Line Implementation of a Mobile Application to Control Forest Fires Near Residential Areas Using an Artificial Neural Network

Abstract

The following article exposes the design and implementation of a mobile application to detect and control forest fires, developed for the iOS platform; where the Bogotá’s bordering forest areas are monitored through a series of weather stations, which capture climate data such as relative humidity, air temperature and wind speed along with their location, sending said information in real time to the user through an REST API service. The system works together with an artificial neural network (ANN) that processes the station’s meteorological data and allows determining the probability of existence of a forest fire, since in many cases the sensors can detect false alarms; improving accuracy when reporting a possible emergency, considering that the accuracy index that the artificial neural network showed is quite high (93.67%), thus reducing reaction times and alerting the control organisms in a timely manner.