DigitalPath's Ethan Higgins On Using AI to Fight Wildfires - Ep. 211

Data Filtering

DigitalPath employs advanced data filtering techniques to manage the vast amounts of data collected from their network of cameras. Ethan Higgins explains that while computer vision is crucial for identifying smoke, the real challenge lies in processing the immense volume of data efficiently 1. The cameras, constantly scanning high-risk fire areas, generate millions of images daily, necessitating a system that prioritizes images for human review 2.

The question was how can we make it so that a human can review all of these images in an efficient manner?

--- Ethan Higgins

By leveraging computer vision and automation, DigitalPath ensures that only the most relevant data reaches human operators, streamlining the detection process.

   

Alert Optimization

To tackle the overwhelming number of alerts generated, DigitalPath has developed sophisticated alert optimization strategies. Ethan Higgins describes how their system reduces the number of alerts from thousands to a manageable level by consolidating multiple images of the same fire into a single alert 3. This process involves using algorithms similar to generative adversarial networks to classify and prioritize alerts for human action.

We needed to send out a single alert per fire, and that's it. No more, no less.

--- Ethan Higgins

By refining this system, DigitalPath significantly reduces the cognitive load on human operators, allowing them to focus on critical incidents.