Understanding the relationship between safety investment and safety performance of construction projects through agent-based modeling

Miaojia Lu, Clara Man Cheung, Heng Li, Shu-Chien Hsu

Research output: Journal article publicationJournal articleAcademic researchpeer-review

92 Citations (Scopus)

Abstract

The construction industry in Hong Kong increased its safety investment by 300% in the past two decades; however, its accident rate has plateaued to around 50% for one decade. Against this backdrop, researchers have found inconclusive results on the causal relationship between safety investment and safety performance. Using agent-based modeling, this study takes an unconventional bottom-up approach to study safety performance on a construction site as an outcome of a complex system defined by interactions among a worksite, individual construction workers, and different safety investments. Instead of focusing on finding the absolute relationship between safety investment and safety performance, this study contributes to providing a practical framework to investigate how different safety investments interacting with different parameters such as human and environmental factors could affect safety performance. As a result, we could identify cost-effective safety investments under different construction scenarios for delivering optimal safety performance.
Original languageEnglish
Pages (from-to)8-17
Number of pages10
JournalAccident Analysis and Prevention
Volume94
DOIs
Publication statusPublished - 1 Sept 2016

Keywords

  • Agent-based modeling
  • Construction industry
  • Information technology
  • Safety investment
  • Safety performance

ASJC Scopus subject areas

  • Human Factors and Ergonomics
  • Safety, Risk, Reliability and Quality
  • Public Health, Environmental and Occupational Health
  • Law

Fingerprint

Dive into the research topics of 'Understanding the relationship between safety investment and safety performance of construction projects through agent-based modeling'. Together they form a unique fingerprint.

Cite this