A New Big-Data-Driven Platform for Safety Behavior Modification in Mega Railway Project Construction

With rapid development of civil engineering and building construction industry in Hong Kong and China,accidents occur frequently and worker's unsafe behavior is considered to be a leading cause.Behavior-Based Safety (BBS)is an effective approach to modify worker's unsafe behavior but pays little attentiontoworkers’ personal characteristics. Therefore, this study proposed a mobile device-basedpersonalized worker behavior modification system in mega-scaled project construction. Firstly, thepersonalized recommendation mechanism is explored with thebig-data-drivenplatform, includingcontent-based personalized recommendation andMapReduce-based collaborative filtering personalizedrecommendation. Second, the automated platform is mainly designed for frontline construction workers.Finally, the application effect of this platform is tested against unsafe behavior rate (S). A pilot study of ametro construction site (Line 3 of Wuhan Metro Railway) shows the value ofS-01 with regard to a gang oflaborers engaged in scaffolding work decreased by more than 70%. Compared with another group oflaborers not using the proposed platform (preliminary stage with limited functions), the value ofSdecreases by more than 60%. From one 6-year research (refer to next section) :80% of average-sizedaccidents were directly related to workers’ unsafe behavior.