Abstract

The advent of Industry 5.0, with its focus on humanmachine collaboration, has amplified the complexity of designing ergonomic work environments. This presents a significant challenge for industrial engineering education, particularly in teaching ergonomic principles that account for human fatigue, operational continuity, and safety in shared workspaces. Traditional ergonomic risk assessments are often slow, subjective, and isolated from the broader production system, hindering their effectiveness and scalability in educational settings where direct exposure to real production lines is impractical. To address these limitations, this study introduces an AI-driven ergonomic automated risk assessment system designed for educational use in simulated Human-Integrated Production Lines (HIPLs). This innovative tool provides students with real-time, objective, and automated posture-based risk evaluations within virtual environments. The core of the system is an AI-based posture classification model, developed using Random Forest, XGBoost, Support Vector Machine, and a Hard Voting ensemble, achieving 94.02% classification accuracy. Integrated into a software platform, the system offers immediate feedback on ergonomic safety and productivity through a timing and monitoring module. It records ergonomic scores, risk levels, performance statistics, and visual data for traceability and personalized feedback. An experimental study involving fifteen participants performing VR-based assembly tasks demonstrated an average learning rate of 95.61%, confirming the system’s effectiveness in enhancing participants’ understanding of ergonomic risk assessment and the impact of work environment design on human performance and safety. The System Usability Scale (SUS) yielded an average score of 75.33, indicating good usability and a high level of learner engagement.