Automating the Wildfire Detection and Scheduling Pipeline with Maneuverable Earth Observation Satellites
For disaster response and satellite operations, this work provides an automated pipeline for wildfire monitoring, though it is a proof-of-concept with simulated experiments.
The paper proposes WildFIRE-DS, a framework automating wildfire detection and satellite scheduling, integrating CNN-based detection, Bayesian updating, and multi-satellite scheduling. Simulated experiments using real wildfire locations and satellite orbits show enhanced monitoring capabilities.
Wildfires are becoming increasingly frequent, with potentially devastating consequences, including loss of life, infrastructure destruction, and severe environmental damage. Low Earth orbit satellites equipped with onboard sensors can capture critical information relative to active wildfires and enable near real-time detection through machine learning algorithms applied to the acquired data. We propose a framework that automates the complete wildfire detection and satellite scheduling pipeline, entitled the WildFire-applicable Intelligent and Responsive Ensemble for Detection and Scheduling (WildFIRE-DS). This paper develops an algorithm to realize the vision of the WildFIRE-DS as a proof of concept, integrating three key components: wildfire detection in satellite imagery, statistical updating that incorporates data from repeated flyovers, and multi-satellite scheduling optimization. The algorithm enables wildfire detection using convolutional neural networks with sensor fusion techniques, incorporates subsequent flyover information via Bayesian statistics, and schedules a constellation of satellites using the state-of-the-art Reconfigurable Earth Observation Satellite Scheduling Problem. Simulated experiments conducted using real-world wildfire locations and the orbits of operational Earth observation satellites to demonstrate that this autonomous detection and scheduling approach effectively enhances wildfire monitoring capabilities.