From Nvidia:

DigitalPath is using computer vision and thousands of cameras to detect signs of fire in real time as part of the ALERTCalifornia initiative. The company processes 8-16 million images per day to identify signs of fire and send alerts to CAL FIRE. They are also exploring the use of AI to improve wildfire prediction and detection, and utilizing high-resolution lidar data to identify areas for prescribed burns. AI algorithms classify images and determine whether they should be reviewed or acted upon.

In addition to using computer vision, the company and UCSD are exploring the use of AI to identify areas for prescribed burns in order to help prevent larger wildfires in the long term.
Through the use of multiple AI models, DigitalPath aims to improve wildfire prediction and detection, though it acknowledges that AI is not perfect. They are also considering tapping generative AI to accelerate new simulation tools and analyze the output of other models for further improvements.
The company uses a network of thousands of cameras to process millions of images daily and detect signs of fire or smoke. The images are classified to determine whether they should be reviewed or acted upon, and alerts are sent to a CAL FIRE command center as needed. This process helps in early detection and response to wildfires in California.



Read more: From Embers to Algorithms: How DigitalPath’s AI is Revolutionizing Wildfire Detection