Types of intruder detection, and which space each suits

What types of intruder detection are there, and which suits which space?

Intruder detection is built in layers: sensors on the site perimeter, on the shell of the building, and inside it. Door and window contacts tell you an opening was used. Glass break and vibration sensors pick up an attack on the shell. Motion sensors cover rooms and corridors, and beams, fence sensors or camera analytics cover the outside. Which type suits a space depends on what the space is like. But on the sites we see, the biggest cause of false alarms is people, not sensors. In New Zealand the sensor is only half of it: NZ Police policy expects the monitoring company to respond first, so what happens after the alarm matters as much as what raised it.

Perimeter: outside the building

• Active infrared beams. A beam runs between two posts, and breaking it raises the alarm. Good along fence lines and building edges with a clear line of sight. Long grass, fog, birds and anything moving through the beam path cause false alarms.
• Outdoor microwave detectors and barriers. A detection field across open ground, less affected by weather than beams. They need clear ground: swaying trees, standing water and moving gates in the field set them off.
• Fence-mounted sensors. They detect cutting, climbing or lifting on the fence itself. Wind and loose panels are the usual cause of false alarms, so fence condition matters.
• Camera analytics and thermal cameras. More and more often the perimeter layer, because the device that detects the intruder also shows an operator what caused the alarm.

Shell: doors, windows and walls

• Magnetic contacts. Fitted to doors, windows, roller doors and hatches, they trigger when an opening opens. The most reliable sensor in most systems.
• Glass break sensors. They listen for the specific sound of breaking glass, so a space can stay armed while people move outside it. They need to sit within their rated range of the glass they cover.
• Vibration and seismic sensors. They detect drilling, cutting or impact on walls, safes, cages and cash machines. Good where an attack would come through the structure rather than a door.

Interior: rooms, aisles and corridors

• PIR motion sensors. They detect the heat of a body moving across their field. Inexpensive and dependable in a stable space. Heaters, air-conditioning outlets, direct sunlight and anything warm that moves cause false alarms.
• Dual-technology sensors. A PIR and a microwave sensor in one unit, which alarms only when both agree. The usual choice for warehouses, workshops and other spaces where a PIR alone would false alarm.
• Ultrasonic sensors. They detect movement through changes in high-frequency sound, and suit small enclosed spaces. Strong air movement can set them off.
• Security fog. A response, not a detector. When an intrusion is confirmed, a fog device fills the space with dense, harmless fog within seconds, so an intruder cannot see what to take. The European standard EN 50131-8 covers these devices. In 2022 the Government began subsidising fog cannons for small shops and dairies, at $4,000 a shop, after a run of ram raids. Best for high-value stock rooms. Check with your fire protection provider before installing, so the fog does not trigger smoke detection.

The standard and the grades

In New Zealand, intruder alarm systems are designed, installed, commissioned and maintained to AS/NZS 2201.1:2007, Intruder alarm systems, client’s premises. It sets minimum requirements for the detection devices, control equipment, warning devices and power supplies on site. Monitoring centres and the link from the premises to them are covered by other parts of the AS 2201 series.

Many systems are also graded to the European standard EN 50131, which rates a system from Grade 1 to Grade 4 by the risk it is designed for: Grade 1 for low risk, Grade 2 for low to medium, Grade 3 for medium to high, and Grade 4 for high risk. The grade sets how hard each part of the system is to defeat. From Grade 3, for example, a motion sensor has to report that someone has tried to cover or mask it. Ask a supplier which grade a design meets, and check it matches what your insurer expects.

Medifab’s alarm, for example, is an Ajax hybrid system specified to Grade 3 throughout: the controller and every connected device.

What drives the cost over five years

The cost of an intruder alarm, at installation and in the years after, sits in these:

• The grade. A Grade 3 design needs sensors that resist masking and tampering, with a control panel and communications to match. It costs more than Grade 2, and it is worth it where the risk, or your insurer, calls for it.
• Wired or wireless. In a new building, cable is cheap to run before the walls close. In an existing one, wireless sensors avoid cabling and disruption, but every one has a battery to replace on a schedule, which turns into a maintenance cost.
• The perimeter. Outdoor detection costs more for the area it covers than anything inside, and it needs the most tuning to keep false alarms down. Decide first whether you need to know when someone reaches the fence or when they reach the building.
• How the alarm reports. A monitored alarm needs a communicator to send signals to the monitoring centre, ideally over two paths, such as cellular and internet, so cutting one does not silence it. New Zealand’s mobile carriers have been retiring their 3G networks, and a communicator that depends on 3G stops reporting when its network goes.
• Monitoring and response. Monitoring is a monthly fee for as long as the alarm is in use, and guard attendance is usually charged per call-out. False alarms are what makes that cost grow.
• Maintenance. Servicing and testing, battery replacement, and re-tuning after changes on site. A system nobody services drifts into false alarms.
• Room to grow. Every control panel has a limit on zones and users. A panel near its limit makes one more door the trigger for a new panel.

When you compare two quotes, line them up on these, not on the number of sensors. Our article on comparing security quotes shows how to put a five-year number on each.

Where false alarms come from, and why it matters

NZ Police policy is clear about who goes first. In most cases, the private security company that monitors an alarm should be the first to respond, and Police respond when they are told of a burglary or suspicious activity. Police also weigh a site’s history of false activations when deciding how to respond. A system that false alarms often costs you twice: in call-out fees, and in how the next alarm from your site is treated.

On the sites we see, the causes are, most common first:

• People. Users who were never shown how the system works, or who forget to arm or disarm it. It is the biggest single cause, and it is fixed with training, not hardware.
• Ageing or poorly installed equipment. Worn sensors, loose contacts, flat batteries and poor original installation raise false alarms and fault alerts.
• Systems not set up for the site. Sensitivity and zones that alert on birds and other animals, a PIR near a heater or a roller door, or racking moved in front of a sensor during a fit-out.

Two fixes do the most. Train everyone who sets the alarm, and train again when staff change. Then add verification: an alarm confirmed by a second sensor, or by a camera an operator can look at, before anyone is sent.

Walk the site with the zone list

1. Walk the site with the zone list and check that every opening has a contact, including roller doors and roof hatches.
2. Ask your monitoring company for the last 12 months of alarm activations, and sort them by zone, cause and time of day. False alarms around opening and closing times usually point to people, not sensors.
3. Check what changed in those zones: racking, heaters, a fit-out.
4. Ask who on site has been shown how to set and unset the alarm, and when. Show anyone who has not.
5. Ask your monitoring company exactly what happens when an alarm comes in: who is called, whether it is verified, and who attends.
6. Ask what your alarm’s communicator uses to report, and check it does not depend on the 3G network.
7. Read your insurance policy for an alarm condition: whether the alarm must be monitored, set whenever the premises are empty, or meet a grade.
8. Ask when the system was last tested and serviced, and get the record.

How Blue Eagle handles it

We design detection in layers, for the risk on each site and the environment of each space, and tell you which grade the design meets. SwiftProtect™ is our monitoring and intelligence service, run with our monitoring partners. Where it is in place with AI analytics, a person verifies what triggered an alert before anyone responds. That verification step is the difference between a dispatch and a false alarm.