Digital ergonomics and AI-based evaluation in agricultural work systems

Résumé fourni par la source

Agriculture remains one of the most physically demanding occupations, with workers regularly exposed to repetitive movements, awkward postures, manual material handling, vibration, heat stress, and other environmental hazards that results in a high prevalence of work-related musculoskeletal disorders (WMSDs). Conventional ergonomic assessment methods, such as Rapid Upper Limb Assessment (RULA), Rapid Entire Body Assessment (REBA), and Ovako Working Posture Assessment System (OWAS), are mainly based on manual observation and are often limited by subjectivity, time consumption, and poor applicability under dynamic field conditions. Recent advances in Agriculture 4.0 have accelerated the integration of digital technologies, enabling objective, continuous, and predictive ergonomic evaluation. This review synthesizes recent developments in digital ergonomics and artificial intelligence (AI)-based ergonomic assessment for agricultural work systems. It examines the application of wearable sensors, Internet of Things (IoT) platforms, computer vision, motion capture systems, Digital Human Modeling (DHM), digital twins, robotics, and machine learning techniques for real-time monitoring of worker posture, physiological responses, and environmental exposures. The review further discusses AI-driven approaches, including machine learning, deep learning, and decision-support systems, for automated posture recognition, ergonomic risk classification, and early prediction of WMSDs. Practical applications in harvesting, planting, weeding, manual material handling, and tractor operation are highlighted, demonstrating the potential of digital technologies to enhance worker safety, improve ergonomic design, and increase operational efficiency. Current challenges, such as technical limitations, data availability, worker acceptance, economic constraints, and the lack of long-term field validation, are also critically discussed. Future research should focus on field validation with robustness, integration of multiple sensors, lightweight AI models, human-centered system design, and intelligent decision-support frameworks to enable the transition toward predictive, data-driven occupational health management in smart agriculture.