Justice, Fairness, Inclusion, and Performance.

By Scott Cameron, Fellow of the National Academy of Public Administration, former Acting Assistant Secretary of the Interior for Policy, Management and Budget, and Senior Advisor at Cornea, Inc.


Machine learning (ML) is being applied across an ever-increasing number of applications in our society. It offers the potential to improve our lives by complementing the work of our biological brains. Traditional computing has long been able to rapidly execute routine and repetitive tasks to enable human brains to work more productively. Machine learning is now increasingly able to quickly make incremental process improvements that would otherwise be tedious tasks for their biological counterparts.

The fields of human endeavor that have benefited the least from technological advances offer the greatest promise for ML applications. One such endeavor is dealing with the ever-growing challenge of managing wildland fire. Congress recently gave the USDA Forest Service and the Department of the Interior a more than $3 billion budget increase in the Bipartisan Infrastructure law last fall to improve wildfire management. In late January, the Administration committed to a 10-year and $50 billion investment in improved wildfire management. Congress and the public naturally expect results from this influx of funding. New technology can help deliver those results.

Wildfires in the United States are increasingly large, severe, destructive, prolonged, and occur over a longer part of the calendar year than historical norms. For example, at the end of December 2021, Colorado experienced its most destructive wildfire in state history, which destroyed nearly a thousand homes in the Denver suburbs. New home construction in fire-prone areas creates more opportunities for loss of life and property from wildfire. In many parts of the country, the soil is drier, and frigid temperatures along with winter snows arrive later and end sooner,  making our already modified ecosystems even more vulnerable to wildfire over an extended period each year. The fires themselves emit hundreds of millions of tons of carbon dioxide and other greenhouse gasses, further exacerbating our long-term wildland fire challenges.

Several conditions have increased the pressure on the wildland firefighting workforce. The transition from a fire season to an ever more intense fire year has increased workloads, leading to understaffing, exhaustion, and greater physical risk. Supply chain problems have led to supply shortages. The aging of the firefighting workforce and increased retirements have led to a shortage of experienced supervisors and middle managers, making errors more likely and potentially more serious.

Wildfire can shift its speed and direction in a matter of seconds, putting firefighters’ lives at risk due to unpredictable changes. The ability to rapidly recognize, process, and shift firefighting tactics in response to changing conditions is a potentially life-saving advantage. Currently, it can take hours or even days of computer processing time to re-crunch models that guide decisions about where to position firefighting resources in response to changing conditions at a single fire. By the time the process finishes, its results may no longer make a difference on the ground.

In contrast, new ML products, such as those developed by Cornea Inc., can provide real-time situational awareness and actionable insights. They enable firefighters to fight fires more effectively, allocate human resources and supplies more efficiently, and ultimately reduce human life and property risk. More effective firefighting also means less carbon dioxide and conventional pollutants emitted from wildfires. This technology can process terabytes of terrain, weather, and fuel data across multiple wildfires and immediately share the results with firefighters and the public.

Extensive fires often span multiple jurisdictions. Responding agencies may include local fire departments, Indian Tribal firefighting agencies, state firefighting agencies, federal firefighters from multiple states, and National Guard units, sometimes working together under mutual aid agreements. Interoperability in communications and data sharing is vital to ensure that all affected parties have the same current, reliable information, can coordinate their actions effectively, and keep safe as they do their jobs. Modern technology makes this possible. Cornea’s products also allow users to communicate directly within the technology platform.

The benefits of ML can address many of the intergovernmental fire community’s pain points, specifically by:

  • Providing reliable information about fires more quickly to those charged with making resource decisions, ML can effectively amplify the workforce, allowing fewer people to accomplish the same number of tasks.
  • Providing management with better information about the workforce, ML increases efficiency, reduces fatigue, and avoids costly, perhaps fatal, errors.
  • Predicting the need for supplies at any given fire and enabling more accurate resource allocation, ML stretches available supplies and increases the odds that the right amount of supplies will be in the right place at the right time. ‘Just in Time’ inventory management comes to fighting wildfires.
  • Reducing person-hours, ML also reduces firefighters’ stress, exhaustion, and exposure to dangerous situations. As a result, senior supervisors and managers may continue pursuing their passions of fighting fire rather than retiring.
  • Improving communication, ML allows all parties access to the information they need to work more cohesively in fighting fire.

Private sector companies with expertise in ML, like Cornea, have the potential to save lives and property and improve working conditions for those involved in protecting our communities and planet from the many dangers posed by wildfire.

The views expressed in this post are those of the author. They do not necessarily reflect the views of the Academy as an institution.

Leave a Reply