PG&E explains how AI, machine learning tools can bolster wildfire prevention efforts

Machine Learning


PG&E shares how it uses the latest technology to prevent wildfires.

At a media briefing on August 28th, PG&E chief meteorologist Scott Strenfell said the situation has been cool so far this fire season. However, he added that Cal Fire has a higher total number of fires in 2025 compared to this point in 2025.

In August and September, Strenfel said most of the PG&E service areas carry a fire risk beyond the usual, citing the National Inter-Agency Fire Centre.

Strenfel looked into tools they use to prevent fires and help with decision-making, such as artificial intelligence and machine learning.

“Some data scientists on the team can take these very robust historical data sets, compare them with devastating fires from the past, run through AI machine learning algorithms, and come up with this model. “We can see in great detail what the risks are across our territory every day, every hour.”

Since 2017, Strenfel said the team has set up more than 1,600 weather stations, with stations reporting data every 10 minutes.

Information from the Stop Probability Weather (OPW) model and FPI is key to know where the Public Safety Stop (PSP) is.

“By taking the PSP scale, we actually prevented some large and catastrophic fires, given the analysis we did,” Strenfel says. “You can't prove it 100%. It's very difficult to prove it negative. But some tools and technologies are working. We'll also review the use of powerline safety settings that are enhanced in a very similar way.”

Strenfel predicted that the tools will become more sophisticated in the future and will keep up with the latest technology.

Report corrections or typos.



Source link

Leave a Reply

Your email address will not be published. Required fields are marked *