A Novel Fire Rescue System with Fog Computing

Fire as a common disaster, because of its perniciousness, which has aroused a lot of attention of public. Fire rescue is one of the most urgent tasks that we need to take care. Normally, firefighters make the rescue plan by the 2D map, but things cannot be predicted well without enough information. And because of the interference of fire, the transmission of data can be easily blocked. In order to play the full potential of the IoT design, and consider all the factors that may influence the indoor-environment to help make the fire rescue. First, we use Raspberry Pi to work as the fog server, Arduino mega 2560 to make the sensor, combined with PC (personal computer) to make the framework of the IoT (Internet of thing) design. And here, we put the crowd and the indoor environment as the main factors. We use PSO-SVM (Particle Swarm Optimization-Support Vector Machine) to reduce the error of prediction, neural network to analysis the image to give the number of people. And according to all the result to make the plan based on a 3D view to help make the rescue. We also make some experiments to show that the algorithms can have a good performance to improve the work.