Agent-based Modeling of Fire Situations in Buildings in NetLogo Using BIM

Abstract

Modern technologies enable the prediction of events, mitigation of risks, and prevention of threat. Agent modeling is one method for simulating emergency situations and human behavior during them. An identified research gap in scientific research is the lack of effective methods to convert BIM data for agent simulation in NetLogo. Thus, the purpose of the study was to develop and test a method for converting BIM data into a format supported by NetLogo and to create a simulation model depicting the course of a fire and the evacuation of people. A simulation model was created, and ten simulations were carried out, obtaining data on the number of evacuated people, casualties and the percentage of areas occupied by the fire. The results show the applicability of the proposed approach in fire safety analyses. The model can be extended to include additional elements, such as psychophysical factors of agents, properties of building materials or the effect of smoke on visibility. The research and development work carried out can be used to develop existing BIM applications or to create new plugins. Future digital twins must be able to perform such analyses, without which it is difficult to expect the ambitious goals of the circular economy to be met.