Optimal Selection of Affordable Wearable Devices that Minimize Construction Stresses on Multiple Body Parts

Abstract

The construction industry is struggling with skilled labor shortages, compounded by health and safety risks associated with repetitive physical stresses experienced by workers. Previous studies experimented with several costly wearable devices (e.g., full-body exoskeletons) to reduce musculoskeletal stresses for construction workers. Conversely, this study focuses on low-cost wearable devices that offer practical and affordable solutions. The study proposes a novel framework comprising three phases: (1) identify high-risk body parts through extensive analysis of publicly available injury data; (2) evaluate the effectiveness of low-cost wearable devices through controlled lab experiments; and (3) develop a training-based optimization framework for wearable recommendations based on worker characteristics and support needs. A preliminary dataset, comprising 18 participants and 11 wearable devices, was used for clustering analysis to categorize the tested wearables into low, medium, and high support levels. A neural network model was then trained to estimate the required level of support based on individual worker characteristics. Lastly, an optimization model was developed and applied to recommend shoulder, back, and leg wearables that meet any worker’s support needs and budgetary constraints. The findings demonstrate the potential of the proposed framework to support scalable, cost-effective improvements in occupational safety within construction operations.