Modelling Autonomous Evacuation Navigation System (AENS) For Optimal Route Using Dijkstra's Algorithm

Evacuation wayfinding is the process of route/pathfinding or searching from one location to a safety destination. During fire evacuation in a building, the safest and shortest path wayfinding, so called as an optimal route is a big challenge faced by people who are unfamiliar with the building environment, especially when the hazards spreads inside the building. Furthermore, the development in high-rise building and complexity of floor plan layout has greatly influenced indoor wayfinding. Currently, the static signages implemented in buildings such as building floor plan system (BFPS) and “exit signage” are unable to provide information and guidance whether the exit is inaccessible or overcrowded. In addition, there is no structured or systematic way to help occupant’s navigation that can provide direction to the safe path. Moreover, fire detection system and evacuation system operate independently whereby fire alarm control panel system (FACPS) only receives the signal and thereafter, the evacuation process takes parts without any information about the hazardous fire location. Current evacuation preparedness in wayfinding using human as an agent and the disintegration of the information will lead to a safety issue. Therefore, this study has modelled a conceptual framework for “Autonomous Evacuation Navigation System” (AENS) by adapting the systems thinking (ST). The ST itself is a conceptual framework that examines, reframes the problem and finds the solution. Through ST adaptation, all subsystems were integrated and using the “Dijkstra’s algorithm” (DA) by modifying its function from shortest path algorithm to safest and shortest algorithm, to the nearest exit. DA modification was done through the restriction of the node directions and additional “pseudo code” function. The values for zone location of fire detectors, which detect any abnormalities, updated into the matrix distance table and will not be considered for the shortest path calculation, in which the distance value will be updated as “∞” or 999. Then, the modified DA has been implemented into the proposed conceptual model. The evaluation and validation have been executed through a case study using Pathfinder simulation software and experimental model as to support the hypothesis. As a result, 79.7% and 44.7% reduction of evacuation time were recorded at two different floor plans layout for unfamiliar occupants. Additionally, the hypothesis result shows a significant difference in evacuation time taken using AENS for simulation and experiment result respectively. In conclusion, AENS using DA modification has contributed to the development of the systematic ways for evacuation preparedness, thus providing a navigation mechanism to solve the problem of indoor wayfinding and specifically to lead the unfamiliar occupants to an optimal route