Distributed path planning for building evacuation guidance

Abstract

This paper proposes a distributed path planning strategy to provide guidance information for efficient building evacuation in emergency situations by utilizing a cyber-physical system (CPS) with networked sensing, information sharing, and distributed computation capabilities. The goal is to reduce casualty caused by stampedes, pushing, and squeezing during evacuation in pubic buildings when guiding evacuees from hazard zones to safe areas. By following the guidance information, evacuees are expected to transit to the exits through safe and time-efficient paths instead of rushing to the nearest exit and generating congestion along the paths and around the exits. The evacuation and hazard spreading models are built based on floor plans of an object building. The cost of each path is determined by time consumption to transit along the relative path and adjusted in real-time according to the hazard spreading model and decisions from each evacuee group. An integrated Bellman-Ford and dual subgradient algorithm is developed to find the minimum time evacuation path for scattered evacuees in a distributed manner. Simulation examples of designing the evacuation paths for occupants in Aerospace Engineering building at Iowa State University are provided to verify effectiveness of the proposed guidance strategy based on CPS.