Ayonix Video Analytics

Industry

Video analytics for utilities, substations and regulated sites

In short

Critical infrastructure video analytics provides perimeter intrusion and restricted-area detection on remote and regulated sites, with air-gapped deployment where outbound connectivity is prohibited. Data control is usually the deciding factor, which is why edge and isolated processing are the normal patterns.

CCTV-protected perimeter fence line of a modern industrial facility at dusk
Conceptual illustration of a monitored industrial perimeter. AI-generated conceptual image, not a customer deployment.

The problem

What critical infrastructure teams are dealing with

  • Remote substations and pumping stations have cameras and no one watching them.
  • Cloud analytics is ruled out because the operational technology network permits no outbound connectivity.
  • Regulatory reporting requires an audit trail that most analytics deployments cannot produce.
  • Bandwidth to remote sites cannot support continuous video backhaul.
  • A nuisance-alarm rate that would be tolerable elsewhere is not tolerable where every response means a vehicle dispatch.

Camera placement

Where cameras need to be, and why

Camera position decides more of the outcome than any software setting. These are the placements that matter in this sector.

Recommended camera placement forCritical infrastructure
Location Purpose What matters
Site perimeter Intrusion detection Stable mounting with verified IR performance at the actual boundary range, not the specification range.
Compound and equipment areas Out-of-permit presence Schedule-aware rules aligned to planned maintenance windows.
Access gates Vehicle and person crossing Crossing events give a timestamped arrival record without needing plate processing.
Control and switch rooms Restricted presence Verified low-light performance is essential in normally unlit rooms.
Remote outstations Edge intrusion detection Edge processing is mandatory where backhaul cannot carry continuous video.

Workflow

From event to acknowledged action

Event workflow

  1. 1 Site cameras are ingested locally; no video leaves the site in an air-gapped deployment.
  2. 2 Objects are detected and classified; zone and line rules run to the permit schedule.
  3. 3 Events are stabilised aggressively, because every alert may trigger a physical dispatch.
  4. 4 Evidence is created and retained locally under the site retention rule.
  5. 5 Events are delivered to the local VMS or the isolated operations network.
  6. 6 Every event, access and configuration change is written to an immutable audit log.

Operator workflow

  1. 1 Control room receives a classified perimeter event with evidence from an unmanned site.
  2. 2 The event is assessed against the site response procedure before any dispatch.
  3. 3 A patrol or engineer is dispatched only on an assessed event, not on a raw detection.
  4. 4 The assessment, decision and outcome are recorded against the event.
  5. 5 Audit logs are exported for the regulatory reporting cycle.

Operations

Dashboards, integration and deployment

Dashboard metrics

  • Perimeter events per site per night
  • Assessed-as-genuine versus assessed-as-irrelevant
  • Dispatches per genuine event
  • Out-of-permit presence by site
  • Audit-log completeness
  • Camera and edge-node availability per site

Integration options

  • VMS alarm delivery on the isolated operations network
  • Webhook delivery into control-room systems where permitted
  • REST retrieval for regulatory reporting
  • ONVIF and RTSP ingestion with no outbound dependency
Compatibility states →

Deployment options

  • Air-gapped deployment, frequently the mandated pattern for regulated operational networks
  • Edge processing at remote outstations with constrained backhaul
  • On-premise processing at staffed sites
  • Multi-site management across an estate, operating within the network isolation boundary
Compare architectures →

Evidence

Deployment pattern

Isolated site pattern

Edge or air-gapped perimeter and restricted-area detection with local evidence, immutable audit logging and delivery into the isolated operations network.

Recall and irrelevant-alert rates are measured on your own sites under real night and weather conditions. No cross-site benchmarks are published.

This is an anonymised pattern describing how a deployment of this kind is structured. It is not a named customer reference. No customer names, logos, testimonials or measured results are published on this site, because none have been supplied for publication. More on how evidence is handled .

Governance

Privacy and governance in this sector

  • Air-gapped and edge deployment keeps video and events inside the operational boundary, which is usually the deciding requirement.
  • Immutable audit logging of events, access and configuration changes supports regulatory reporting.
  • Rules operate on object class, not identity; no biometric processing is performed by these analytics.
  • Retention rules should be set explicitly per site and evidenced, not left to default.
  • Where staff work areas are monitored, employee monitoring obligations apply alongside the security purpose.

Frequently asked questions

Can this run with no internet connection at all?

Yes. Air-gapped deployment is a published Ayonix deployment model, and for regulated operational networks it is frequently the only acceptable one. Detection, rules, evidence and audit logging all operate locally with no outbound dependency.

Why does nuisance-alarm rate matter more here?

Because a response often means dispatching a vehicle to a remote site. An alert rate that a staffed urban control room would absorb becomes an unacceptable operating cost when every event is a journey. Rules are tuned more conservatively as a result.

What audit evidence is produced?

Events, operator assessments, evidence access and configuration changes are logged immutably. That record is usually what makes the deployment acceptable to a regulator, so it is designed in from the start rather than added afterwards.

Other industries

Author
Gabriel Bamola, Chief Marketing Officer, Ayonix
Technical review
Dr Sadi Vural, Founder and Chief Executive Officer, Ayonix
Published
Last reviewed

Design a critical infrastructure pilot around your own cameras

Start with a camera readiness review and a log-only baseline, then measure against acceptance criteria we agree before the pilot begins. That is what turns an analytics evaluation into a decision.