ML/AI in Intelligent Forest Fire Observer Network

This paper discusses advantages of using machine learning and artificial intelligence (ML/AI) techniques in forest fire observer. Observer network was proposed as a framework for artificial perception software system relying on sensor network deployed on monitoring locations and used for detection of phenomenon. The architecture of the system distinguishes three main parts where data is analyzed - network observer or proprioceptor who checks upon validity of system parts and data available from sensors, phenomenon observer for detection of scenario taking place in the environment that is monitored and system observer for analysis of system usefulness and identification of future improvement of the system. In each of these parts ML/AI techniques are used. This framework was implemented as vital part of intelligent forest fire monitoring and surveillance system deployed in coastal part of Croatia and some results are presented.