Application of Wearable Biosensors to Construction Sites. I: Assessing Workers' Stress

Houtan Jebelli, Byungjoo Choi, Sang Hyun Lee

Research output: Contribution to journalArticle

1 Citation (Scopus)

Abstract

One of the major hazards of the workplace, and in life in general, is occupational stress, which adversely affects workers' well-being, safety, and productivity. The construction industry is one of the most stressful occupations. Current stress-assessment tools rely either on a subject's perceived stress (e.g., stress questionnaires) or an individual's chemical reaction to stressors (e.g., cortisol hormone). However, these methods can interrupt ongoing tasks and therefore may not be suitable for continuous measurement. To address this problem, the authors aim to develop and validate a framework for noninvasive and nonsubjective measurement of worker stress by examining changes in workers' physiological signals collected from a wearable biosensor. The framework applies various filtering methods to reduce physiological signal noises and extracts the patterns of physiological signals as workers experience various stress levels. Then, the framework learns these patterns by applying a supervised-learning algorithm. To examine the performance of the proposed framework, the authors collected a physiological signal from 10 construction workers in the field. The proposed framework resulted in a stress-prediction accuracy of 84.48% in distinguishing between low and high stress levels and 73.28% in distinguishing among low, medium, and high stress levels. The results confirmed the potential of the proposed framework for assessing workers' stress in the field. Automatic predictions of workers' physical demand levels based on physiological signals is described in a companion paper. This study, along with the companion paper, contributes to the body of knowledge on the in-depth understanding of construction workers' stress on construction sites by developing a noninvasive means for continuous monitoring and assessing workers' stress. The proposed stress-recognition framework is expected to enhance workers' health, safety, and productivity through early detection of occupational stressors on actual sites.

Original languageEnglish (US)
Article number04019079
JournalJournal of Construction Engineering and Management
Volume145
Issue number12
DOIs
StatePublished - Dec 1 2019

Fingerprint

Biosensors
Workers
Construction sites
Productivity
Cortisol
Hormones
Supervised learning
Construction industry
Learning algorithms
Chemical reactions
Hazards
Health
Monitoring

All Science Journal Classification (ASJC) codes

  • Civil and Structural Engineering
  • Building and Construction
  • Industrial relations
  • Strategy and Management

Cite this

@article{ccfa7ef109e547689657c0738ff02fdc,
title = "Application of Wearable Biosensors to Construction Sites. I: Assessing Workers' Stress",
abstract = "One of the major hazards of the workplace, and in life in general, is occupational stress, which adversely affects workers' well-being, safety, and productivity. The construction industry is one of the most stressful occupations. Current stress-assessment tools rely either on a subject's perceived stress (e.g., stress questionnaires) or an individual's chemical reaction to stressors (e.g., cortisol hormone). However, these methods can interrupt ongoing tasks and therefore may not be suitable for continuous measurement. To address this problem, the authors aim to develop and validate a framework for noninvasive and nonsubjective measurement of worker stress by examining changes in workers' physiological signals collected from a wearable biosensor. The framework applies various filtering methods to reduce physiological signal noises and extracts the patterns of physiological signals as workers experience various stress levels. Then, the framework learns these patterns by applying a supervised-learning algorithm. To examine the performance of the proposed framework, the authors collected a physiological signal from 10 construction workers in the field. The proposed framework resulted in a stress-prediction accuracy of 84.48{\%} in distinguishing between low and high stress levels and 73.28{\%} in distinguishing among low, medium, and high stress levels. The results confirmed the potential of the proposed framework for assessing workers' stress in the field. Automatic predictions of workers' physical demand levels based on physiological signals is described in a companion paper. This study, along with the companion paper, contributes to the body of knowledge on the in-depth understanding of construction workers' stress on construction sites by developing a noninvasive means for continuous monitoring and assessing workers' stress. The proposed stress-recognition framework is expected to enhance workers' health, safety, and productivity through early detection of occupational stressors on actual sites.",
author = "Houtan Jebelli and Byungjoo Choi and Lee, {Sang Hyun}",
year = "2019",
month = "12",
day = "1",
doi = "10.1061/(ASCE)CO.1943-7862.0001729",
language = "English (US)",
volume = "145",
journal = "Journal of Construction Engineering and Management - ASCE",
issn = "0733-9364",
publisher = "American Society of Civil Engineers (ASCE)",
number = "12",

}

Application of Wearable Biosensors to Construction Sites. I : Assessing Workers' Stress. / Jebelli, Houtan; Choi, Byungjoo; Lee, Sang Hyun.

In: Journal of Construction Engineering and Management, Vol. 145, No. 12, 04019079, 01.12.2019.

Research output: Contribution to journalArticle

TY - JOUR

T1 - Application of Wearable Biosensors to Construction Sites. I

T2 - Assessing Workers' Stress

AU - Jebelli, Houtan

AU - Choi, Byungjoo

AU - Lee, Sang Hyun

PY - 2019/12/1

Y1 - 2019/12/1

N2 - One of the major hazards of the workplace, and in life in general, is occupational stress, which adversely affects workers' well-being, safety, and productivity. The construction industry is one of the most stressful occupations. Current stress-assessment tools rely either on a subject's perceived stress (e.g., stress questionnaires) or an individual's chemical reaction to stressors (e.g., cortisol hormone). However, these methods can interrupt ongoing tasks and therefore may not be suitable for continuous measurement. To address this problem, the authors aim to develop and validate a framework for noninvasive and nonsubjective measurement of worker stress by examining changes in workers' physiological signals collected from a wearable biosensor. The framework applies various filtering methods to reduce physiological signal noises and extracts the patterns of physiological signals as workers experience various stress levels. Then, the framework learns these patterns by applying a supervised-learning algorithm. To examine the performance of the proposed framework, the authors collected a physiological signal from 10 construction workers in the field. The proposed framework resulted in a stress-prediction accuracy of 84.48% in distinguishing between low and high stress levels and 73.28% in distinguishing among low, medium, and high stress levels. The results confirmed the potential of the proposed framework for assessing workers' stress in the field. Automatic predictions of workers' physical demand levels based on physiological signals is described in a companion paper. This study, along with the companion paper, contributes to the body of knowledge on the in-depth understanding of construction workers' stress on construction sites by developing a noninvasive means for continuous monitoring and assessing workers' stress. The proposed stress-recognition framework is expected to enhance workers' health, safety, and productivity through early detection of occupational stressors on actual sites.

AB - One of the major hazards of the workplace, and in life in general, is occupational stress, which adversely affects workers' well-being, safety, and productivity. The construction industry is one of the most stressful occupations. Current stress-assessment tools rely either on a subject's perceived stress (e.g., stress questionnaires) or an individual's chemical reaction to stressors (e.g., cortisol hormone). However, these methods can interrupt ongoing tasks and therefore may not be suitable for continuous measurement. To address this problem, the authors aim to develop and validate a framework for noninvasive and nonsubjective measurement of worker stress by examining changes in workers' physiological signals collected from a wearable biosensor. The framework applies various filtering methods to reduce physiological signal noises and extracts the patterns of physiological signals as workers experience various stress levels. Then, the framework learns these patterns by applying a supervised-learning algorithm. To examine the performance of the proposed framework, the authors collected a physiological signal from 10 construction workers in the field. The proposed framework resulted in a stress-prediction accuracy of 84.48% in distinguishing between low and high stress levels and 73.28% in distinguishing among low, medium, and high stress levels. The results confirmed the potential of the proposed framework for assessing workers' stress in the field. Automatic predictions of workers' physical demand levels based on physiological signals is described in a companion paper. This study, along with the companion paper, contributes to the body of knowledge on the in-depth understanding of construction workers' stress on construction sites by developing a noninvasive means for continuous monitoring and assessing workers' stress. The proposed stress-recognition framework is expected to enhance workers' health, safety, and productivity through early detection of occupational stressors on actual sites.

UR - http://www.scopus.com/inward/record.url?scp=85072940974&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=85072940974&partnerID=8YFLogxK

U2 - 10.1061/(ASCE)CO.1943-7862.0001729

DO - 10.1061/(ASCE)CO.1943-7862.0001729

M3 - Article

AN - SCOPUS:85072940974

VL - 145

JO - Journal of Construction Engineering and Management - ASCE

JF - Journal of Construction Engineering and Management - ASCE

SN - 0733-9364

IS - 12

M1 - 04019079

ER -