Ayonix Perimeter & Zone Intelligence
Protect Every Boundary. Understand Every Zone.
Detect unauthorized entry, monitor restricted areas, understand zone occupancy and identify configured safety risks using AI-powered video analytics.
Interface illustration. Not measured data.
One platform
Security, safety and operations from the same feed — because a camera that only raises alarms is a camera nobody reviews.
Existing IP camera over ONVIF or RTSP.
People and vehicles found in the frame.
Object type resolved so a rule can filter on it.
Your zone, line, direction, dwell or occupancy rule.
Event raised with severity and destination.
Snapshot and playback for an operator to confirm.
Search by camera, zone, object and time.
Export or push to the systems you already run.
One platform. Multiple security scenarios.
Every module below is the same detection and tracking layer with a different rule on top. Configure the ones your site actually needs.
Intrusion detection
Mark a protected area and detect people or vehicles entering it within the camera view. The rule is a shape you draw, not a fixed box.
Perimeter & boundary crossing
Define a virtual boundary around a site and focus attention on the crossings that matter, with the direction of travel recorded.
Virtual line crossing
A virtual line turns a movement into an event. Filter it by object class and by direction so an entry and an exit are not the same thing.
Occupancy & capacity
Maintain how many people are inside a zone and raise an event before a configured threshold is exceeded rather than after.
Loitering & dwell
Measure how long a person remains inside a zone and raise an event when dwell passes the threshold you set for that area.
Wrong direction
Where traffic should only travel one way, a directional rule on a line or a zone turns the exception into an event.
Workplace vehicle-path safety
Draw the area where vehicles move and raise an event when a person is detected inside it. A configured zone rule — useful for reviewing near-misses and for training, and never a replacement for the physical controls that keep people out of that path.
Visual rule builder
Zones and lines are configured per camera during commissioning, then tested against real footage before they are saved.
Security operations
Live feeds, the event stream, and the trend behind them — so an operator can act now and a manager can see the pattern later.
All values shown are illustrative interface data.
Investigation
An alert is only useful if the next four clicks are obvious. Every event carries what an operator needs to confirm it and what an investigator needs to find it again.
VMS integration
Surface perimeter and zone events alongside the video workflows in place today.
Confirm plugin availability, supported VMS versions, licensing, and event capabilities with Ayonix before specifying a deployment.
Ayonix works with camera and VMS environments including Genetec, Milestone Systems, Network Optix, Hikvision and Axis Communications. Compatibility is confirmed per deployment.
Discuss VMS integrationExternal systems
Perimeter events rarely end in the video system. They open a door, raise a ticket or land in a log.
Doors, gates and barriers triggered on a rule.
Events into the platform that already runs the site.
Security events into the monitoring pipeline.
Counts and events for reporting and analysis.
Your own systems, over REST and webhooks.
Scheduled exports and CSV for review cycles.
Integration availability depends on product version, licensing, site topology and project scope. Requirements are validated during solution design.
Ayonix AIBOX · Edge AI
AIBOX runs the analytics beside the cameras, so video stays on site and only events and metadata travel.
Deployment
Latency, bandwidth, data control and site autonomy decide this — not preference.
Local inference on AIBOX beside the cameras. Low latency and modest bandwidth, with video staying on site.
Best for remote sites and gatesAn AI server inside the customer network, aggregating streams centrally within your own environment.
Best for single large sitesLocal inference for site autonomy, with central management and a single view across locations.
Best for estates and networksWhere it is used
Camera evaluation
The only honest way to know whether a view supports a rule is to run the rule against footage from that view.
A busy period and a quiet one, from the cameras in question.
Where the boundary is, and which way across it matters.
Object classes, thresholds, schedules and severity.
Reviewed against what actually happened in that footage.
Buyer questions
Software that detects people and vehicles in a camera view and raises an event when one crosses a virtual boundary or enters an area you have drawn. The boundary is configured in software rather than built in hardware, so it can be changed without touching the site.
Often yes. Suitable RTSP or ONVIF streams can be used. Camera height, angle, lens, resolution, lighting and the distance to the boundary all affect what a given view can support, so representative footage is reviewed before deployment.
A shape drawn on the camera view — a polygon — that the analytics treat as a protected area. Entering it, leaving it, staying in it too long or exceeding a headcount inside it can each raise a different event.
Yes. A virtual line raises an event when a tracked object crosses it, and the direction of travel is recorded so entries and exits are distinguished. Several lines can be configured in one camera view.
Yes. Live occupancy is maintained per zone against a configured capacity, and an event can be raised as the threshold is approached or exceeded.
Yes. A dwell threshold is set per zone, and an event is raised when a person remains inside that zone for longer than the configured time. Different areas can carry different thresholds.
Yes. Detected objects are classified, and rules can filter on class — so a vehicle crossing a boundary and a person crossing it can raise different events, or one and not the other.
A vehicle-path area can be drawn and an event raised when a person is detected inside it. This is a configured zone rule for review, near-miss analysis and training. It is not a collision-avoidance system, it carries no safety certification, and it does not replace barriers, segregation, training or any other physical or procedural control.
Perimeter and zone events can be surfaced alongside video workflows. Confirm plugin availability, supported VMS versions, licensing, and event capabilities with Ayonix before specifying a deployment.
Yes. An AI server inside your own network can process streams centrally, keeping video within your environment and exporting only the events and metadata your policy allows.
Yes. AIBOX runs inference beside the cameras, which keeps latency low and bandwidth modest and lets a site keep working when the link to a central system is down.
Safer spaces. Stronger tomorrow.
Bring us the camera watching the boundary you care about most. We will help define the zone, choose the rule and evaluate it on your own footage before anything is specified.
