Skip to content
Illustrative public plaza with people walking, overlaid with a discreet camera-network graphic

Public safety · Face surveillance

See the moment. Understand the risk.

Turn authorized camera networks into real-time watchlist alerts and faster, evidence-led investigations — without replacing human judgment.

Plan a field evaluation Explore the workflow
Tokyo · since 2007
Japanese engineering
NIST FRTE participant
Independent evaluation
Edge · on-prem · cloud
Flexible architecture
VMS · API · ONVIF
Open integration layer

Mission outcomes

From hours of video to moments that matter.

Public-safety teams do not need more feeds. They need a controlled way to find, verify and act on relevant events across cameras, locations and time.

Live watchlist alerts

Screen authorized video streams against one or more controlled watchlists, and route candidate matches to trained operators with the camera, time, score and reference record.

  • Multi-watchlist policies
  • Zone and schedule rules
  • Operator verification queue

Missing & vulnerable persons

Support time-critical searches across permitted cameras and help teams focus review on the most relevant appearances.

  • Rapid gallery enrolment
  • Cross-camera sightings
  • Case-scoped access

Post-event investigation

Search recorded footage by face and reconstruct a timeline across locations without manually watching every minute.

  • Video file and VMS search
  • Event bookmarks
  • Evidence-ready export

Critical-site awareness

Apply different identification and notification policies to transport hubs, government facilities, borders and major-event zones.

  • Site-specific thresholds
  • Restricted-area workflows
  • Escalation by role

Operational workflow

Ayonix does not miss out!

Ayonix recognizes all walking people in the streets. No missing face, no missing identities based on its 20 years of experience in the field.

Ayonix Command
Two people walking along a city street, each face framed by a detection box drawn over the video
Tested and deployed in 2019 in Nagoya, Japan by Lets Corporation, Japan.
  1. DetectFind usable faces in live or recorded video.
  2. CompareSearch only against authorized, purpose-bound galleries.
  3. VerifyGive trained operators the context to assess a candidate.
  4. RespondTrigger a defined workflow and preserve the audit trail.

Investigations

Reconstruct the journey. Keep the evidence connected.

Start from an image, a clip or a previous sighting. Search permitted recordings, group likely appearances and build a reviewable sequence for investigators.

  1. Search by faceUpload a probe image or select a detected face from video.
  2. Correlate across camerasReview likely appearances by time, location and similarity.
  3. Build a case timelineBookmark events, add notes and export approved evidence.
See an investigation case study
A person walking into a shop entrance, framed by a detection box, with three captured face images stacked in a review panel beside the video
Tested in 2020 in seven SoftBank branches in Kofu, Japan for customer demographics.

Deployment & integration

Add intelligence. Keep your VMS.

Use Ayonix as a biometric intelligence layer within existing video operations. Plugin and connector scope is confirmed against the customer's versions, licenses and workflow.

Milestone XProtectGenetec Security CenterNetwork Optix Nx WitnessONVIFRTSPREST APIWebhooks

Edge

Low-latency processing near cameras with reduced upstream bandwidth.

On-premises

Keep biometric data and operations inside customer-controlled infrastructure.

Approved cloud

Scale centralized workloads where policy and connectivity permit.

Hybrid

Process video locally and federate governed events to a central command layer.

Integration availability is project-specific. Certification, version support and required licenses must be validated before procurement.

Request an integration review

Responsible operations

Control what the system can do — and prove what it did.

Public-safety deployments succeed when the operational safeguards are designed with the technology.

Purpose-bound watchlists

Separate galleries by mission, authorization, geography and validity period.

Human confirmation

Treat candidate matches as leads. Require trained review before consequential action.

Threshold policy

Set and test thresholds by camera conditions, operating risk and response workflow.

Retention controls

Apply event and evidence retention rules aligned with lawful purpose and policy.

Role-based access

Limit enrolment, search, export and administration to authorized functions.

Auditable actions

Record searches, reviews, decisions, exports and administrative changes.

What buyers need to know

Can Ayonix work with existing CCTV cameras?
Often, yes. Suitability depends on face size, camera angle, lighting, motion, compression and scene density. Ayonix recommends validating performance with representative footage before deployment.
Does a face match make an enforcement decision?
No. A candidate match is an investigative lead or alert for trained human review. Operational policy should require confirmation and define when further action is permitted.
Where can the platform be deployed?
The solution can be designed for edge, on-premises, approved cloud or hybrid environments, based on latency, scale, connectivity and data-governance requirements.
Which VMS platforms can be integrated?
Ayonix offers project-based integration options for Milestone XProtect, Genetec Security Center and Network Optix Nx Witness, as well as ONVIF, RTSP, APIs and webhooks. Compatibility and plugin scope are confirmed during solution design.
How should accuracy be evaluated?
Use the intended cameras, representative people and conditions, the real gallery size and the planned decision threshold. Measure false matches, missed matches, image-quality failures, speed and the complete operator workflow — not one headline number.

Start with your environment

Bring one site, one watchlist and a real question.

We will map camera suitability, expected throughput, integration points, governance controls and a measurable evaluation plan.

Plan a field evaluation Request the solution brief
Camera and scene assessmentIntegration architecturePoC success criteria