Local data fusion algorithm for fire detection through mobile robot

Abstract

Multisensor data fusion is a technique that combines the readings of multiple sensors to detect some phenomenon. Data fusion applications are numerous and they can be used in smart buildings, environment monitoring, industry and defense applications. The main goal of multisensor data fusion is to minimize false alarms and maximize the probability of detection based on the detection of multiple sensors. In this paper a local data fusion algorithm based on luminosity, temperature and flame for fire detection is presented. The data fusion approach was embedded in a low cost mobile robot. The prototype test validation has indicated that our approach can detect fire occurrence. Moreover, the low cost project allow the development of robots that could be discarded in their fire detection missions. © 2013 IEEE.