Smart Fire Evacuation Service Based on Internet of Things Computing for Web3D

Abstract

Fire hazard is the primary cause of crowd death in public buildings. To decrease the serious disaster from fire hazard, a rational IoT-based fire evacuation system fully considering manifold aspects of smoke hazards in Multi-agent path planning has been developed and is illustrated for Web3D. In this smart evacuation system, a WebVR technique based on volume rendering on Web and smoke diffusion data have been calculated to found an accurate and realistic fire scenario for the objective of high rate Multi-agent AI evacuation, which allows Multi-agent to practice an authentic and yet non-threatening smoke environment. Moreover, a comprehensive evaluation model of smog and narcotics hazards is also created as an important safety element for Multi-agent evacuation along optimization path planned using IoT-a ACO (Internet-of-Things-based adaptive Ant Colony Optimization) algorithm, which allows Multi-agent to identify how to find the optimal path. The position of every agent comes from every people’s position using Internet of Things technologies. And the optimization path of every people will be sent to their mobile phones and environment-TVs through IoT technologies. There are two study cases of a rectangular metro station and a curve space structure metro station, which demonstrate the characteristic of accuracy, real-time and smooth interactive performance of the proposed Multi-agent evacuation system. The system is thus shown to be valuable for people who might become trapped in fire to learn the proper evacuation procedures.