EMG-BASED DETECTION OF PERFORMANCE FATIGUE IN MULTI-ACTIVITY MANUAL HANDLING TASKS
Authors
Armin Bonakdar1, Catherine Disselhorst-Klug2, Karla Beltran Martinez1, Ali Golabchi3, Mahdi Tavakoli4, Hossein Rouhani1,5*
1Department of Mechanical Engineering, University of Alberta, Canada
2Department of Rehabilitation & Prevention Engineering, RWTH Aachen University, Aachen, Germany
3Department of Civil and Environmental Engineering, University of Alberta, Canada
4Department of Electrical & Computer Engineering, University of Alberta, Canada
5Glenrose Rehabilitation Hospital, Alberta Health Services, Canada
*Corresponding author
Abstract
Physical fatigue plays a critical role in work-related musculoskeletal disorders, yet tracking perceived fatigue during realistic multi-activity manual handling tasks remains challenging. This study investigated how surface electromyography (sEMG)-based myoelectric manifestation of fatigue (MMF) indicators relate to perceived exertion across repetitive lifting, carrying, and lowering activities. By segmenting sEMG with joint angles from wearable inertial sensors, linear and complexity-based fatigue features were extracted and evaluated against Borg Rate of Perceived Exertion (RPE) scales. Complexity-based indicators such as mobility, fuzzy entropy, and Dimitrov’s index showed stronger correlations with perceived fatigue than traditional linear metrics. A deep learning model using these indicators achieved meaningful multi-stage fatigue classification, suggesting a practical framework for personalized fatigue monitoring in occupational settings.