Optimizing Evacuation Routes: Rapid Pathfinding through Image-Based Grid Modeling and A* Algorithmic Route-Planning

Abstract

When faced with an emergency situation, humans are largely unsuccessful in identifying the shortest paths. This can become increasingly difficult when there are many walls and obstacles present. The purpose of this experiment was to create a computational software that can expeditiously help the user navigate any given environment in the case of an emergency evacuation. Using pixel-level brightness analysis, an image-to-grid conversion pipeline was developed in order to distinguish walls and obstacles from walkable space. Building interiors were converted to discrete state spaces in order to apply them to real-life architecture. The A* algorithm was then implemented to find optimal routes under the detected constraints. The user is also able to refine the positions of walls, obstacles, exits, and their starting point. The resulting evacuation path is laid on top of the original floor plan for easy interpretation. After modeling a given environment as a grid, the A* algorithm was used to compute a cost function based on the nodes of each grid cell and the distances of these nodes from the start position and the exits. This computational software is able to find the most feasible paths without requiring comprehensive architectural models of the surrounding environment. This software bridges the gap between theoretical pathfinding algorithms and implementable safety protocols, with numerous applications including evacuation planning in schools, offices, and all other public facilities.