MyoLearn: Using a multimodal armband sensor for vocational safety problem identification

Abstract

This project looks at using the Myo armband and machine learning techniques to detect irregularities when performing a vocational assembly task. A Lego car assembly task is performed while Myo data are collected, which include accelerometer, gyroscope, magnetometer, and EMG. Principal Component Analysis is then used to reduce dimensionality and discover if long-term deviations from the assembly task can be detected, indicating a potential health and safety risk.