CAM-902
Pointed at the wall
The same street, except the camera no longer sees it. Three quarters of the frame is the façade of the building opposite: the ochre wall, an iron balcony with terracotta pots, the stone mouldings, a shop’s yellow awning and the arched doorway. The street it was supposed to watch has been squeezed into a narrow strip down the right-hand side, where a tree, a piece of pavement and the end of a crossing just show. The picture is good: in focus, well exposed, recording without a break. The only trouble is that there is nothing inside it to watch. IRIS sees that the field of view is now a wall. Who turned it and why is told by the video, not by IRIS.
Turn on IRIS Vision
AI-generated image
What IRIS understands
What it recognises in this scene
- wall filling the frame
The rule
You decide once. IRIS applies it always.
If
a camera’s field of view becomes a flat, motionless surface and the scene it was watching no longer appears
Then
- alert maintenance with the camera, the time and the video of the moment it turned
- mark the area that camera covered as unwatched, so nobody assumes it is being seen
- log the time at which it stopped seeing the scene
The instruction that set it up«Tell me if a camera stops seeing the street and ends up staring at a wall, a ceiling or the ground.»
What changes
What changes for whoever answers for the cameras working
A camera turned to the wall is the quietest fault there is, because technically nothing is faulty. It gives signal, it gives video, it writes to disk and the status panel shows it green. The whole installation believes that junction is covered. It is not. And since nobody goes through forty thumbnails one by one every morning, it can stay that way until somebody needs one specific clip: a motorbike that knocked over a pedestrian, on a Tuesday, at half past seven. That day the archive opens and holds three weeks of a balcony with geraniums. With IRIS somebody finds out that same morning, with the before and after image, and a ladder fixes it in ten minutes. What you get back is not the camera: it is the three weeks.
What people usually ask
How does it tell a wall apart from an empty street at four in the morning?
By what is there, not by what moves. An empty street at night still has a roadway, a kerb, pavements, street lights and a background at a distance. A façade two metres away has none of that: it is flat, it is close and it has no depth. Above all, nothing that used to appear appears any more. The comparison is made against the framing that this particular camera had saved as good, so the hour and the traffic make no difference.
If it is still recording, why is this so serious?
Because the installation believes that area is covered and it is not. A status panel sees signal, sees video and sees the disk being written, so it marks the camera as working and nobody goes out to check. Meanwhile, the place that camera was watching is being watched by nobody. And that is not discovered on the day it happens, but weeks later, when somebody asks for the video of one particular moment and it turns out not to exist. Recording and watching are not the same thing.
What about cameras that are meant to look at a wall?
They exist and they are perfectly normal: one covering a door a metre away, one reading a panel, one watching a machine that is standing still. That is why the check is not against a general idea of «this is a wall», but against the framing that this particular camera had saved as good. If that camera has always looked at a door, that door is its reference and it will never raise anything. It only raises something when what it sees stops being what it saw.
The real interface, step by step
We won't just tell you. We'll show you.
It is a summary of the real screen: it plays by itself, step by step, and shows only the path of one case. The actual tool has far more than fits in here.
- It lands in the tray
- You open it
- You confirm or dismiss
Next step
How long does it take to find a hand played two hours ago?
12 cases: 9 raise an alert when something happens and 3 only measure. You see each image as it is, and then with what IRIS understands on top.
What does a shopping centre see beyond the visitor count?
12 cases: 4 raise an alert when something happens and 8 only measure. You see each image as it is, and then with what IRIS understands on top.
And the best part
You already record.The hard part was always watching it.
Years of video that only get opened once something has gone wrong. IRIS understands those same images as they arrive, so the answer no longer lives only in the archive.
Tell us your problem and we will say whether IRIS understands it, or not yet.
We reply the same working day.