security analytics

Restricted area monitoring for zones that should normally be empty

In short

Restricted area monitoring raises an event when a person or vehicle is present inside a zone that should be empty or access-controlled at that time. Ayonix lists restricted zones as a published video analytics capability, with loading docks as a worked example. Rules are schedule-aware so routine authorised activity does not generate noise.

Request a pilot Send a sample video

Conceptual illustration of distribution-centre operations. AI-generated conceptual image, not a customer deployment.

Scope

What this analytic does, and what it does not

What it detects

Presence of a person inside a defined restricted zone
Presence of a vehicle inside a defined restricted zone
Presence outside an authorised schedule window
Continued presence beyond a configured duration

What it does not guarantee

That the person detected is unauthorised — the system detects presence, not permission
Detection where the subject is fully occluded by plant, racking or a parked vehicle
Any life-safety function inside hazardous areas
Replacement of access control, interlocks or permit-to-work procedures

The problem

What customers are actually dealing with

These are the situations that lead teams to look at this analytic in the first place.

A loading dock is open during shifts and should be empty overnight, but nobody is watching the camera when it is not.
Plant rooms, substations and roof areas are entered without a permit and this is only discovered during an audit.
Access control records show a door opening but not what happened in the space afterwards.
Contractors work in a bonded or high-value area outside agreed windows and there is no timely record.

How it works

Detection workflow

Every event carries the rule that produced it, so an operator can see why they were alerted.

1

Ingest

The camera covering the restricted space is read and monitored for stream health.
2

Detect and classify

People and vehicles inside the frame are detected and classified.
3

Track

Objects are tracked so that presence is measured as a continuous duration.
4

Apply zone and schedule

The rule tests zone membership against the active schedule for that day and time.
5

Stabilise

A dwell threshold prevents momentary pass-throughs from raising an event, and cooldown limits repeats.
6

Create evidence and deliver

Frame, clip and event record are produced and routed to the VMS or operator queue.

Restricted area monitoring workflow diagram

Configuration options

One or more polygon zones per camera, each with its own schedule
Minimum presence duration before an event is created
Object classes that trigger the rule
Escalation after continued presence beyond a longer threshold
Cooldown between repeat events for the same track
Optional exclusion sub-zones for permanently occupied positions such as a picking station

Alert quality

How irrelevant and duplicate alerts are reduced

No video analytic eliminates false alerts. These are the mechanisms that reduce them, and the residual rate is measured on your own cameras.

The presence threshold removes people who legitimately walk past the edge of a zone.
Exclusion sub-zones prevent a fixed workstation inside the wider area from generating continuous events.
Schedules remove the dominant source of noise — authorised daytime activity.
Escalation thresholds let a short presence be logged quietly while a long presence raises an operator alert.
Irrelevant alerts are reduced but not eliminated; the remaining rate is measured per camera during the pilot.

Prerequisites

Camera requirements and environmental limits

Camera suitability decides more of the outcome than any software setting. These are assessed per camera before commitment.

Camera requirements

The zone boundary must be unambiguous in the camera view so the polygon matches the physical space
Sufficient light during the hours the rule is active, including IR performance if the area is unlit
A view that is not dominated by racking, plant or vehicles that occlude the floor area
Object height of at least 8–10% of frame inside the zone
Stable mounting so the polygon remains aligned with the physical area

Environmental limitations

Large parked vehicles or moved stock can occlude a substantial part of a dock zone.
Steam, dust and exhaust in industrial zones reduce detection contrast.
Reflective wet floors can produce mirrored detections in some views.
A zone drawn across a large depth range will have very different pixel-on-target at its near and far edges.

Delivery

Alerts, evidence and where they land

Event and evidence fields

Zone name and rule
Schedule state at trigger
Object class and confidence
Presence duration at trigger
Camera and site
Timestamp with timezone
Still frame and clip
Event identifier for acknowledgement

VMS integration

Zone events map to VMS alarms with the camera bookmarked at the entry time, which lets an operator scrub backwards to see how the person reached the area. Where access control is integrated separately, correlating a door event with a zone event is normally done in the VMS or PSIM rather than in the analytics layer.

See compatibility states →

Deployment options

Edge processing for remote plant rooms and substations with limited bandwidth
On-premise processing where docks and internal zones are already on a site VMS
Air-gapped deployment for utilities and regulated sites
Multi-site rule templates so the same dock pattern is applied consistently across depots

Compare architectures →

Measurement

Dashboard metrics and pilot acceptance criteria

Acceptance thresholds are agreed with you before the pilot starts. This site publishes no benchmark figures, because they do not transfer between sites.

Dashboard metrics

Out-of-schedule entries per zone per week
Mean presence duration per entry
Escalated events versus logged-only events
Zones with the highest event counts
Acknowledgement rate and time to acknowledge
Camera availability per restricted zone

Measurable pilot criteria

Entry recall measured with scripted entries at the start, middle and end of the rule schedule
Irrelevant alerts per zone per day during normal operation
Correct schedule behaviour verified across a full week including weekend variation
Presence-duration accuracy against stopwatch ground truth
Escalation behaviour verified for a sustained presence

Governance

Privacy and governance

The analytic reports presence by object class, not identity.
Where restricted zones are staff work areas, employee monitoring law and consultation duties may apply.
Evidence access should be limited to named roles and audit-logged.
Retention should be set to the shortest period that supports the investigation or audit purpose.

Frequently asked questions

Can it tell whether the person is authorised?

No. This analytic detects presence and classifies the object as a person or vehicle. Authorisation is an access-control question. Sites that need this correlate a zone event with an access-control transaction in the VMS or PSIM, or use the schedule to encode when presence is expected.

Our dock is busy all day — will we be flooded with alerts?

Not if the schedule reflects reality. The rule is normally inactive during operating hours and active only when the space should be empty. Where part of a zone is always occupied, an exclusion sub-zone removes that region from evaluation.

Does this work in an unlit plant room?

Only if the camera has usable IR illumination at the required range. A camera that produces a near-black frame at night cannot support a reliable rule, and the camera readiness review will flag it for a lighting or camera change.

Keep reading

A controlled pilot establishes what this analytic actually does in your environment, against acceptance criteria we agree before it starts. Send a sample video first if you would rather see an assessment before committing to a pilot.

Request a pilot Send a sample video

Restricted area monitoring for zones that should normally be empty

What this analytic does, and what it does not

What it detects

What it does not guarantee

What customers are actually dealing with

Detection workflow

Ingest

Detect and classify

Track

Apply zone and schedule

Stabilise

Create evidence and deliver