A multidisciplinary approach for the detection of workplace injuries’ precrussors

AI4WorkplaceSafety project aims to automate the detection of 1) non-use of personal protective equipment and 2) non-ergonomic execution of repetitive work assignments in an industry workspace. The proposed concept assumes the development and fusion of algorithms for the human pose estimation, object detection, and classification with algorithms for the analysis of corresponding acting forces, EEG and EMG signals measured with wearables. The interdisciplinary study is expected to enable a deeper understanding of the correlation between employees’ activities and risks from injuries at repetitive work workplaces. Automatic detection of injuries precursors is expected to significantly reduce efforts and costs of both workplace safety management and the healthcare system – while the objectivity of the safety reporting will be significantly increased.

This research was supported by The science Fund of Republic Serbia #6524219, AI – AI4WorkplaceSafety.