Automatic biomechanical workload estimation for construction workers by computer vision and smart insoles

Abstract

Construction workers are commonly subject to ergonomic risks due to awkward working postures or lifting/carrying heavy objects. Accordingly, accurate ergonomic assessment is needed to help improve efficiency and reduce risks. However, the diverse and dynamic nature of construction activities makes it difficult to unobtrusively collect worker behavior data for analysis. To address this issue, an automatic workload approach is proposed for the first time to continuously assess worker body joints using image-based three-dimensional (3D) posture capture smart insoles, and biomechanical analysis to provide detailed and accurate assessments based on real data instead of simulation. This approach was tested in an experiment, indicating that the method was able to automatically collect data concerning the workers’ 3D posture, estimate external loads, and provide the estimated loads on key body joints with an error rate of 15%. In addition to helping prevent construction workers’ ergonomic risks, the method provides a new data collection approach that may benefit various behavior research fields related to construction safety and productivity management.