PRESS RELEASE –

University of California San Diego’s ALERTCalifornia public safety program announced a new collaboration with Microsoft’s AI for Good Lab. Research teams will work together to strengthen ALERTCalifornia’s statewide camera network’s platform resilience, use advanced Artificial Intelligence (AI) to develop new approaches to natural disaster response, and utilize AI to dive into ALERTCalifornia’s massive data archive to understand climate change impacts across the Golden State. This collaboration aims to foster innovation and present novel opportunities to use ALERTCalifornia’s camera network and AI to prepare for, respond to, and recover from natural disasters like wildfires, while gaining a better understanding of California’s diverse habitats.

“Wildfire resilience requires innovation at every level—from sensing and analytics to infrastructure and decision support,” said Neal Driscoll, ALERTCalifornia director and professor of geology and geophysics at Scripps Institution of Oceanography at UC San Diego. “By collaborating with Microsoft, we’re expanding the scientific and technical foundation of ALERTCalifornia to better equip emergency managers and communities across the state.”

The aim is to strengthen ALERTCalifornia’s resilience by ensuring that critical systems, like the ever-growing network of more than 1,200 natural hazard monitoring cameras, remain secure, scalable and operational when they are needed most. The cameras are used by emergency managers at the local, state and federal level, and are an essential tool for wildfire and natural disaster response in California. The cameras will continue to remain open source and available to the public.  Additional support from Microsoft’s Public Safety Industry Team will leverage secure cloud infrastructure, develop data pipelines, prepare for future scalability, and ensure network continuity during times of high-demand, such as during large wildfires.

The Microsoft AI for Good team will work with ALERTCalifornia’s researchers and extensive data archive to explore how AI and predictive technologies can be better applied to natural disaster response. Teams will evaluate how machine learning and emerging AI capabilities have the potential to go beyond ALERTCalifornia’s existing award-winning AI tool to help emergency managers accelerate early wildfire detection, enhance situational awareness and support faster, more informed decision-making. Preliminary efforts in collaboration with Microsoft AI For Good Lab reveal that wildfire detection is up to 10 to 30 minutes earlier than the existing platform. This early detection helps reduce response time and improve the likelihood of preventing catastrophic fires.

“Microsoft is working with ALERTCalifornia and UC San Diego, combining Azure cloud and AI with a powerful camera network to give first responders earlier, clearer situational awareness, often before the first 911 call,” said Brad Smith, Vice Chair and President of Microsoft. “That early insight can help stop small fires from becoming devastating ones and better support those protecting lives, homes, and communities.” 

The collaboration will also create new opportunities to utilize ALERTCalifornia’s massive data archive for environmental monitoring. All of ALERTCalifornia’s natural hazard camera footage is archived, making it a wealth of long-term environmental monitoring data that has yet to be thoroughly explored. These invaluable long-term data give researchers the opportunity to study, with the help of AI, live recordings of how climate change has impacted habitats and vegetation across California over seasons, years, and, in some camera locations, decades. These data will also support Microsoft AI for Good’s efforts to detect and monitor endangered species in sensitive habitats. Endangered species, like California condors, are frequently seen roosting within view of ALERTCalifornia’s cameras.

To help with both ecology research and natural hazard monitoring, the teams are developing edge computing for AI and imagery that will improve the camera transmission frame rate from 20 seconds per frame to one frame every second.  Understanding the first few minutes of ignition is critical to assess the severity of the wildfire. New edge computing will give emergency managers the opportunity to have near-real recording to assess the characteristics of the smoke, providing a better understanding of the early behavior of the fire to inform rapid response.

ALERTCalifornia initially began as a National Science Foundation-funded initiative to create internet connectivity in remote Southern California regions for hazard detection. For more information and to view the cameras, visit alertcaliforna.org