Towards automated physical fatigue monitoring and prediction among construction workers using physiological signals: An on-site study

Waleed Umer, Yantao Yu, Maxwell Fordjour Antwi Afari, Shahnawaz Anwer, Arshad Jamal

Research output: Contribution to journalArticlepeer-review

Abstract

Physical fatigue has been recognized as a serious health and safety risk among construction workers. As a result, numerous studies have endeavored to monitor/predict it using physiological measures. While the results are promising, their methodologies seem inappropriate. First, many studies utilized inappropriate benchmarking methods for physical fatigue monitoring. Importantly, a few of them utilized physical exertion scales as a surrogate for physical fatigue benchmarking. Second, many of them collected data in highly structured tasks in controlled environments. To assess these potential flaws, this research monitored fourteen construction workers' fatigue onsite by gathering physiological measures and fatigue data simultaneously. The results show that while the physical exertion scale was on average moderately correlated with a valid physical fatigue scale (average correlation coefficient 0.65), correlation coefficients varied widely among workers with the lowest of 0.05 and the highest of 0.89. This variation could be attributed to numerous factors including nature of the task, pacing and breaks during work, and individual factors. This might suggest that the physical exertion scale cannot serve as a good surrogate for physical fatigue. Additionally, the results found that workers’ physiological measures were weakly correlated to fatigue than previous laboratory studies. Overall, this study contributes to the body of knowledge by highlighting the methodological issues in the previous studies related to physical fatigue monitoring using physiological measures and the need to re-evaluate the usefulness of the measures, entailing appropriate methods. More importantly, the current study has challenged the status quo for monitoring/predicting fatigue using physiological
Original languageEnglish
Article number106242
JournalSafety Science
Volume166
Early online date27 Jun 2023
DOIs
Publication statusPublished - Oct 2023

Bibliographical note

© 2023 The Author(s). Published by Elsevier Ltd. This is an open access article under the CC BY license (http://creativecommons.org/licenses/by/4.0/).

Fingerprint

Dive into the research topics of 'Towards automated physical fatigue monitoring and prediction among construction workers using physiological signals: An on-site study'. Together they form a unique fingerprint.

Cite this