Autonomous firefighting drones are undergoing testing in California and Alaska as part of efforts to combat increasingly severe wildfires. These AI-powered systems are designed to detect fires in their initial stages and deploy suppressants before they can spread.
CAL FIRE, in collaboration with FireWERX and Seneca, conducted field tests of Seneca's autonomous suppression drones on July 15. These drones utilize onboard sensors, AI, and computer vision to locate potential fires. Once identified, they can deliver aerated Class A foam onto targeted areas. According to Seneca, their full systems can deploy four to six drones, delivering between 500 to 1,000 pounds of suppression capacity per trip.
While these drones cannot match the payload of large airtankers, they are intended for early attack operations, nighttime use, and deployment in difficult terrain where ground crews may face delays. San Bernardino County Fire previously tested an earlier Seneca Argo-1 system in December 2025, describing it as semi-autonomous with thermal tracking capabilities for early wildfire response. These systems, while capable of significant autonomous flight and targeting, still operate under human supervision, with a single pilot potentially overseeing multiple aircraft.
Wildfire Competition in Alaska
In a separate development, teams competing in the $11 million XPRIZE Wildfire competition underwent field testing in Alaska during June 2026. This competition focuses on autonomous systems designed to detect and suppress early-stage fires without human intervention. Three finalist teams demonstrated their technologies, with judges evaluating speed, accuracy, and effectiveness.
German company Dryad Networks was one of the finalists, utilizing its Silvanet sensor network combined with a drone system called Silvaguard. The Silvanet sensors, powered by solar energy, monitor for combustion gases and particulates that indicate the earliest stages of a fire, communicating potential ignitions through Dryad's network. Upon detection, an observation drone can launch automatically, using infrared and optical imaging to confirm the fire. A separate suppression drone can then be dispatched, carrying up to 100 liters of fire suppressant. In a November 2025 demonstration, Dryad reported its system detected and extinguished a controlled fire in under 12 minutes autonomously.
Moving Toward Real-World Deployment
Aspen Fire Protection District in Colorado has entered into a five-year agreement with Seneca, valued at several million dollars, for a five-aircraft Seneca Strike Team and a mobile operations base. Seneca stated the system has a capacity of approximately 500 gallons of finished foam per sortie and is designed for a single pilot to oversee multiple aircraft due to its autonomy features. The system is slated for delivery in summer 2026, with Aspen firefighters set to train on the aircraft before operational integration. This contract represents a significant step from demonstration toward real-world application.
Existing AI-driven systems are also contributing to earlier fire detection. CAL FIRE collaborates with UC San Diego's ALERTCalifornia program, which uses over 1,200 cameras and sensors. Its AI system reportedly detected more than 1,200 fires in its first season, identifying incidents before the first 911 call more than 30% of the time. Additionally, Google's FireSat project, which launched its first three operational satellites on July 7, 2026, aims to use specialized satellites and AI for early wildfire spotting, with operational data expected to become available in late 2026.
The integration of early detection systems like ALERTCalifornia and FireSat with rapid response capabilities of drones could create a more robust wildfire management strategy. While these technologies offer promise for faster intervention in remote areas, their effectiveness will depend on factors such as weather, terrain, and coordination with existing firefighting resources. Their limited payload compared to traditional aircraft means they are suited for initial attack rather than large-scale suppression.