Development of optimal routing service for emergency scenarios using pgRouting and FOSS4G

Abstract

This study aims to implement a system for Emergency Routing Decision Planning (ERDP) based on Service Oriented Architecture. A Web-based system is implemented to facilitate ubiquitous dynamic routing services on up-to-date road network data. Integration of Dijkstra’s shortest path and Analytic Hierarchy Process (AHP) algorithms has facilitated improved weighted travel-time computation. Route computations are done considering situation at source, transit and destinations. The AHP is used to prioritize amongst possible destinations considering impedance factors affecting travel time. The routing algorithm is deployed as Web Processing Service (WPS) using the ZOO-Project Platform. Two scenarios for application of the ERDP Web services are demonstrated. In the first scenario of medical emergency, the ERDP computes routes between patient’s location, emergency car to hospital in proximity of accident site considering dynamic factors such as conditions of road network, the patient’s state and availability of medical facilities and expertise in the target hospital. In the second scenario of a disaster situation, the GRASS GIS r.sim.water simulation model for overland flow under excess rainfall conditions was integrated into the ERDP system as a WPS. The result of the simulation is used to automatically update the road network database and new routes are computed based on existing conditions. The system is developed using Free and Open Source Software for Geoinformatics (FOSS4G) stack and is amenable to customization to support other emergencies such as fire, debris flow and tsunami. Integration with Sensor Observation Services for automatic data updates from CCTV camera and weather stations could further improve utility as a real-time ERDP system.