Evacuation Management in Mixed Industrial Buildings Through Dynamic Human Flow Modeling

Ensuring safety in complex industrial environments requires a transition from static evacuation plans to dynamic models. This paper aims to optimize evacuation flows within a mixed-use production facility, employing an advanced digital approach to identify zones of vulnerability. The methodology combines architectural reconstruction in AutoCAD with computer simulations of fire scenarios in Pyrosim and evacuation flows in Pathfinder. The computational model reproduces the complexity of human behavior by integrating varied reaction times, adapting movement speeds based on the occupants’ demographic profile, and modeling their tendency to use familiar access routes rather than the nearest emergency exits. The results demonstrate that implementing a phased evacuation strategy, through the temporal staggering of alarms across levels, eliminates critical congestion phenomena. By prioritizing the evacuation of the ground floor and intermediate levels, followed by a controlled delay for the upper floors, vertical flow fluidization and a significant reduction in total egress time are achieved. The study thus validates that the strategic planning of alarm sequences can replace costly structural modifications, providing an applied solution for eliminating bottlenecks and optimizing the evacuation process for occupants in mixed industrial structures.