Keras. Data was captured using an ESP32 and MSA301 3-axis accelerometer taped to a ceiling fan. Each sample is about 200 samples of all 3 axes captured over the course of 1 second.