An international team of scientists presented a new system of drones and sensors. She is able to track even small forest fires with a very high accuracy – up to 99%.
The researchers explained that the rate at which a wildfire spreads over a large area is difficult to predict and dangerous. Their early detection is crucial for fighting fires – then the population can quickly call specialists to fight them or, if the fire is too large, the authorities will have time to evacuate nearby settlements.
Early detection of fires in remote areas is now usually done by satellite, but these devices can be hampered by excessive cloud cover. Moreover, even the most advanced satellite systems detect fires when the burning area reaches an average of 18.4 square meters. km.
For early detection of wildfires, the researchers presented a new solution that uses an Internet of Things (IoT) sensor network and a drone fleet. Researchers tested their approach with simulations, according to a study published in the IEEE Internet of Things Journal. Engineers have found that they can detect fires covering an area of 2.5 square kilometers. with near-perfect accuracy. Scientists noted that this idea is timely, as climate change is leading to an increase in the number of forest fires in many regions of the world.
Researchers want to create a network of IoT sensors scattered across problem regions, such as national parks or forests located near settlements. When a fire breaks out, IoT devices located in the area detect it and wait until a patrolling drone is in range to report its measurements. If the drone receives multiple positive results from IoT devices, it notifies the nearby fire department that fire is spreading in the area.
Researchers evaluated several drones and sensors, their costs and characteristics to determine the optimal combinations. They then tested their approach with simulation, in which 18 drones patrolled 400 sq. km., 420 sensors were located on each square kilometer. The system was able to detect fires in an area of 2.5 sq. km. with an accuracy of more than 99%. For small fires, the accuracy was 70%.
Scientists suggest that such systems may find other uses. While primarily intended for forest fire detection, it can be used to measure wind speed, moisture content, and estimate temperature. She also knows how to monitor oil spills in water bodies.