Anomaly Detection Based on Tiny Machine Learning: A Review

Abstract

Anomaly detection (AD) is the detection of patterns in data under expected behavior. In an industrial environment, any equipment or system that breaks down will affect productivity. Therefore, Tiny Machine Learning (TinyML) is introduced to address this problem. TinyML can undergo anomaly detection to detect if any equipment did not act within expected behavior and notify the user if an anomaly detection has been detected. Anomaly detection is an unsupervised learning algorithm. It aims to identify the patterns in data that do not follow the expected behavior. By using TensorFlow Lite Micro, the TinyML can be trained to undergo anomaly detection. However, the machine learning algorithm had to be exported from TensorFlow, then TensorFlow Lite, and finally TensorFlow Lite Micro in order to upload the machine learning algorithm into TinyML. This paper highlights the state of the art of the current works on TinyML. Some suggestions on the research direction are also introduced for potential future endeavors.