Loitering detection AI that spots people who stay where they should not
Most people pass a gate, a cash counter or a car park in seconds. Someone who stays is worth a second look. Eye AI loitering detection runs on the CCTV cameras you already have. It times how long a person stays still in an area you choose. When the time passes your limit, it logs the event with a snapshot, so your team can decide what to do.
ILLUSTRATION WITH SAMPLE DATA
The problem
Nobody watches every camera, so a person standing near a gate for twenty minutes goes unnoticed until something happens.
Searching recordings afterwards means scrubbing through hours of footage to find one person.
Guards cannot tell, from a glance at a wall of screens, who has been there a long time and who just arrived.
How it works
- Pick the spots.In the free camera-readiness check we look at your camera views and agree where loitering matters: a gate, an entrance, a cash area, a car park corner.
- Set your limit.You choose how long is too long for each area. Near gates, Eye AI flags sooner.
- The on-site computer watches.Video is analysed on a computer at your site. It is not sent away to be analysed.
- You see the event.When someone stays past the limit, the event appears on your password-protected dashboard with the time, the place, how long they stayed, and a snapshot.
What you get
- Loitering events with the duration of each stay.
- Gate-intrusion warnings when someone lingers at an entrance.
- A live "in view now" list of who is currently in the area.
- A searchable history, so you can look back by day, hour or camera.
Honest limits
- This analytic is in beta. It has been demonstrated and tested on simulated movement. It has not yet been validated on large amounts of real footage from customer sites, so we publish no accuracy figure.
- A person who stays still is not necessarily doing anything wrong. Eye AI tells you who stayed; it does not judge why.
- People half-hidden behind others or objects can be missed or merged. Poor light, glare and long distances make small figures harder to see.
- Alerts to WhatsApp or email are a separate add-on that is launching. Time-of-day schedules, for example "alert only after hours", are not available yet.
Questions about loitering detection
Can Eye AI work on my existing CCTV cameras?
Yes, if your cameras or recorder provide a standard video stream on your network. The free camera-readiness check confirms this before you spend anything.
How long does someone have to stay before it is logged?
You decide. The limit is set per area, and areas near gates are flagged sooner.
Does it know if the person is a threat?
No. It reports that someone stayed past your limit and shows a snapshot. A person on your team decides what it means.
Is the footage sent to the cloud?
No. Video is analysed on a computer at your site and the dashboard is on your network.
Is it proven?
It is in beta. It works today and has been tested on simulated movement. Real-footage validation is ongoing, which is why we describe the limits openly.
Tell us about your site.
We read every request and reply within 24–48 hours (working days, Monday to Friday) with questions and a proposed solution.
Request a solution