Citation

Shayesteh S, Ojha A, Liu Y, Jebelli H. Safety Sci. 2023; 159: e106019.

Copyright

(Copyright © 2023, Elsevier Publishing)

DOI

10.1016/j.ssci.2022.106019

PMID

unavailable

Abstract

Occupational safety has become a major issue in the construction industry over the years. Studies have shown that work-related accidents are mostly caused by the unsafe behaviors of construction workers, implying that they can be avoided with the appropriate safety training. With emerging technologies being increasingly implemented in the construction industry, there is a growing need to improve pedagogical techniques to equip workers with the multidisciplinary skills required to safely accomplish construction tasks. Construction robots are fine examples of such technologies that necessitate more effective training, as their adoption in the field is closely tied to workers' safety. However, there is a lack of robust safety training methods for working with robots that can ensure the effectiveness of the training process. To address this issue, this study proposes a virtual avatar-based training platform for collaborating with construction robots. The proposed training platform uses workers' physiological signals to evaluate the training process throughout a human-robot collaboration (HRC) task. Using a deep neural network architecture, workers' cognitive load, a crucial factor of effective learning, was identified and linked to safety performance during the HRC task. The study results highlighted the effectiveness of the proposed training platform and its capability to evaluate the cognitive load during the HRC training in construction.

Language: en

Keywords

Cognitive load; Human-robot collaboration; Immersive environments; Physiological sensing; Safety training; Training evaluation