Design of an Indoor Disaster Routing Protocol

Background/Objectives: We suffer a lot of casualties every year from some disaster situations, such as fire or gas leaks that occur indoors. The research is needed to minimize the harm of disaster.

Methods/Statistical analysis: When the disaster happens, the evacuation route changes every minute due to obstacles or dangerous situations. To solve this problem, we must collect map data and sensor information periodically. Then this information is used to calculate the evacuation route. In addition, the sensor data of the node transfer to a server in order to utilize the prediction of evacuation path using the RNN.

Findings: In many developed countries, a lot of research are launched to prevent disasters rather than response when the disaster occurs. However, guiding evacuation routes in real situations is an aspect of the disaster response, which requires disaster sensor data. In addition, in a disaster situation, some areas cannot go through by people due to fire or gas leakage. In order to efficiently guide the evacuation route, techniques for producing a map in real time and sensor (actual and predicted) data are required. At this time, the map data is converted into the topology map by using the image obtained through the SLAM technology, and the sensor data predicts the next value based on the previous data through the RNN (using the sensor data to predict). Consequently, the proposed Routing Protocol in Indoor Disaster (RPID) guides the evacuator to a path with a short cumulative damage amount.

Improvements/Applications: Depending on the routing protocol used in the simulator or the technique of locating the user's location information, we can develop an evacuation system which can help victim find a safe way to escape as fast as possible from the dangerous situation.