Problem Statement

Based on the statistics from the National Interagency Fire Center (NIFC), wildfires exhausted 10 million acres of land in 2016 and brought $6 billion irrecoverable damages from 1995 to 2014 in the United States. Wildfires not only impact the wildlife, but more importantly endanger human lives. Therefore, early detection of wildfires before they get out of control is an urgent requirement. Wildfires are often initiated in remote forest areas where the common fire detection methods such as lookout tower stations fail to detect such fires in a timely manner. Moreover, conventional detection approaches can barely provide sufficient fire information about the precise fire locations, the orientation of fire expansion.


We aim to build a network of drones that can autonomously detect and track wildfires. The system will utilize technology such visible light and infrared imaging, software-based radio communication, path planning and machine learning algorithms

Project Depiection Diagram


Our ultimate goal is to have a network of drones that will cover a large area of forest for a long period time and communicate with each other, to detect and relay about all the fires detected in an area. The data will be transmitted to a ground station using a custom software-defined radio system. Then the information is sent to first reponders. The goal of project is to drastically improve wild fire detection times