Integration of BIM, Bayesian Belief Network, and Ant Colony Algorithm for Assessing Fall Risk and Route Planning

摘要

The majority of routing decision approaches in emergency response operations aim at fire emergency in buildings and normally focus on building operation and maintenance phase. However, in the construction industry, falls are the most frequently occurring types of accidents resulting in fatalities, and there exists limited research considering potential fall risk on working routes where construction workers perform tasks such as material handling at the construction stage. This research contributes to providing an approach to consider potential fall risk on working routes and suggest one route with relatively lower risk by integrating BIM, Bayesian belief network and ant colony algorithm. Building information modeling can retrieve building geometry, integrate information from surrounding environment and thus help identify and evaluate potential risk at the construction stage. Based on the geometry information retrieved from the BIM model, Bayesian belief network is applied to assess potential fall risk of different identified fall scenarios. The obtained data after Bayesian belief network analysis is input into ant colony algorithm to plan safe working routes on a typical construction site. An example is presented to demonstrate the simulation result.