Monitoring muscle fatigue following continuous load changes

Abstract

Human-Robot collaboration (HRC) have been actively applied in industry in a form of interactacting with workers by detecting their muscle fatigue. During repetitive movements such as dynamic muscle contractions, the information of decrement of electromyography (EMG) center frequency can suggest muscle fatigue occurred. However, the method was not evaluated both for development and recovery of muscle fatigue during such dynamic situations. If the decrement can be detected during the work without interruptions, the method can be highly useful in the industry. This study aims to confirm validity of the dynamic fatigue evaluation method using the wavelet transform in both direction of fatigue (development and recovery) by varying load intensity. Seventeen healthy males conducted four sets of repetitive elbow flexion-extension exertions work with one set of 2-minute work with a 2 kg weight (‘heavy work’) and then another with an 1 kg weight (‘light work’). Muscle fatigue was quantified by calculating the EMG center frequency from dynamic evaluation and static evaluation for each set of exertion work task. In the result, the EMG center frequency was significantly decreased during every ‘heavy work’ and increased during every ‘light work’, which was consistent with the result from dynamic fatigue evaluation based on the wavelet transform. The result suggests the possibility of monitoring muscle fatigue in real-time in industry and providing a guideline in designing a human-robot interaction system.