Saturday, November 21, 2026
Translate
PrivacySubmit newsShare your story Newsletter Advertise
Breaking
AI news

Gemini Tool Spots Wildfire Dispatch in Under 30 Seconds

A Gemini-powered tool built with Google.org support helps Watch Duty turn noisy fire radio into searchable text and flag dispatches in under 30 seconds.

Gemini Tool Spots Wildfire Dispatch in Under 30 Seconds
Image: Google

Google said a Gemini-powered tool is helping Watch Duty catch wildfire dispatches from noisy radio in under 30 seconds, including a Railroad Fire dispatch after the fire ignited at night in July 2026.

Why it matters

Google said minutes can decide whether people get out safely when a wildfire starts, but the first word of a new fire often travels over decades-old analog VHF radio.

Engine noise, wind, static and rapid-fire tactical jargon can make that audio hard to hear, while Watch Duty reporters monitor up to 20 feeds manually.

The TechUpscale Brief. The day's cyber, AI and tech news in one short email, every weekday morning. Free. Unsubscribe anytime.

Google said the Gemini-powered tool removes the added strain of deciphering noisy audio so volunteers can focus on verifying incidents and getting alerts out.

How the tool works

A dedicated team of Google engineers and researchers spent six months embedded full-time with Watch Duty, a nonprofit wildfire alert app run by active and retired firefighters, dispatchers, and first responders.

Through the Google.org Fellowship, they used more than 20,000 hours of past radio recordings to build a tool that wipes out heavy static and translates specialized firefighting terminology into clean, searchable text.

Google said the tool has helped close coverage gaps in areas with no other automated dispatch detection by sending reporters a signal in less than 30 seconds.

Results and live test

Google reported that the tool reduced word error rate on degraded VHF radio audio from 38% down to 22.34%.

It transcribed more than 200 hours of radio speech a day from about 400 active feeds and has been designed and tested to handle more than 10,000 feeds at once.

Google also said volunteer reporter satisfaction with radio transcription tools increased from 35.3% to 92.3% as a result of the collaboration.

In July 2026, the Railroad Fire ignited at night in Washington, and the Gemini-powered transcription tool transcribed the first 911 dispatch from noisy radio audio and flagged it in under 30 seconds.

Google said that head start gave Watch Duty's team time to verify the fire and push a Critical Alert, which breaks through do-not-disturb mode in urgent situations. It was the first time Watch Duty had sent a Critical Alert overnight.

What to watch

Google said the goal is simple: get people the warning they need while there's still time to act when the next fire starts.

Key facts and where they come from
  • Google.org Fellowship support helped build a Gemini-powered tool that turns noisy fire radio into real-time text.
    Built with pro bono support from a Google.org Fellowship, a new Gemini-powered tool turns noisy fire radio into real-time text
  • The tool helps Watch Duty spot new wildfire dispatches in under 30 seconds.
    helping Watch Duty spot new wildfire dispatches in under 30 seconds
  • The tool reduced word error rate on degraded VHF radio audio from 38% to 22.34%.
    Reduced word error rate on degraded VHF radio audio from 38% down to 22.34%
  • The system transcribes more than 200 hours of radio speech daily from about 400 active feeds and is designed and tested for more than 10,000 feeds.
    Transcribed more than 200 hours of radio speech a day from about 400 active feeds, and has been designed and tested to handle more than 10,000 feeds at once.
  • The Railroad Fire ignited at night in Washington in July 2026.
    In July 2026, a fast-moving wildfire — the Railroad Fire — ignited at night in Washington.
  • The Gemini-powered transcription tool transcribed a dispatch from noisy radio audio and flagged it in under 30 seconds.
    the Gemini-powered transcription tool transcribed it straight from the noisy radio audio and flagged it in under 30 seconds.
  • Volunteer reporter satisfaction with radio transcription tools increased from 35.3% to 92.3%.
    volunteer reporter satisfaction with radio transcription tools increased from 35.3% to 92.3%.

The open question: How should wildfire alert teams verify AI-transcribed radio dispatches before issuing public alerts?

Tell us what you think · Discuss it in the AI room

Reader poll: Should AI transcription be used to trigger urgent wildfire alerts?

A show of hands from readers, not a scientific poll. Say why in the comments.

Add TechUpscale as a preferred source on Google

The TechUpscale Brief

The day's cyber, AI and tech news in one short email, every weekday morning. Free. Unsubscribe anytime.

I'm most interested in

How should wildfire alert teams verify AI-transcribed radio dispatches before issuing public alerts?

Comments are moderated. Be civil and stay on the story: no personal attacks, hate, spam or ads, and no one else's private details. Your email is never shown. Privacy policy.

More AI