Computing

Watch Duty Uses Google's Gemini AI to Speed Up Wildfire Alerts

Tendela Briefing ·
4 views · 0 Comments
Watch Duty Uses Google's Gemini AI to Speed Up Wildfire Alerts

Image: Google · Source

Watch Duty, a nonprofit wildfire alert app, utilizes Google's Gemini-powered tool to transcribe noisy fire radio in under 30 seconds, improving early wildfire detection and faster alerts for communities.

Watch Duty, a nonprofit wildfire alert platform operated by active and retired firefighters, dispatchers, and first responders, has integrated a new artificial intelligence tool powered by Google's Gemini model to significantly improve its wildfire alert capabilities. This Gemini-powered system converts noisy and static-filled fire radio transmissions into clear, real-time searchable text. The collaboration emerged through a Google.org Fellowship, in which a dedicated team of Google engineers and researchers spent six months working alongside Watch Duty. Using more than 20,000 hours of past radio data, they developed an AI tool tailored to reduce static and accurately transcribe specialized firefighting jargon, easing the task of identifying new wildfires amid noisy audio feeds.

Wildfire alerting hinges on speed; minutes can be critical for safe evacuations. However, many early fire warnings still come through old analog VHF radio channels, which often suffer from engine noise, wind, static, and complex, rapid-fire communications. Watch Duty's volunteer reporters monitor as many as 20 radio feeds manually, a challenging and error-prone process.

The Gemini tool addresses these issues by lowering the word error rate on degraded VHF audio from 38% to 22.34% and streamlining transcription to handle over 200 hours of radio speech daily from around 400 active feeds. It has been tested to scale to more than 10,000 simultaneous streams. This real-time transcription capability enables Watch Duty to spot new wildfire dispatches in less than 30 seconds, even when no volunteer happens to be listening at that moment, effectively closing critical coverage gaps.

During a real-world test in July 2026 amid the Railroad Fire in Washington, the tool rapidly transcribed a 911 dispatch through heavy static and flagged the incident within 30 seconds. This rapid detection allowed Watch Duty to verify the fire quickly and send a Critical Alert to users overnight—the first such alert sent during nighttime—giving residents more time to evacuate safely.

This AI-powered improvement has raised volunteer reporter satisfaction with transcription tools from 35.3% to 92.3%. Beyond transcription, the partnership between Google and Watch Duty also includes $2 million in funding and Google Cloud credits provided in 2024 to help validate new fires using satellite, camera, and radio data. Additionally, Google Search has featured Watch Duty's alerts alongside official emergency advisories during significant wildfires, and Firebase Cloud Messaging scaling has facilitated timely push notifications to millions of users.

Ultimately, this ongoing collaboration aims to enhance early wildfire detection nationwide, delivering faster alerts that can save lives. By harnessing Gemini's advanced AI capabilities, Watch Duty advances its mission to warn communities promptly, especially when every second counts.

Watch Duty rendering
Watch Duty rendering · Source ↗

Sources and original reporting

Comments (0)

No comments yet. Start the discussion.

Write a comment

Comments are published after moderation. Your name and comment will be visible publicly. Account