Development of a multi-model sensor node with data recovery ability to monitor environmental factors which could be used for forest fires detection

Abstract

Forest fires induce significant threats to ecosystems, human lives, and the climate. Human activities, fuel accumulation, and climate change are the primary factors contributing to their occurrence. To effectively manage and prevent the spreading of forest fires, it is essential to take support from improved forest fire monitoring systems. Existing fire detection systems have limitations, including high costs, limited coverage, and false alarms. Wireless Sensor Networks offer a promising solution due to their ability to cover large areas, provide rapid response, and collect relevant ambient data. This research proposes the design and development of a multi-model sensor node for forest fire detection. The sensor node integrates temperature and humidity sensors, along with a communication system and data recovery method. An artificial intelligencebased algorithm is employed for fire detection. The performance of the sensor node detection algorithm is evaluated through experiments. Also, the system life, data recovery ability, and packet loss of the sensor node are also evaluated experimentally. As the results of these evaluations, there is a system life of more than 4 years, excellent recovery ability for sensor module failure, and promising fire detection ability for a 30 cm x 30 cm wide artificial fire with 2 Kg of forest fire fuel compared to the existing systems giving better deployment ability.