Skip to main content

Grand Challenges Team Seeks to Observe Wildfires Through UAVs and Fire Imaging Technologies to Improve Understanding and Response

A University of Maryland Grand Challenges Grant supported effort led by Principal Investigator and Assistant Professor of Fire Protection Engineering Fernando Raffan-Montoya, Professor of Fire Protection Engineering Arnaud Trouvé, and Dr. Lina Castaño (AE) seeks to leverage recent advances in technology — in particular, Unmanned Aerial Vehicles (UAVs) and fire imaging technologies —to increase the scientific understanding of wildland & WUI fire behavior and provide innovative solutions to real-time incident response.

The overarching goals of the project are twofold: a) to develop an autonomous UAV to track a wildland fire in real time while providing key information to ground personnel; b) to develop a UAV-based sensor package capable of 3-D tracking of firebrands in flight. 

Student Participation

Ph.D. student Adetola Koiki joined the team last Fall to work on the 3-D tracking problem. Upon joining UMD, she was awarded a Clark Doctoral Fellowship, a program that supports first-year doctoral students conducting research in strategic and vital areas, helping them grow into tomorrow’s leaders and innovators, launching technology companies and spearheading lifesaving advances to improve society.

Since its conception, the team’s project was designed with undergraduate learning opportunities in mind. In line with this philosophy, Professor Trouvé and Assistant Professor Fernando Raffan-Montoya have co-mentored a multidisciplinary team called SUPA'HOT, comprised of undergraduate students from engineering, computer science and environmental science under the Gemstone Honors Program at UMD. This four-year program introduces teams to the world of research while advancing knowledge and exploring topics related to urgent issues in today’s society. Last year, the 14 teams in the cohort delivered an oral poster presentation summarizing their work and its impact as part of the Do Good Challenge. A panel of five judges selected three teams to receive monetary support for their research efforts, one of which was the SUPA'HOT team mentored by Professors Trouvé and Raffan-Montoya. 

Now in their final year, the students are finishing the construction and testing of a low-cost, self-contained device capable of capturing stereoscopic images of hot firebrands in flight. Processing these images can reveal 3-D information about firebrand concentrations (i.e., number of firebrands per volume), as well as their physical location and velocity. These parameters are known to affect the probability for firebrands to ignite structures in the built environment if they land on decks, roofs, etc. For this reason, modeling the transport of firebrands is a highly active research problem within the fire science community and data obtained with this sensor package can be used to improve and validate such models. Given the device’s size and cost it also has the potential for wide deployment in the vicinity of the so-called wildland-urban interface (WUI) to trigger early warnings that can alert of possible ignitions of structures.

Unmanned Aerial Vehicles and Wildfires: XPRIZE Wildfire

In the few months since starting the project, there has been significant interest by the applied research community in implementing UAV-based technology to address multiple problems associated with wildland fires.  One of these initiatives, XPRIZE Wildfire (a 4-year, $11 million competition), was recently launched with the goal of incentivizing the innovation of firefighting technologies that can end destructive wildfires in their early stages. The competition consists of a space-based track and an autonomous suppression track, the latter of which closely aligns with the goals of our Grand Challenges Project. The XPRIZE Wildfire challenge requires a scalable, cost-effective, climate-friendly solution that can autonomously (i.e., with no human intervention) detect and suppress an incipient wildfire located within a 1,000 square kilometer area. Furthermore, the challenge must be completed within a 10-minute timeframe and any decoy fires must be left untouched. 

Recognizing the obvious overlap between XPRIZE Wildfire and the current Grand Challenges project, the PIs assembled an extended interdisciplinary team of UMD faculty, including engineers, computer scientists, and atmospheric and geographical scientists. Team Crossfire, currently led by Prof. Arnaud Trouvé, Department Chair of the Department of Fire Protection Engineering and Co-PI of the Grand Challenges project, recently presented a Qualifying Technical Submission outlining a system-level version of the proposed solution. Team Crossfire successfully advanced to the next round in the competition and will share a $750,000 purse with the other 28 teams that have moved on to the technical verification stage. Team Crossfire is currently testing subsystems in the areas of detection, suppression and system integration. For more information, please visit: https://crossfire.umd.edu/

For more information about the Grand Challenges project, visit: https://research.umd.edu/wildfires-uavs

Back to Top