Sensor AI pitches early-wildfire detection pilot to Laguna Beach preparedness committee

Ryan Honore, founder of Sensor AI, told the Laguna Beach Emergency Disaster Preparedness Committee on July 6 that his company’s sensor network and on‑device artificial intelligence can detect very small, incipient wildfires and notify fire authorities within minutes.

Honore said he began the project after a 2020 campfire and has developed the system with support from university partners and local agencies. "I am Ryan Honore, the founder of Sensor AI," he said, adding that Sensor AI had become an XPRIZE finalist and in April detected a small wildfire in Calabasas. Honore described a two-part system of risk‑assessment units (measuring wind, vegetation and moisture) and detectors that register three signatures—infrared flame, smoke and a thermal‑fluids heat plume—and then pass images through layered AI to reduce false alarms.

The presentation and slides emphasized speed and coverage. Honore said the detectors can identify fires smaller than 3 by 3 feet, and reported performance numbers in which the system detected that Calabasas fire from up to 2.6 miles with a clear line of sight; partial line‑of‑sight detection at about 1.6 miles; and thermal‑plume detection with no direct line of sight at roughly 1,500 feet. "By the time the fires are detected, they've already grown out of hand," he said, arguing that earlier notification can limit spread.

Committee members pressed technical details. Honore said what3words coordinates will be added to alert messages to make location useable for responders, and that the platform offers role‑based user interfaces: fire authorities get detailed operational pages while residents see limited views and notifications. On operational control he said the system is a decision‑support tool: detectors notify an on‑duty battalion chief, but "we keep the power in the people," meaning final dispatch decisions remain with fire personnel.

A major line of questioning focused on fuel‑moisture measurement. Some members and outside experts questioned whether soil‑moisture probes alone can reliably represent live and dead fuel moisture; Honore and a company colleague said a combination of many small soil probes and separate dead‑fuel sensors—coupled with lab work at UCLA and UCI—was their proposed approach. David Horn of the Greater Fire Safe Council said he has permits and field sampling underway and cautioned that ground‑truth sampling remains important.

On pricing and production, Honore said costs vary with configuration and volume; small orders can be built in days in Orange County, while citywide procurement may take two to three months for component lead times. He said Sensor AI has worked with land managers and the Orange County Fire Authority on pilot deployments and expects to deploy roughly 12 units in Laguna Beach for phase‑1 testing in collaboration with the Laguna Beach Fire Authority.

Committee members offered local data and operational assistance, including sharing fuel‑moisture sampling procedures and pilot sites. Honore said the company is open to selling through homeowners associations and partnering with insurers if the sensors demonstrate reduced risk.

The committee did not take any formal action on procurement at the July 6 meeting; members invited Honore to return and continue coordination as pilots proceed.