Skip to content
IRIS NEURALContact

Solutions by typeInfrastructure integrity

Who sees a fallen tree across the road at five in the morning?

The tree is across the road and nobody knows

Ten past five in the morning, after a windy night. On the road out of the village a tree has come down and covers both lanes with all its branches. Loose branches lie across the asphalt. The camera on the pole has been watching it since the first minute. The first car to arrive from that side is doing ninety, and the bend hides the view until twenty metres out.

11 cameras across 4 sectors

This is how you ask for it

«Tell me if a tree comes down across the road, and how many lanes are blocked.»

Nobody calls until somebody meets the tree head-on. With luck, that person brakes in time and phones it in; without luck, there is a crash and now the lane is blocked with a car in it. And the same scene has other versions: a tree jammed across a canal with water backing up behind it, or a heavy branch resting on a railway track. In all of them, the time between the fall and someone finding out is time in which only luck is working.

You decide once

You decide it once and it applies across the whole camera network: what counts as an obstruction, on which stretches it matters — a straight with clear sightlines is not the exit of a bend — and who is told depending on the hour, because five in the morning is not the same shift as midday. IRIS applies it on windy nights and on calm ones, on every camera at once, with nobody having to remember to look harder when a storm warning is issued.

A fallen tree is not a hard alert to understand: either it is on the carriageway or it is not. What is hard is that it happens exactly where nobody is watching, and almost always at night or in a storm — when the road is at its most dangerous and the control room has far more cameras switched on than eyes available.

And it is not only the road. The same thing that blocks two lanes also jams across a canal from wall to wall, and then the water stops behind it and starts rising where it should not. Or it comes to rest on a railway track. The consequence changes; what does not change is that someone has to go and remove it, and that the later it is known, the more it costs.

IRIS watches the carriageway, or the canal, and sees that something large is there that was not there before and is not moving. It does not need to know what species the tree is: it needs to know that it blocks the way, how many lanes, and since when. With that, the alert arrives with the camera open, and whoever receives it can already decide whether to send signage, a machine or a patrol. On a windy night that may be ten alerts in an hour — and that is precisely the night when one person cannot walk three hundred cameras.

And there is a second half that almost nobody talks about: when it can reopen. Once it is cleared, somebody has to say the way is free, and until now that was a radio call from someone already on their way to the next incident. The same camera that saw the tree come down sees the empty lane, and whoever is in the control room checks it without sending anyone to look. On a night with six open incidents, that is one hour less of closed road.

What changes

Before, the first news of a fallen tree was a call from someone who had already met it. Now the first news is the camera, and it arrives while the road is still empty. What gets dispatched changes: instead of an ambulance and a recovery truck, a sign and a chainsaw. And what can be explained afterwards changes too, because the time it fell and the time it took to clear are both on record.

Where we have seen it

  • A tree has come down from its pit and its branches now cover both lanes. Cars are stopped on either side and loose branches are scattered over the tarmac.

    CAM-33Fallen tree blocking the road

    «Tell me if a tree falls onto the road, and how many lanes are blocked.»

    See this cameraThe city

  • A whole tree, torn up by the roots, has come to rest across the canal from wall to wall: the crown against one bank, the root ball wedged into the other. Behind it the water has gone still and dark; in front it forces through and foams. Two workers in helmets are setting up a barrier and cones to close the walkway, while a group watches from the railing opposite. IRIS sees that the flow section is blocked and how much the water has risen upstream.

    CAM-176A tree jammed across the canal

    «Tell me if something big gets jammed across the canal and the water starts backing up behind it.»

    See this cameraCritical infrastructure

  • Mid-morning at the intake channel. A fallen tree lies across the channel and its branches are pressed against the rack bays, mixed with bottles and plastics carried down by the water; only the section on the right is still clear. Upstream the water is still, with nothing moving. A crew member in a vest and white helmet is leaning over from the walkway, and up on the road a utility truck with a boom is parked.

    CAM-403A tree has come down the channel and is blocking the intake

    «Tell me if something large lodges across the racks at the intake.»

    See this cameraIndustry

  • After the storm. A pine has come down whole inside the enclosure: the trunk crosses the yard diagonally and rests beside the transformer, its splintered base against the fence. The crown has swept the far side of the perimeter and there are branches all over the gravel. Two panels of blue fencing are bent and one pole is lying down. Nobody has gone in yet. IRIS shows the site before and after the fall, with the exact time of the impact.

    CAM-351A whole pine tree down inside the substation

    «Alert me right away if a tree or a pole comes down inside the substation, and show me the before and after.»

    See this cameraIndustry

  • A narrow street with the roadway under water and, in the middle of it, the gully vanished beneath a heap of branches, cardboard boxes and bags. The water swirls beside it looking for a way through, and the parked cars already have it around their wheels. IRIS sees that the grate is covered and how much of it is still clear; it does not watch people and it does not read any number plate.

    CAM-167The street gully choked with branches

    «Tell me when a gully gets covered and stops taking water, before the street floods and a resident has to phone me about it.»

    See this cameraCritical infrastructure

  • Summer midday at a neighbourhood substation, houses all around. Scrub has climbed the wall and creepers hang over the yard. From the tree on the left, a long branch has pushed inside the enclosure: it crosses above the transformer and its twigs reach the insulators and the busbar. A few months ago that gap was clear. IRIS compares today’s image with the usual one, from the same camera and the same framing, and sees where the vegetation has come in.

    CAM-350A branch touching the substation busbar

    «Tell me if vegetation grows inside the yard or if a branch gets close to the busbar.»

    See this cameraIndustry

  • The mouth of the spillway seen from above. Against the screens of the bays on the right there is a raft of logs, branches and a bit of plastic, crossed over one another and resting on the concrete; underneath, white water is still coming through. The bay next to it has a clear screen and the water goes in unobstructed. On the walkway above you can see the gate hoists. Nobody is in shot.

    CAM-146Logs piling up against the spillway

    «Tell me when wood or rubbish builds up against the spillway screens, and tell me whether the pile is still growing.»

    See this cameraCritical infrastructure

  • The same substation, nobody inside, with tall weeds everywhere: between the transformer and the disconnectors, against the fence and all over the gravel. You can barely make out the walkway any more. This did not happen today: it has been growing for weeks, a little each day, and nobody has seen it because nobody has come.

    CAM-161Weeds have taken over the yard

    «Tell me when the vegetation starts taking over the yard, without waiting for the inspection: compare it with what was there a month ago.»

    See this cameraCritical infrastructure

What it does not do

IRIS does not know whether the tree is intact or whether another one is about to come down beside it. It does not measure wind, predicts nothing and does not say whether the embankment is unstable: whoever arrives judges that.

And it only sees what the camera sees. A tree down on a stretch with no coverage is still seen by nobody, and in thick fog or very heavy rain the image degrades just as anyone’s eyesight would.

And it cannot say whether what is on the carriageway is a tree, a pole or a load that fell off a lorry. It says there is something large blocking the way and not moving, which is exactly what you need in order to set out. The species is added by whoever arrives, and it does not change what gets dispatched.

Questions

Can it tell a fallen tree from a shadow or a branch swaying in the wind?
Yes, and the key is time. A shadow changes with the light and a swaying branch returns to where it was; a fallen tree lies still on the carriageway and is in exactly the same place a minute later. That is why the rule does not fire on what appears, but on what appears and stays. That detail is what separates an alert people actually look at from one that gets muted on the first windy night — the very night it was needed most.
Can it tell me how many lanes are blocked?
Yes, if the camera sees the whole carriageway. You mark the lanes on the image once, and IRIS says which of them has something on it. That is the difference between «there is a tree on the road» and «there is a tree across both uphill lanes», which is exactly what decides whether the road is closed or traffic is routed past. If the camera only sees half the width, the alert says half the width: nothing that is not in the image is invented.
Does it work for a canal or a railway track, or only for roads?
It works anywhere there is a way that has to stay clear. In a canal the problem is not the impact: it is that water stops behind it and starts rising, and that shows in the image long before it reaches anywhere. On a railway track, what matters is that nothing can be there when a train comes. What changes is the area you draw and who is told; what is being watched is the same: something large blocking a way and not moving.
On a stormy night many come down. Does the control room not get flooded?
It is a reasonable worry, and it is solved by deciding in advance. On a night like that what you need is not one alert per event but an ordered list: which stretches are blocked, how many lanes and since when. That lets you dispatch by severity instead of by the order calls came in. You can also mark in advance the stretches where a blockage is most serious, so they rise to the top. The bad night is exactly what this is installed for.

See all the questions

See it for yourself

This isn’t a video. It’s a slice of the real screen.

A summary of the real interface, played hands-free. You see one case from start to finish; what you do not see are the other screens, and there are many.

  1. Four cameras at once
  2. Each understands its own view
  3. The numbers climb by themselves
IRIS NEURAL
30ES
IRISDirectoGrabacionesGISIncidencias3996SituaciónCasosLPRAutomatizacionesIRIS DATA
Ask a question or make a request…Send
IRIS · Infinity Neural

Not seeing yours here?

These words are the ones we have built, not the ones IRIS understands. The list falls short on purpose: you ask for it in a sentence, and sentences do not run out. Tell us what is in front of your cameras and we will tell you whether it understands it, or not yet.

See all solutionsTell us what you see

And the best part

Nothing changes on the wall.On the screen, everything changes.

Same brackets, same cabling, same cameras. The only difference is that those images are now understood as they come in, instead of waiting for someone to open them.

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

We reply the same working day.