Determining the Optimal Recovery Time for Fatigued Construction Workers: Machine Learning Approach Based on Physiological and Environmental Measurements

Abstract

Construction workers often engage in extended periods of intensive physical labor, resulting in sustained physiological stress. This study aimed to develop a tailored recovery time model for construction workers to determine the optimal recovery time necessary for improving their well-being. Field studies were conducted with 211 construction workers across five construction sites in mainland China, where participants performed their daily tasks until voluntary exhaustion, followed by monitored recovery on-site. A series of physiological and environmental indicators were systematically tracked to develop machine learning-based recovery time models. The developed model exhibited good fitting with high accuracy. It was found that the recovery rate of construction workers was influenced by recovery time, Wet Bulb Globe Temperature (WBGT), Air Quality Index (AQI), worker age, and worker clothing. Under conditions of an AQI of 59 and a worker age of 44 years, construction workers could achieve fatigue recovery of 68%, 56%, and 43% after a 15-min rest in the WBGT of 10–20°C, 20–30°C, and 30–40°C respectively, and 77%, 69%, and 58% after a 30-min rest in the WBGT of 10–20°C, 20–30°C, and 30–40°C respectively. Depending on the recovery process and considering the managerial expectations for recovery levels and durations, the optimal recovery time for construction workers in different environments can be determined. This study offers clear guidelines and practical recommendations for the industry to enhance the occupational health and safety of construction workers.