AI-enabled Drone Swarms for Fire Detection, Mapping, and Modeling

Tremendous scientific and technology advances have been made in the past decade toward detecting, mapping, and predicting behavior of wildfires, mainly due to or building on the expansion in airborne and spaceborne remote sensing. These advances have left important gaps, including the ability to monitor a wildfire's minute by minute change and conditions in its near environment, to quickly yet accurately anticipate near-term dynamic fire behavior, and to integrate diverse sources of intelligence and predictions and test action scenarios in an accessible framework. Patrolling UAVs have been proposed to fill data gaps but have had limited use as one or a few independent radio-controlled drones controlled by a nearby human, where limited bandwidth communication requires most collected information to be stored onboard for later processing. To both extend existing capabilities built on NASA data and address the limits of current UAS-based fire management technology, our objective is to develop integrated AI-based formation and onboard computing methods for a fleet of heterogeneous drones. We will accomplish this through proposal thrusts in coordinated fire detection and mapping, short term UAS-based fire spread modeling, and a synthesized data and modeling visualization environment (digital twin) to aid analysis and management of a complex, evolving wildfire and its environment. Thrust 1 will develop a hierarchical platform of multiple UAVs to provide long-term coverage of the fire using both high altitude platforms to serve as leaders, managing a fleet of small, low altitude UAS or fixed wings. The work will develop optimal coalitions to provide full fire coverage and optimally assign resources while maintaining space and communication coverage. Thrust 2 will develop low-computation real-time collaborative learning methods for fire detection and mapping onboard the drones to mosaic multi-vehicle data into a common image, distill the detected fire area, and only transmit the final fire map. Thrust 3 will develop AI-based onboard fire spread modeling using deep learning methods training from airborne data, satellite active fire data, and a library of 1-minute frequency model output from validated coupled weather-fire modeling fire event case studies. Rapid short-term predictions will anticipate alignments of fire environment conditions that enable rapid fire growth. Thrust 4 will develop a Digital Twin Environment - a 3D enhanced augmented reality environment that can integrate real-time data from aircraft and UAS assets, fire management data from SIT-209 reports, plans for fire management activities, and fire behavior modeling outputs into a semi-immersive geospatial platform so that fire management teams can spatially and interactively assess risks and opportunities across active wildland fires. Our four thrusts build upon NASA-sponsored projects, data, and target key gaps encountered in previous work across observation, modeling, and communication. They address FireTech topic areas enabling unprecedented measurements from multiple vantage points through model-directed, coordinated observations using autonomous tasking, addressing computational challenges for modeling and for data acquisition, fusion, and processing in a real-time environment, and facilitating machine learning and artificial intelligence to create new data products needed for wildfire management and for management of the constellation of observing platforms. As new discoveries and technology, they will benefit from demonstration and testing in prescribed fires, managed wildfires, and capstone field tests at FireSense airborne field campaigns.

Enhances the capabilities for existing science instruments for monitoring pre-fire, active-fire, and post-fire situations, reduces the power and mass of these instruments, and enables unprecedented observations in support of wildfire science through distributed observing systems and the information technologies needed for their support.

This is early/mid-stage (TRL 4) — the most realistic path in is NASA SBIR/STTR, which funds small businesses and research institutions to develop technology aligned with NASA's needs (equity-free, phased funding). Check whether a current SBIR/STTR solicitation topic overlaps with this project's technology area, or contact the project directly (above) to ask.

None of these are guaranteed paths for this specific project — TechPort itself doesn't have an "apply" button. Reaching out to the contact(s) above with a specific question is usually the fastest way to find out what's actually open.