Lake County News, California – PG&E Testing Artificial Intelligence for Growing Network of High Definition Cameras

LAKE COUNTY, Calif .– Pacific Gas and Electric said it is testing artificial intelligence and machine learning capabilities in its growing network of high-definition cameras in northern and central California to see how it can improve performance. fire surveillance and response capabilities.

The company said this year, working with ALERTWildfire, it installed 138 new HD cameras in high fire risk neighborhoods, in line with its 2021 wildfire mitigation plan.

Of these 138 cameras, 46 of them are included in the new artificial intelligence, or AI, testing program in partnership with Alchera and ALERTWildfire.

A similar pilot was carried out with Pano as part of participation in EPRI’s 2021 Incubatenergy Labs Challenge. PG&E began installing HD cameras in 2018, as part of its community forest fire safety program.

As of October 31, 487 cameras were now in service, including 45 in Lake, Mendocino and Humboldt counties. None of them are in the new AI testing program.

In Lake County, 10 cameras have been installed. In 2019, cameras were installed on Mount Konocti, two each on Buckingham Peak, Cobb Mountain and Cow Mountain, with the last two cameras installed on Walker Ridge in January.

“Even with the two major rainstorms in October and November, we are still in a historic drought and California, along with other western states, continues to experience an increased risk of wildfires and an increase in the risk of wildfires. longer forest fire season. We are using all the new tools and technologies at our disposal to improve situational awareness and intelligence to mitigate and prevent wildfires, including this new AI capability, ”said Sumeet Singh, Director PG&E risks. “Every bit of data and intelligence that reaches us could potentially save a life.”

PG&E said the pilot program is already demonstrating the potential of AI to reduce the size expansion of fires.

On August 4, PG&E’s Howell Mountain 1 camera located in Placer County and equipped with Alchera’s AI software, detected smoke one minute before the actual fire started and several minutes earlier than manual movement of the camera. This smoke eventually became the Fire River. This is just one example among many noted over the course of two pilots confirming the value of early fire detection technology.

Expert staff from the company’s Wildfire Safety Operations Center, or WSOC, external agencies and first responders use fire surveillance cameras to monitor, detect, assess threats, and respond to wildfires.

AI testing programs include PG&E determining a way to get new data to the right people quickly and efficiently. The faster data is received, the faster first responders and PG&E can confirm fires and move the right resources to the right place.

“The software analyzes the video stream and if it thinks it sees smoke, we get an email and SMS alert telling us that it has just detected smoke. Our analysts then identify where the smoke is coming from and determine if it is a car fire, dumpster, or even a wildfire. Depending on the location, we can assess the threat to the public or the PG&E facilities, ”said Eric Sutphin, WSOC PG&E supervisor who is in charge of the camera installations. “AI filters out a significant number of false positives, for example, excluding dust, fog or haze.”

Sutphin explained that the recent installation of AI testing software with its machine learning capabilities means the WSOC team is getting smarter over time with more experience and more data collected.

“We know that cameras are good at detecting billows of smoke from long distances. We plan to evaluate our initial implementation, continue to collect data and develop a plan to use this advanced technology on a more extensive basis, ”he said.

The cameras offer 360 degree views with pan, tilt and zoom capabilities and can be viewed by anyone through the ALERTWildfire network at www.alertwildfire.org.

By the end of 2022, the company plans to install around 600 cameras, making it possible to see in real time more than 90% of the high-risk areas it serves.