Intruder detection · Senstar
If the fence shakes in the wind, the fence is the problem.
Senstar makes perimeter protection for large sites: sensor cable on the fence, buried cable, fibre, taut wire and microwave. Here the sensor is the fence itself or the ground itself, so the result depends on the civil works as much as on the equipment.
What Senstar is and in what sites it turns up.
A Canadian perimeter manufacturer for large sites: sensor cable on the fence, buried cable, optical fibre, taut wire and microwave. And its own video software, which in its world makes sense: a two-kilometre perimeter with no image cannot be operated.
It turns up where the perimeter is long and the response costs money: industrial plants, substations, depots, data centres, ports, airports, quarries. What they share is not the size: it is that a patrol takes time to get there and you have to know which point to go to.
And that is what really sells this family of technologies: not just that something has happened, but where. Knowing the fence has been touched at metre four hundred of two thousand changes the response, because there you can turn a camera and look before moving anybody.
What each technology really detects.
Sensor cable on a fence measures the fence: it detects the vibrations of somebody climbing, cutting or lifting the mesh, and tells them apart —that is the fine work— from the vibrations of the wind. The consequence has to be said before signing: the sensor is the cable plus the fence, and a slack fence is a bad sensor.
Buried cable cannot be seen, and that is half its value: it detects whoever walks over it, in a narrow corridor with nothing on show. It is useful where it must not be obvious that there is detection, it follows the ground with its slopes, and light and fog make no difference to it.
Fibre detects somebody touching the cable by the change in the light travelling inside it: it carries no voltage along the perimeter and covers a long distance with little equipment at the ends. Taut wire is the one that gives the fewest false alerts, in exchange for civil works. And microwave covers the stretches where there is no fence.
Where each one falls down. This is the bit to read.
Here almost every problem is one of civil works and ground, not of electronics.
- Cable on a fence fails with the fence: loose mesh and it gives alerts with every gust; a sign or a tarpaulin tied on acts as a sail; a gate that does not shut tight bangs all night; and a hedge up against the fencing rubs it the moment there is a breeze.
- And it fails with what goes on on the other side: fencing that gives onto a road with heavy traffic or onto a manoeuvring area picks up vibration through the ground. You work on it by tuning stretch by stretch, but you have to know it on the day of the project.
- The buried one suffers with water and ice: ground saturated after days of rain does not behave like dry ground, and frozen or snow-covered soil changes what the cable feels. In climates with a real winter, that means a seasonal inspection.
- The buried one detects mass going past, so a wild boar or a loose dog go past too: what you adjust is the threshold, and you pay for it in sensitivity. And it cannot be installed where there are going to be works, where there is a lot of buried metal or where heavy vehicles pass alongside.
- Fibre and cable on a fence share a blind spot: doors, gates and corners. A door does not vibrate like a fence and has to be its own zone. Most perimeters that give trouble give it at the doors.
- And they all share the most expensive mistake: the calibration that was done once. It has to be readjusted every time a piece of fencing is repaired or a hedge is trimmed. A perimeter calibrated in September and never looked at again is not, by March, the one that was handed over.
The installation is eighty per cent of the result.
With these technologies that is not a phrase: the sensor is literally the civil works.
The cable is fixed with a set spacing and a set tension, and both are a technical decision, not a knack. Slack, it does not carry the vibration it is meant to detect; tight, it carries all of it, the wind included. And it goes on to the same criteria along the whole stretch, because the system learns what is normal there.
Before that the fence has to be repaired, and that is not an extra on the quotation: missing ties, loose mesh, posts that move, a flattened stretch. If it is not repaired, what you are calibrating is a system that tolerates the noise of a broken fence — and tolerating noise is ceasing to detect.
Zoning is what changes the response most: short zones let you turn a camera to the exact spot; long zones come out cheaper and leave a patrol walking two hundred metres with a torch. And the buried one is a trench: constant depth and drainage, documented with photographs before it is covered, because a deeper stretch detects less and that is never seen again.
Verifying: what turns an alert into a decision.
When a fence zone raises an alert, the only thing you know is that the zone has moved: somebody climbing, the wind on a stretch that has worked loose, a dog. And the two options for whoever is on duty are both bad: send a patrol to every alert, which soon stops happening, or send nobody and hope.
With an image, the system says which zone, the camera goes to that point and the operator looks. That is what turns a long perimeter into something one person can operate at night, which is exactly the problem that has to be solved.
And here something has to be declared: in the same house there is IRIS Neural, the NVMS from Infinity Neural, the other company in this group. It is our own product, and we say so because recommending your own without warning wrecks everything else. It adds the image of the stretch that generated the alert; it does not replace the sensor or the panel. With other video software it is done too, including this manufacturer's. With nothing that lets you look, it is not.
There is no good technology. There is technology suited to that perimeter.
A real perimeter is not a line: it is fence, wall, a lorry gate, an embankment and a gap where the fence does not close because a pipe goes through. Covering them all with the same answer is what produces perimeters that go off five times a night on two particular stretches.
The typical split on a large site: cable on the fence along the long, sound stretches, microwave in the gaps, a sensor and a contact of their own on the doors, buried cable where nothing can be seen, and cameras with verification over the top of all of it. Two technologies that fail for different reasons, holding each other to account on the difficult stretches, is what makes the perimeter credible.
We are not going to say this brand is better than the other four on the list; we will say what a catalogue that wide is for: when a site calls for four technologies, solving them with one manufacturer saves integrations and spares. And the wholly honest part: on a large perimeter what decides most is not the brand, it is the state of the fence, the drainage and somebody coming back to recalibrate after the winter.
What we do with it.
- Walking the perimeter on foot, all of it, before looking at any catalogue: the state of the fence metre by metre, doors, corners, embankments, where it floods and what is on the other side.
- The fencing condition report, kept separate from the detection quotation. If the fence has to be repaired, it goes in with its own figure: it is the client's money and they have a right to see the two items apart.
- The zoning drawn out, with the doors as their own zone, the overlaps marked and which camera looks at each zone — decided on the same drawing.
- The civil works: fixing with the specified spacing and tension, the trench documented with photographs before it is covered, watertight ducting, backup power and tamper detection on every box.
- The calibration stretch by stretch, walking and climbing, against what really happens on that stretch and not against a catalogue value. With wind if you can wait for a windy day: it is the best calibration there is.
- And the seasonal adjustment in the contract, plus recalibration every time a stretch of fence is repaired or a door is changed. If that is not written down, it does not get done — and it is what separates a perimeter that is alive after three years from one nobody looks at.
The grade, the alarm receiving centre and what comes after.
A perimeter sensor is a zone on a panel. The EN 50131 standards classify systems into security grades, and each grade demands specific things of the equipment, the cabling, the power supply and the tamper detection. Which grade each piece of equipment supports is read in its documentation; whether the installation meets that grade is another matter. We do not say here which one we hand over: it depends on the project, and it is stated in writing by whoever signs.
Connecting to an alarm receiving centre is a separate decision, with its own regulations and its own contract. On a large site the protocol also has to be written down zone by zone: what is done with the stretch that gives onto the road, what with the lorry gate, and what gets verified before moving anybody. Who provides that service is settled in a quotation, not on a website.
The day to day is the client's: arming by zone, looking at the log and knowing which stretch is which number. What they do not touch is the calibration or the threshold, because turning it down so it stops being a nuisance is exactly how a stretch stops detecting without anybody noticing. And the isolation that got left in place is the way in: each one with an expiry date and with that date in the report.
The other eight in this group.
Four more perimeter houses and the four control panels all of this connects to. None is better than another: what changes is the ground they hold up on.
Get started
Before the quotation, the state of your fence.
Tell us how many metres of perimeter there are, what condition the fencing is in, how many gates it has, what the ground is like and how many alerts come in each month. We will tell you which technology each stretch calls for, what has to be repaired first and what can wait.