Building the CBR-based Identification System Framework for Construction Accident Precursors

Construction accident precursors(CAPs) are potential upcoming accident signals and offer the possibility for improving safety management performance. Case-based reasoning (CBR) as an artificial intelligence(AI) tools can be used to improve the efficiency and quality of CAF's identification. This paper developed a system framework for identifying CAPs on construction sites using CBR. There were ten indicators, including six problem indicators and four solution indicators were identified to describe the CAPs case between the accident characteristics and corresponding precursors. Especially, a technique case-based adaptation was implemented in the adaptation stage of CBR called inner CBR and three indicators were identified to describe the adaptation case. Eventually, the CAPs identification system framework was presented and a case study based on actual accident case from US Department of OSHA was used to illustrate how to seek CAPs. This research provides a new way to acquire more useful information from historical accident records in order to improve safety on construction sites for future ongoing project.