Fi-Ona: Fire Warning Alarm System Using Internet of Things Based on Fuzzy Logic

摘要

The system proposed in this paper comes in form of an alarm system designed to detect any fire that happen in nearby environment, using the principle of Internet of Things (IoT) and a machine learning logic called Fuzzy Logic. The IoT technology is used for sharing data between node-node of sensor in a network that connected to database from the main system. Sensors that have been used in this system is system is divided into 3 types, Termokopel-K for temperature sensor, MQ-2 for smoke sensor, and the last one is flame sensor. Each of this sensor will convert the input into analog voltage. After that, data from each sensor will be processed using Fuzzy Logic to determine the level of a fire before the already processed data will be sent into the owner the property where the fire is ignited or the firefighter officer using realtime database. The result of testing the system shown that there are 2 conditions that can triggered the alarm. The first one is when there is 2 "medium" state from 2 sensor, for example 31° degree from temperature sensor and 170 Part Per Million(PPM) from smoke sensor. The second one is when there is "extreme" state from the sensor, for example, flame sensor output is 115 ADC. The increasing number in integrated sensor will increased the accuracy and computing time for environment data while maintaning the reliability of the decision taking process.