Exploring the Viability of Detecting Pole Fires
Since November 2022, Dryad has been collaborating with Pacific Gas and Electric on pilot projects to test the effectiveness of its wildfire detection technology when it comes to pole fires.
PG&E is seeking a solution to effectively 'smell' fires from their wooden power line poles. This use-case was a new one for Dryad, and involved new training for its artificial intelligence algorithms back at its R&D laboratory in Eberswalde, Germany.
We have extensively trained our AI at Dryad to be able to detect different types of fire from all over the world and have therefore been able to also train it to accurately and reliably detect early-stage fires from PG&E's poles.
Wildfires are a significant environmental, societal and economic problem and like other utilities companies, PG&E must protect their infrastructure, assets and mitigate their liabilities caused by wildfires. Dryad's ultra-early wildfire detection technology allows them to do this.
"[Dryad's technology] uses more of a smell-type sensor which can be very beneficial as you can detect smell for longer distances.
This technology has potential in our rural areas where we have limited communications and can give us a very granular approach in addition to the wildfire detection cameras we have in place.
After three months of pilot testing in our labs, we feel pretty confident that the technology is reliable for us to move over to the next pilot where we will better learn and understand how these devices work in rural areas, how the solar panels behave in charging these devices, and give us really good early detection for any wildfire risk."
Gavin Fong, Principal, Wildfire Risk Management, PG&E
The fact that Dryad's sensors can updated over-the-air was also a key feature that attracted PG&E to the collaboration. Once deployed, our sensors can be enhanced remotely with patches and software updates, that is: over-the-air (OTA). For example, once our AI learns a new smell or use-case, we can quickly redeploy these updates across our large-scale mesh gateway, ensuring the most up-to-date and reliable wildfire sensing possible.