You hear sirens in town. You wonder what's happening. We wondered: could technology help answer that question without making everyone listen to a scanner all day?
The idea began when Scott tried a phone app that was supposed to transcribe police and fire radio traffic in real time. Unfortunately, it didn't do a very good job. The conversations came through as garbled words, misunderstood locations and half-finished sentences.
Scott tried to improve the transcription himself. That didn't solve the bigger problem: even a perfect word-for-word transcript wouldn't necessarily tell a listener what was going on.
That question became the inspiration for the NewingtonWeather Scanner Summary Service. With the help of AI and a lot of local knowledge, Scott built a system that listens for Newington police, fire and EMS radio traffic and turns meaningful calls into short, readable updates.
Here's the simple version
The system monitors scanner audio and detects when radio traffic is being transmitted.
AI converts speech to text, interprets local terminology and sorts actual calls from routine chatter.
A brief explanation appears on our scanner page, helping visitors understand recent activity.
Teaching AI a little about Newington
The hard part wasn't just recognizing words. Dispatchers use radio codes, shorthand and local place names that aren't always familiar to a general-purpose AI system. Scott spent considerable time refining how the service understands those details and decides which transmissions belong to the same incident.
We continue to fine-tune the system as new situations arise. We believe this is an unusual approach for a community weather website, although we can't verify that it was the first of its kind.
Curious about what's happening around town? Stop by the scanner page. And if you're a tech enthusiast who enjoys looking under the hood, we've also written a more detailed explanation of the system.
