What is a perimeter intrusion detection system?

Any system that detects someone crossing a defined boundary and reports it. The main types are fence-mounted vibration sensors, buried seismic cable, infrared beams, microwave or radar, and video analytics. They differ mainly in what they can tell you afterwards: physical sensors report that something happened at a point, video analytics reports what it was and gives you a clip.

What does a perimeter intrusion detection system cost?

The dominant factor is whether power and network already run along the perimeter. Where they do, cost scales with the number of detection points. Where they do not, most approaches require trenching the length of the boundary, and that civil work routinely exceeds the equipment cost. Solar and cellular units avoid the trench, which is why they change the arithmetic on unpowered perimeters.

Which perimeter detection technology is most reliable?

Buried seismic cable is exceptionally reliable and invisible — and needs a trench along the whole perimeter. The right question is not which detects best in isolation, but which is best given your ground, your budget for civil work, and whether you need to know what crossed rather than only that something did.

Do I need to cover the entire perimeter?

Almost certainly not, and trying to is what makes these projects unaffordable. Perimeters are crossed at specific points — where there is cover, where a vehicle can stop unseen, where the fence is already damaged. Cover those first, run it for a month, and let the events tell you what phase two should be.

Will it alert on animals and weather?

Motion-based systems will, constantly. Classification is what fixes it: detecting a person rather than movement removes most nuisance alerts, because on an empty perimeter at 3am there usually is no person. Some residual false alerts are unavoidable, and any vendor claiming zero is describing a demo rather than a fence line.

Can it work on a perimeter with no power?

Yes — that is the common case rather than the exception. Solar units with cellular connectivity run without any site infrastructure, and analysis happens on the unit so only events and short clips travel over the cellular link. The practical limits are line of sight and correct battery sizing for winter.

Use case

Perimeter intrusion detection: what works and what it costs

A perimeter is long, mostly empty, and almost never has power along it. That combination is why most perimeter projects are priced by the trench rather than by the sensor.

August 17, 2026VuePoint

Perimeter intrusion detection is the problem of knowing when someone has crossed a boundary they should not have crossed, in time for that knowledge to be useful. It sounds like a sensor question. In practice it is a geometry and infrastructure question, and the sensor is close to the last decision you make.

The difficulty is that a perimeter has three properties at once: it is long, it is almost entirely uneventful, and it usually has no power or network anywhere along it. Any approach that ignores one of those three fails in a predictable way.

The approaches, and how each one fails

Approach	What it detects	How it fails
Fence-mounted vibration sensors	Someone climbing or cutting the fence itself	Wind, heavy rain, passing lorries, and a loose fence panel. Needs a fence in good condition to work at all
Buried cable / seismic	Someone walking over a buried detection line	Excellent detection, invisible installation — and it needs a trench along the whole perimeter, which is the cost
Infrared beams	Something breaking a line between two posts	Animals, fog, snow, and drifting alignment. Cheap and genuinely useful over short straight runs
Microwave / radar	Movement within a volume, at long range	Very long range and weather-tolerant; needs clear ground and careful setup, and reports movement rather than what moved
Video analytics	What the thing is — person, vehicle, animal — and where it went	Bounded by line of sight, lighting and pixels on target. Cannot see through a building or round a corner

There is no bad option here — there are options matched or mismatched to a site. The failure mode is the useful column.

Why perimeter projects cost what they cost

Ask what a perimeter intrusion detection system costs and the honest answer is another question: does power reach the fence line? That single fact moves the total more than the choice of technology.

Where power and network already run the length of the perimeter, cost scales roughly with the number of detection points, and the sensor choice is a genuine engineering comparison.
Where they do not, most approaches need a trench along the perimeter to work, and that civil work routinely exceeds the equipment cost — before permits and reinstatement.
Perimeters are long, so trenching is not a one-off cost at one point. It scales with distance, which is what makes it dominate.

This is why solar and cellular units change the arithmetic here more than almost anywhere else. They do not detect better than a buried cable — a buried cable is superb. They detect well enough while removing the trench entirely, and on an unpowered perimeter the trench was the price. The full cost breakdown is here.

You do not need to cover the whole perimeter

This is the most useful thing on this page, and it is the step most projects skip. The instinct is to protect the entire boundary uniformly. Almost no site needs that, and pursuing it is what makes perimeter projects unaffordable.

How to scope a perimeter properly

1
Find where people actually cross
Perimeters are not crossed at random. They are crossed where there is cover, where a vehicle can pull up unseen, where the fence is already damaged, and where the ground outside is easy to reach. Walk it and look for the worn patch — there usually is one, and it is more informative than any survey.
2
Work out what is worth reaching
People cross a boundary to get to something: fuel, copper, tools, equipment, a vehicle compound. Coverage that watches the approach to the valuable thing outperforms coverage spread evenly along a fence that mostly borders nothing.
3
Separate detection from evidence
The camera that sees the whole approach is rarely the one that identifies a vehicle or a face. Decide which positions are for knowing and which are for proving. Sites that get useful outcomes have both, placed deliberately.
4
Decide what happens when it fires
An alert with no defined response is a notification nobody reads. Write down who receives it, at what hours, and what they do. This determines how sensitive the system should be — and it is the step that most changes whether the deployment is still in use in six months.
5
Start with the two or three worst points
Cover those, run it for a month, and look at what the events tell you. Perimeter behaviour is far more concentrated than people expect, and a month of real data is a better guide to phase two than any amount of planning.

What makes video analytics workable on a perimeter

The nuisance problem above is exactly what classification solves. Detecting motion on a fence line at 3am produces alerts every night. Detecting a person on a fence line at 3am produces very few, because there usually is not one.

That said, the reliability is bounded by physics, not by the model. Detection quality depends on how many pixels land on the target, which means a camera covering 300 metres of fence from one position will not do this well no matter what is running on it. Realistic coverage is a sequence of positions each watching a manageable span, which is another argument for units that can be placed where they are needed rather than where the cable reaches.

The related failure modes at night — IR illuminators, insects, rain lit by the illuminator — are covered in fixing night-time false alerts, and they matter more on a perimeter than anywhere else, because a perimeter is dark by definition.

What this does not do

It does not stop anyone. It detects a crossing and documents it. What happens next is a decision made by people on your side.
It does not replace a fence. A boundary that is trivial to cross stays trivial to cross; you will simply know about it.
It does not identify individuals. It classifies a person as a person, not as a named one.
It does not work where it cannot see. Line of sight is a hard constraint, and a building, a stack of materials or a hedge is the end of the conversation for that position.

Common questions

What is a perimeter intrusion detection system?: Any system that detects someone crossing a defined boundary and reports it. The main types are fence-mounted vibration sensors, buried seismic cable, infrared beams, microwave or radar, and video analytics. They differ mainly in what they can tell you afterwards: physical sensors report that something happened at a point, video analytics reports what it was and gives you a clip.
What does a perimeter intrusion detection system cost?: The dominant factor is whether power and network already run along the perimeter. Where they do, cost scales with the number of detection points. Where they do not, most approaches require trenching the length of the boundary, and that civil work routinely exceeds the equipment cost. Solar and cellular units avoid the trench, which is why they change the arithmetic on unpowered perimeters.
Which perimeter detection technology is most reliable?: Buried seismic cable is exceptionally reliable and invisible — and needs a trench along the whole perimeter. The right question is not which detects best in isolation, but which is best given your ground, your budget for civil work, and whether you need to know what crossed rather than only that something did.
Do I need to cover the entire perimeter?: Almost certainly not, and trying to is what makes these projects unaffordable. Perimeters are crossed at specific points — where there is cover, where a vehicle can stop unseen, where the fence is already damaged. Cover those first, run it for a month, and let the events tell you what phase two should be.
Will it alert on animals and weather?: Motion-based systems will, constantly. Classification is what fixes it: detecting a person rather than movement removes most nuisance alerts, because on an empty perimeter at 3am there usually is no person. Some residual false alerts are unavoidable, and any vendor claiming zero is describing a demo rather than a fence line.
Can it work on a perimeter with no power?: Yes — that is the common case rather than the exception. Solar units with cellular connectivity run without any site infrastructure, and analysis happens on the unit so only events and short clips travel over the cellular link. The practical limits are line of sight and correct battery sizing for winter.

Keep reading

Got a boundary that keeps getting crossed?

Tell us the length, where the power is, and where you think people are coming over. We will tell you what is coverable and what it would take.

Request a quote