Abstract:- Detection of forest fire should be fast and accurate as they may cause damageand destruction at a large scale. Recently, Amazon forest confronted adevastating forest fire which remained obscured for over 15 days. Hence resultingin huge loss of ecosystem and adversely affecting the global conditions. As thetechnology is developing, Wireless Sensor Networks (WSN) is gaining importancein recent research areas as it has shown its usefulness in warning disastersand save lives[1]. As soon as an unusual event is noticed in the networks, anevent is detected through the sensor devices placed at distributed locations.This event detection information is passed to the base stationand decision is taken. Due to the static configuration of such sensor datain WSN generally lead to false alarm generation [2]. In such a scenario we canuse machine learning algorithms to prevent false alarm since they getconfigured efficiently in dynamic nature, that too automatically .Therefore foreliminating the static essence of WSN, we present a machine learning algorithmimbibed with WSN. In this paper, we propose a decision tree machine learningapproach for detecting events.
Keywords-- PICMicrocontroller, Speed, Distance, L293D Motor Driver, Ultrasonic sensor, LEDDisplay, Buzzer.