CAM-929
Getting worse for weeks
A square in the old town at mid-morning, seen from a camera up high. The stone fountain with its statue in the middle, the balconies and shutters all around, the café terrace with its white parasol, the bins beside the planter and seven or eight people crossing. You can see all of it. And here is the odd thing about this one: if this were the only image you had, you would say the camera is fine. A little flat, a little washed out, the corners slightly dark. Nothing that justifies climbing up to clean it. The fault is not in this image: it appears when you put it next to the one from a month ago. IRIS is not seeing a breakdown. It is seeing a slope. What it does not know is the cause: dome, seal or focus is for whoever goes up.
Turn on IRIS Vision
AI-generated image
What IRIS understands
What it recognises in this scene
- worse every week
The rule
You decide once. IRIS applies it always.
If
a camera’s image quality drops day after day for as many consecutive days as the site has set, without any single image looking faulty
Then
- alert maintenance with the camera, the time and today’s video next to the first one in the series
- attach the line of the last few days, so the drop is visible and not just today’s picture
- put it on the list for the next maintenance visit and close the alert only when the image is back at its starting level
The instruction that set it up«Show me the cameras that see worse today than a month ago, even if none of them looks broken, and sort them by how far they have dropped.»
What changes
What changes for whoever answers for the cameras working
No camera fails all at once. The dome slowly hazes over, the seal starts letting damp through, the focus drifts a little. Every week it sees slightly worse than the week before, and no single week is bad enough for anyone to notice. Months go by like that. On the day the video is needed, the video is worth nothing, and only then does anyone discover it had been sliding for half a year. What IRIS does is store, every day, how this camera looks, and show you the line. The failure has not happened yet: somebody can go up next Tuesday, with a ladder and a cloth and no urgency. It is the one of the thirty-four you fix before it breaks.
What people usually ask
If it gets worse bit by bit, when exactly does it raise the alert?
When the rule you wrote says so, not us. IRIS stores a measurement of each camera’s image every day and draws the line. You decide how far it has to fall, and over how many consecutive days, before the alert fires. On the camera in this card, the site set fourteen days of decline. A camera watching a square can tolerate a slow slide; one reading an entrance cannot. Setting the same number for all of them would mean making it up.
And if the picture drops because of the weather and not the camera?
It happens, which is why you never look at a single image: you look at the line and at the neighbouring cameras. A week of fog or haze pulls quality down on every camera in the area at once, and it climbs back when the weather clears. A camera that drops on its own while the ones next to it stay flat, and that does not recover on any day, is not the weather. When IRIS is not sure, it says so: it raises it as something to look at, not as a fault.
Do you have to keep weeks of video to be able to compare?
No. Comparing does not need the whole recording: a daily measurement of the image and one reference frame a day are enough. That takes up very little and can be kept for months without touching your video retention policy, which stays exactly as you set it. The video is kept for as long as your rules say; the quality line lives separately and survives even once the video has been deleted.
The real interface, step by step
Everything comes in. Only what matters gets through.
This sums up what IRIS does with one alert, from the moment it appears until it is closed. The real screen holds far more; this is the main thread.
- The rule gets built
- It goes to every camera
- Published, it watches alone
- And then it fires
Next step
Who sees the child who has fallen at the far end of the playground?
13 cases: 12 raise an alert when something happens and one only measures. You see each image as it is, and then with what IRIS understands on top.
Who is watching a site where there is almost never anybody?
44 cases, and in every one of them IRIS raises an alert when something happens. You see each image as it is, and then with what it understands on top.
And the best part
If the camera does not show it,IRIS does not see it. We say so up front, not afterwards.
No camera sees through a wall or in complete darkness. We would rather show you where the limit is, and place what you have well, than promise something that never appears on screen.
Tell us your problem and we will say whether IRIS understands it, or not yet.
We reply the same working day.