Abstract
The ergonomic evaluation of several production tasks is still manual filling of evaluation sheets where the results might differ even with the same dataset. To improve this manual process, motion capture (MOCAP) systems, such as intrusive (inertial sensor-based) and non-intrusive (vision-based) systems, have been intensively tested over the years for precise data collection and analysis. This study combines the strengths of two different MOCAP systems, an AI-based vision system and an inertial sensor-based system, for the human motion posture evaluation on an automotive production use case. For the experiments, a few workstations were selected and used throughout. First, an AI-based vision system with multiple algorithms and cameras was evaluated. Second, inertial sensor-based systems were evaluated. Later, an AI-based vision system and an inertial sensor-based system were combined and evaluated. The results have shown that the AI-based vision system has provided better performance with a minimal number of cameras when combined with a few inertial sensors.
Keywords: upper body postural assessment, forklift driving, depth camera, OpenPose
How to Cite:
Elango, V. & Petravic, S. & Hanson, L., (2022) “Evaluation of upper body postural assessment of forklift driving using a single depth camera”, Proceedings of the 7th International Digital Human Modeling Symposium 7(1): 38, 12 pages. doi: https://doi.org/10.17077/dhm.31780
Rights: Copyright © 2022 the author(s)
Downloads:
Download PDF
View
PDF