Situation-Aware Evacuation Model Simulated with Tehran Data Found 17 Percent Reduction in Vehicle Evacuation Time.

Microsimulation Study of A Situation-Aware Evacuation Model Reduced Route Congestion and Resulted in Reduction in Both Vehicle and Pedestrian Evacuation Time.

Date Posted
07/30/2025
Identifier
2025-B01975

A situation-aware emergency evacuation (SAEE) model using multi-agent-based simulation for crisis management after earthquake warning

Summary Information

Helping crowds evacuate safely and find shelter is crucial for protecting lives during natural disasters or other emergency situations. Sudden increase in travel demand, shifts in public behavior, and the change in the regular transport supply, would make evacuation conditions even more challenging. This study introduced an approach for quick decision-making and timely evacuation response by establishing a Situation-Aware Emergency Evacuation (SAEE). Two situations were considered: i) designing the emergency evacuation plan, and ii) finding the shortest or safest routes to reduce travel time for evacuees. The method proposed in this study was simulated in the traffic network of one district (District 6) of the 22 districts in Tehran, Iran using data from 2018 and 2021.

METHODOLOGY
This study explored automated situation identification during emergency evacuations in a foreshock scenario using a multi-agent simulation involving crisis managers, pedestrians, and vehicles. A combination of rule-based reasoning and pattern-recognition techniques was used to understand and extract evolving conditions at various levels. After determining the agent situation, movement patterns, and other spatial components, the criteria of the streets and routes to safe places were inferred, and the crisis manager provided real-time information to other agents using the situational awareness system.

FINDINGS

  • Results showed a 17 percent reduction in vehicle evacuation time.
  • For pedestrian evacuation time, results showed a 43 percent reduction.
  • Results also showed that the evacuees’ spatial knowledge and perception, and awareness of the situation of other agents and their surroundings, led to a 40 percent reduction in the complete evacuation time.
  • It was also found that in the situation-aware evacuation process, route congestion was estimated to be 30.5 percent points lower than the without awareness case.

A situation-aware emergency evacuation (SAEE) model using multi-agent-based simulation for crisis management after earthquake warning

A situation-aware emergency evacuation (SAEE) model using multi-agent-based simulation for crisis management after earthquake warning
Source Publication Date
11/23/2023
Author
Keykhaei, Mahdi; Najmeh Neysani Samany; Mohammadreza Jelokhani-Niarakia; and Sisi Zlatanova
Publisher
Geo-Spatial Information Science Journal
Goal Areas
Results Type
Deployment Locations