Optimized Machine Learning Approach for Assessing Impact of Back-Support Exoskeletons on Muscle Strain and Fatigue in Manual Material Handling

This study evaluates the efficacy of back-support exoskeletons in alleviating muscle strain and fatigue during physically demanding manual material handling tasks, addressing a critical need in occupational ergonomics. Musculoskeletal injuries due to repetitive strain are common in industries reliant on manual labor, affecting worker health and productivity. Back-support exoskeletons, as assistive devices, offer potential benefits for reducing physical demands on workers, but their real impact on muscle activation and fatigue requires further investigation. To quantify these effects, we collected electromyography (EMG) data from five participants as they performed lifting and carrying tasks, both with and without exoskeleton support. Wearable EMG sensors attached to seven muscle groups in the lower limbs and back provided detailed data on muscle activation and exertion patterns. Data from these sessions was processed using machine learning models optimized through feature extraction and parameter tuning, allowing for precise analysis of how exoskeleton use influences muscle strain. Our initial findings indicate that exoskeletons may significantly reduce muscle strain and lower fatigue levels, potentially decreasing injury risks and enhancing long-term worker well-being. This machine learning approach enhances predictive accuracy  and provides interpretable insights into the biomechanical effects of assistive technology, offering a scalable assessment tool for workplace safety. By integrating advanced data analytics with wearable sensing technology, this research contributes to human factors and ergonomics by providing actionable insights for ergonomic interventions. It supports the adoption of cost-effective, data-driven solutions in high-risk industries, promoting sustainable practices that prioritize worker safety and productivity.