Skip to content
IRIS NEURALContact
Choosing a language opens this same page in that language.

Camera health

CAM-917

Dust and sand

A seaside promenade in mid-morning, seen from the lamp post. The sea and the wall on the left, the palm trees and the people walking in the middle, the road with parked cars and the café awnings on the right. It is all there, but as if seen through a dirty window: the whole frame is covered by a brown veil that dulls the blue of the water and the green of the palms and leaves tarmac and pavement the same sand colour. There is no single stain: the dirt is spread evenly. IRIS sees that this camera has spent days giving less detail and less colour than it used to. What it does not know is whether the dust is in the air or stuck to the glass: how long it lasts is what tells you.

A seaside promenade in mid-morning, seen from the lamp post. The sea and the wall on the left, the palm trees and the people walking in the middle, the road with parked cars and the café awnings on the right. It is all there, but as if seen through a dirty window: the whole frame is covered by a brown veil that dulls the blue of the water and the green of the palms and leaves tarmac and pavement the same sand colour. There is no single stain: the dirt is spread evenly. IRIS sees that this camera has spent days giving less detail and less colour than it used to. What it does not know is whether the dust is in the air or stuck to the glass: how long it lasts is what tells you.
CAM-917IRIS VisionLive

Turn on IRIS Vision

AI-generated image

CAM-917 · OPTICA · POLVO EN SUSPENSION

What IRIS understands

What it recognises in this scene

  • dust veil

The rule

You decide once. IRIS applies it always.

If

a camera has spent several days in a row giving less detail and less colour than its own reference picture

Then

  1. alert maintenance with the camera, the time and the video of what it looks like now
  2. keep the day it started to fog over and how far it has dropped since
  3. put it on the cleaning list and take it off by itself when the picture returns to the reference
The instruction that set it up«Tell me which cameras are giving less detail than last week, and alert me if any of them has been like that for more than two days.»

What changes

What changes for whoever answers for the cameras working

A veiled camera is not broken. It keeps recording, it stays online and it shows up green on the list. That is why nobody touches it. The problem comes the day that video is needed — who left the bag on the pavement, which car clipped the lamp post — and all there is is a sand-coloured picture where nothing can be told apart. Today the dust comes off when somebody climbs up for another reason, or when the cleaning round reaches it on the calendar. With IRIS it stops being a calendar: there is a list of specific cameras, each with the day it started to fog over. The one that needs cleaning gets cleaned, and it is written down when it got dirty and when it could see again. Maintenance stops being a habit and becomes a fact.

What people usually ask

In a quarry every camera is dirty all the time. Is this not a permanent alert?

No, because it is not compared with a factory-clean camera: it is compared with itself. Every camera has its own reference, which is how it sees that spot when it has just been cleaned. In a quarry that reference already has dust in it, and that is fine. The alert fires when that particular camera gets worse than itself and does not recover on its own. The threshold is set by whoever runs the site: in a quarry you set it higher than in a lobby.

How does it tell dust on the glass from fog or haze?

By how long it lasts and how it spreads. Fog and haze roll in, last a few hours and the picture comes back on its own; dust stuck to the glass looks the same at night, the next day and in full sun. Fog also eats mostly into the background and leaves the foreground fairly clean, while dirt on the glass veils the whole frame evenly. That is why IRIS does not alert on the first bad hour: it waits to see whether the picture recovers by itself.

Should the cleaning contract not take care of this?

The contract says how often somebody goes up. It does not say which ones need it. Today they all get cleaned every so often: the ones that were fine and the ones that had been seeing nothing for weeks. IRIS does not clean any camera; what it does is give the list of which ones and since when, so the visit goes where it is needed. And it leaves the two dates side by side, the day it got dirty and the day it saw again, which is what lets you check whether the round arrives in time.

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.

  1. The hour, as a grid
  2. You look where it's dark
  3. You jump to that minute
IRIS NEURAL
30ES
IRISDirectoGrabacionesGISIncidencias3996SituaciónCasosLPRAutomatizacionesIRIS DATA
Ask a question or make a request…Send
IRIS · Infinity Neural

And the best part

You have cameras to spare.What you run out of is eyes to watch them.

One person can follow two or three screens. Nobody can follow forty. IRIS watches them all and speaks up when someone needs to decide.

Tell us your problem and we will say whether IRIS understands it, or not yet.

We reply the same working day.