CAM-86
People sleeping on the benches
Five people are asleep on the benches in the concourse, before the first train leaves. Their blankets and bags are beside them.
Turn on IRIS Vision
AI-generated image
What IRIS understands
What it recognises in this scene
- person resting
- concourse benches
- belongings
The rule
You decide once. IRIS applies it always.
If
a person has been lying on a bench in the concourse for a good part of the night
Duringthe small-hours window the station decides on
Then
- note the spot, the time and how many people there are in the morning report
- let the outreach team or the local social services know
- raise no security alert: this is not an incident
The instruction that set it up«Tell me each morning which benches people slept on and how many there are, so the outreach team knows where to go.»
What changes
What changes for whoever runs the station
In many cities there is a team that goes out at night to see who is sleeping rough and to offer them a bed, a blanket or a medical check. Today that team works from what it is told over the phone, and from a big station almost nothing useful reaches it: not how many people there are, not which part of the concourse, not at what time. And five people on the benches is not the same as one: you do not go with the same staff, the same blankets or the same amount of time. With IRIS the station supervisor has, first thing in the morning, a list of places, times and numbers he can pass on, and that team sets out knowing where to go and what it will find. The station changes nothing about what it does: the only change is that the information exists and reaches the people who can help in time.
What people usually ask
Is this for moving people out of the station?
No, and it is not installed for that. This camera sits in the looking-after-people family, not in incidents, and its rule raises no security alert: what it does is put in writing where and when someone is sleeping, so a team that can offer something knows where to go. If what you want is a system for clearing people out, this is not it — and the rule would have to be written differently anyway. It is not written here.
Does IRIS know who that person is?
No. IRIS sees that someone is lying on a bench and has been there for hours. It does not know who they are, why they are there, whether they want help or whether it is the same person as last night. It does not recognise faces, does not compare them against any database and stores nobody's identity. All of that is found out, if it needs to be, by a person who goes over, sits down beside them and talks to them.
What if it is a traveller who fell asleep waiting for the first train?
It comes up too, and rightly so. IRIS cannot tell those two situations apart, and it should not try: from the outside they are the same thing, someone asleep on a bench at five in the morning. The difference is drawn by the person who walks over and asks whether they need anything — and that question works just as well in either case.
Simulation · how IRIS works
You set the rule once. IRIS applies it every time.
You will see a summary of the interface, played step by step and hands-free. Each round starts with another case. The complete tool does not fit in a demo.
- Four cameras at once
- Each understands its own view
- The numbers climb by themselves
Next step
When does a bad review get decided?
15 cases: 11 raise an alert when something happens and 4 only measure. You see each image as it is, and then with what IRIS understands on top.
Who is watching a whole quay at three in the morning?
14 cases: 13 raise an alert when something happens and 1 only measure. You see each image as it is, and then with what IRIS understands on top.
And the best part
The cameras you bought years agostill hold up. What they were missing was not image quality.
It was someone to look. IRIS connects to the installation you already have — same equipment, same cabling — and those images go from being stored to being understood.
Tell us your problem and we will say whether IRIS understands it, or not yet.
We reply the same working day.