Keywords: emergency information management, emergency response, sensor data cleaning, sensor data processing, sensor data storage, emergency management, real-time information, simulation, wireless sensor networks, WSNs
Sensor data processing for emergency response
A gap with current Emergency Response (ER) is the lack of real-time information on the status of an incident. Wireless sensor networks have shown great potential in addressing this gap. However, sensor data streams often contain 'noise', thus they are too difficult to be directly interpreted. In addition, sensor data are often meaningless unless it is associated with time and location information. This paper investigated the process of making sensor data work effectively for ER. The requirement analysis demonstrated that, in order to make sensor data work for ER, it needs to be properly pre-processed, it must be managed and stored efficiently, and meaning must be extracted from the data so that this can be conveyed to the emergency responders. A data cleaning approach and a data storage mechanism were proposed. Simulation results show that outliers were effectively reduced, and the query efficiency can be maintained as the database size increases.