Record-breaking wildfires that result in the loss of life and property have become the new norm in California and other parts of the world. The tools and methods that we’ve relied on to keep us safe are now struggling to keep pace with this new wildfire threat. We need a new generation of tools that can be deployed immediately to help us prepare, respond, and prevent these devastating events.

Silicon Valley is home to some of the world’s greatest innovators and problem solvers who have transformed entire industries and changed the world as we know it. We have the talent and the technology to help solve this threat. With new advances in artificial intelligence, machine learning, and predictive analytics, we can change the course of the 2019 wildfire season.

Combating the wildfire threat is a complex, multidisciplinary endeavor. A coalition of diverse stakeholders–including firefighters, utilities, lawmakers, arborists, geospatial and data scientists, homeowners, and communities–will need to assemble and collaborate closely to implement an effective solution. On March 20 and 21, these stakeholders will come together for the first time in Sacramento, Calif.,at the Fire Tech Summit. As data scientists and geospatial analytics experts, we’re grateful to have been invited to help aid in this complex challenge.

The Challenge:

Wildfires are extremely complex and dynamic events. Not all causes are preventable. But for those that are, we should have zero tolerance. It would be unrealistic to think that we can eliminate all wildfires. However, we can work towards a comprehensive solution that ends all preventable wildfires, minimizes damage to communities, and eliminates wildfire-related fatalities. In a society as advanced as ours, no one should have to fear perishing in a forest fire. Of course, moving people and towns out of wildfire-risk regions is not an option; the urban-wilderness interface will continue to get larger and more populated. Instead, we must work to minimize the damage to homes and buildings by limiting wildfire growth and enhancing the wildfire-resilience of our communities.

The Solution:

There are three main components of a wildfire solution that would meet these goals.

First, we must be proactive, not just reactive. We need to invest in the tools and processes that predict and prevent ignition. We have to be forward looking and continuously update our knowledge of changing wildfire risk, inform how to better prepare and protect our communities, and stop preventable ignition events.

Next, we have to address every stage of the wildfire’s life cycle. Wildfires are dynamic, evolving events. To minimize the impact of fires, we have to think about every stage from wildfire preparedness and prevention to response and recovery.

Lastly, it must support tactical actions from statewide areas. Wildfires are so large that they transcend towns and counties; they truly require a statewide solution. Within these massive regions, a very specific, localized action must be taken in the field to save a life, protect a home, or prevent ignition. The solution must be capable of generating these granular insights from massive areas and datasets.

All wildfire preparation and response is predicated on having timely, complete, and accurate knowledge of wildfire risk and status. Our frontline firefighters and decision-makers need “wildfire information superiority” to effectively plan, prepare, respond, and recover from these devastating events. Central to any solution is a comprehensive and accurate view of the Earth’s land, features, weather, and infrastructure. This view is provided through measurements and analysis of the Earth, also known as geospatial information. Geospatial information can provide a digital view of the physical world and, paired with artificial intelligence (AI), can give all stakeholders the informational edge they need.

Geospatial AI is the application of advanced machine learning and large-scale computing to automatically analyze massive remote sensing datasets rapidly enough for stakeholders to act on it. Proven geospatial AI technologies can be used to automatically build and update real-time, high-resolution wildfire-risk maps for the entire state of California. With those maps we can help homes and communities survive wildfires by proactively identifying where to create defensible space and predict wildfire growth. Utilities can use geospatial AI to detect vegetation near power line equipment that must be trimmed to prevent contact ignition. Geospatial AI can also automatically detect ignition in real-time to give firefighters and communities more notice and provide firefighters real-time situational awareness.

This is all possible today. Geospatial information has been used for decades. 

Farmers use geospatial remote sensing to monitor the health of their crops and AI to determine what actions will maximize crop yield. Retailers use massive geospatial datasets fused with consumer buying patterns to determine the optimal location of their physical stores. Map apps fuse real-time traffic data with complex algorithms to give us dynamically updated turn-by-turn directions on our phones. Why can’t geospatial information and AI be combined to predict, prevent, and fight wildfires? It can, and is, but we have to act quickly.

We have now experienced two record-setting wildfire seasons in a row. There is a unique window of opportunity to use the lessons from the past two years and act now, before the 2019 wildfire season starts. This will require vision and commitment. Lawmakers must move quickly to help fund and deploy statewide solutions. Utilities must adopt new technology and practices. Fire and forestry experts will need to use their knowledge to inform new solutions. And technology providers must field new capabilities to do what we do best. Innovate.

 

Article originally published on LinkedIn: https://www.linkedin.com/pulse/fighting-wildfires-ai-powered-geospatial-insights-san-gunawardana/