New Orleans 911 system uses AI to filter repeat auto-crash reports
The Orleans Parish Communication District has deployed an AI triage tool that answers certain repeat 911 calls about already-reported car crashes, keeping human operators free for other emergencies.
In New Orleans, the Orleans Parish Communication District has integrated Carbyne's AI-V Emergency Call Triage into its 911 center to manage duplicate reports of a single auto accident. The system activates solely when every human call-taker is occupied and the incoming call is detected within roughly 200 meters of a crash already logged in the system. Callers are asked whether they are calling about that specific crash; confirming keeps them connected to a human operator, while any other answer routes the call back into the normal queue.
The AI does not have the authority to dispatch emergency services, prioritize calls, or terminate the call, and it reverts to a human agent if it cannot interpret the response. It supports multiple languages and has been trained on local recordings to recognize regional accents and street names. All AI-handled calls are logged with standard metadata and audio recordings, and the agency reports no false positives or negatives during its initial three-month review. Carbyne claims the tool reduced redundant crash calls by more than 30% and shortened response times for distressed callers by up to 40 seconds, though these figures come from the vendor.
Why it matters
AI can help 911 centers handle call surges, but safeguards are needed to ensure human oversight in emergencies.
In this story
