Abstract

The shift from Industry 4.0 to Industry 5.0 introduces challenges in integrating assistive technologies into manufacturing. Existing international regulations and normatives, lack AI-driven task allocation and fatigue management strategies. This study addressed these gaps by implementing AI-based scheduling using reinforcement learning to dynamically reassign tasks based on fatigue levels. The AI model improved the critical risk minimization by 49.21% and increased the work efficiency by 69.92% in the simulations. Additionally, a Human Digital Twin simulation evaluated prosthetic users' movement, revealing 30% higher energy expenditure and reduced mobility compared to full-abled workers. These findings emphasise the need for ergonomic redesign and adaptive scheduling to enhance workforce efficiency and prevent injuries. Without these adaptations, industrial inclusivity would remain limited. This study provides a framework for AI-powered workforce integration, bridging policy gaps and supporting Industry 5.0's human-centred approach to accessibility and safety.