AI As (Wild)fire Support

This week, the National Wildfire Coordinating Group formally welcomed the Department of Defense as a primary member of its executive board. The group “provides national leadership to enable interoperable wildland fire operations among federal, state, local, Tribal and territorial partners.”

The capabilities already exist for greater situational awareness to fight wildfires smarter, faster, and safer – to reduce staging and response times in a fire’s crucial early hours. What’s needed is mass adoption and deployment – and the mindset that AI is now as essential a tool as the layflat fire hose.

Wildfire statistics from the Congressional Research Service paint a grim picture: from 2012-21, an average of more than 61,000 wildfires impacted more than 7.4 million acres per year. Last year, nearly 59,000 wildfires burned 7.1 million acres – a devastating amount, but still less than that recorded in just the first ten months of 2022.

Mapping a wildfire traditionally can take hours. After determining the size, location, high-heat signatures, shape, that data must be synchronized with the type of terrain, possible accelerants, and wind and weather conditions to provide actionable intelligence for the fire command and control team to make educated decisions on their strategy. Often, by the time the data is relayed, the fire has shifted, and people are forced to make decisions with outdated information.

In short, the process is archaic and arduous; it that takes too much time when time is often the factor between life and death – and billions of dollars in structural damage.

Data from satellites, software-enabled AI can predict the location, speed, and direction of fire-prone conditions. An AI algorithm can then confirm fires once they exist, where it will spread, and what fuel or accelerants are in its pathway. It can relay that information in minutes. That allows fire commanders to immediately start moving assets into place with a rapid mobilization.

The tools exist to combat wildfires more efficiently and effectively. It’s time for forestry services as well as state and federal firefighting agencies to leverage them.

For example, in 2021, NVIDIA and Lockheed Martin partnered with the U.S. Forest Service and Colorado Division of Fire Prevention & Control to stand up the world’s first AI lab for predicting and responding to wildfires. The companies leverage “digital twin” technology to assesses a fire’s magnitude and environment. An AI application combines this information with data on topography, wind, and other factors to predict the fire’s spread. With real-time simulations, responders can engage fires faster and more effectively.

Leveraging capabilities like these, fire commanders can communicate to operations teams faster, enabling quicker, smarter on-the-ground decisions in deploying assets – like fixed-wing or rotor aircraft. AI can also be used to predict weather patterns through partnerships with satellites that help predict fire conditions (vegetation types, heat patterns, and dry conditions), sometimes even before the fire even starts.

To effectively reform firefighting operations proven military skills, techniques and technology can and should be used in the civilian sector for faster and more informed decision-making. Tackle the fire the same way the military tackles combat operations against the enemy. When using joint all-domain operations, multiple domains are integrated to provide the most accurate and current data available. Operators and commanders can utilize joint assets – AI, aircraft, satellites – in a fully integrated system to make a decision and act in a multifaceted operation and fast-changing environment.

As a former U.S. Army helicopter pilot, one of the earliest lessons you learn in combat is that human beings have a limitless ability to surprise you. But while people are inherently unpredictable, fire is not. As a physical phenomenon, even the largest blazes behave in knowable ways. Given a set of conditions – terrain, heat index, wind direction, etc. – wildfire data can be analyzed to shorten response times, saving lives, homes, and businesses.

As General Eisenhower said, “Plans are nothing; planning is everything.” With AI, wildfire first responders have an incredibly powerful tool for planning and responding. We should use it.

This is not to argue that natural disasters are not enormously complex and dynamic, or that their rapidly changes can’t catch even the most veteran responders by surprise. It is to argue that we should deploy every available tool to reduce the number and frequency of those surprises.

AI is at the top of that list.

Amber Smith is the bestselling author of Danger Close, former Deputy Assistant to the Secretary of Defense and U.S. Army Kiowa Warrior helicopter in Iraq and Afghanistan. Visit her at OfficialAmberSmith.com and follow on Twitter.