Massive parallel capture
Each edge node continuously demodulates 16+ aviation AM and marine NFM channels in parallel. No scanning, no missed transmissions when two channels key together.
01 / 09
VOICE INTELLIGENCE · v4.12 · AVIATION + MARINE
A distributed sensor network that captures, transcribes, and classifies voice traffic across the public aviation and marine VHF spectrum, time-correlated to the aircraft and vessels you already track. Mayday to map pin in under five seconds.
THE GAP · WHY THIS NOW
A flight ops center, harbor master, search-and-rescue coordinator, or air-ambulance dispatcher who wants real-time situational awareness today is forced into an option that misses something.
One channel at a time. No transcript, no replay. Depends on a human paying attention 24 hours a day.
Community-fed audio with random coverage gaps, 10 to 30 second delay, no metadata, no chain of custody.
One field, audio only. No transcription, no marine, no SAR alerting, no cross-band correlation.
Every keyed transmission across your area of interest, transcribed and classified, time-correlated to live tracks. Machine readable.
CORE CAPABILITIES · 09
Each capability is independently shippable; together they are a single operational picture for the voice traffic over your area of responsibility.
Each edge node continuously demodulates 16+ aviation AM and marine NFM channels in parallel. No scanning, no missed transmissions when two channels key together.
01 / 09Every keyed transmission is stored as a discrete clip with site, antenna, frequency, RSSI, SNR, hash, and a UTC stamp disciplined to under 50 ms drift.
02 / 09Domain-adapted speech recognition for ICAO aviation and IMO marine procedure words. Sub-10% WER on aviation, sub-15% on marine. Two to five second latency.
03 / 09Mayday, pan-pan, sécurité, declared emergency, minimum fuel, person overboard, NORDO patterns. Multi-label, with operator-tunable false positive thresholds.
04 / 09Lat/lon, intersections, mile markers, buoys, runways, named landmarks. Parsed from speech and dropped on the map automatically.
05 / 09Spoken callsigns and vessel names linked to registry, then matched against the live aircraft and vessel tracks already in your common operating picture.
06 / 09One incident produces traffic on four or five frequencies. Distress, controller, dispatch, bridge-to-bridge, fused into a single chronological incident record.
07 / 09SHA-256 audio hashes, signed sidecar metadata, end-to-end audit log. Exportable for incident review, regulatory inquiry, and after-action documentation.
08 / 09English, Spanish, French, Japanese, Korean. Additional language packs available for border, Caribbean, Great Lakes, and Pacific operations.
09 / 09THE MOAT · CROSS-BAND CORRELATION
A single mayday produces traffic on the distress channel, the controller, the rescue coordinator, dispatch, and bridge-to-bridge, all within a minute. Voice intelligence reconciles them into one record, attached to the live track.
“Mayday, mayday, mayday.” Vessel taking on water, two souls aboard, one mile west of buoy three-one.
“Position three-seven decimal three-two-five north, eight-nine decimal five-seven-six west.” → pin dropped
Coast Guard sector acknowledges. Initiating coordination on 22 alpha. Requesting bridge-to-bridge on thirteen.
“Any vessel in vicinity of buoy three-one, please respond.” Two acknowledgements received and logged.
“Air Evac eight-zero, divert SAR, position three-seven point three-two-five north.” ADS-B match: AE80, 18 nm SE, ETA 6 min.
Air ambulance dispatch confirms tasking. Pre-arrival channel set. Receiving facility on standby.
Center reroutes inbound traffic clear of the recovery box. Block altitude assigned to the asset.
SPECTRUM COVERAGE
Voice intelligence operates entirely on openly broadcast aviation and marine spectrum, by design intended to be received by anyone in radio range. Public safety bands are deliberately out of scope.
FIELD CONFIGURATIONS
Voice intelligence ships as edge nodes plus a central fusion tier. Configurations are chosen by area of responsibility: a single field, a metro, a coastline, a navigable river system.
FUSION INTEGRATION
Voice intelligence is not another window. It feeds the map your operators already watch: voice transcripts ride alongside the same aircraft and vessel tracks the rest of your common operating picture is built on.
Each transcribed transmission becomes a track update or distress event on the maps your operators already use, with a clickable link to the original audio.
Voice events appear as sensor observations correlated to airframe and vessel tracks, available to any operator inside the area of responsibility.
A live, joinable dataset: clip, frequency, transcript, classifier labels, extracted entities, matched track, audio URL. Built for analytics and dashboards.
Where authorized, voice transcripts attach to flight history records so reviewers see what was said alongside the track that was flown.
Equivalent for vessels: voice events attached to the AIS track of the boat that produced them, for ops review and post-incident analysis.
Recent uncontrolled-field activity surfaces in pilot-facing apps for the destination airport, ahead of the first call to common traffic.
Inbound vessel hails on bridge-to-bridge and distress automatically surface in marina management for slip and fuel coordination.
A streaming gRPC API with backpressure-aware delivery, REST and GraphQL surfaces, a WebSocket live stream, and outbound webhooks for one-off routing.
COMPLIANCE · REGULATORY POSTURE
Every band the system operates on is openly broadcast and intended to be received by anyone in radio range. There is no encrypted-traffic ambiguity, no public-safety political surface, and no wiretap exposure. Commercial deployment is unambiguously lawful in the United States and most allied jurisdictions.
DEPLOYMENT · INQUIRY
Pilot one site in a week. Cover a metro, a coastline, or a river system in a quarter. Voice events land on the maps you already operate, attached to the tracks you already trust.