Intrusion detection and alarm

An intruder alarm is worth having in proportion to how early it goes off. Detection layered from the outside in — perimeter, then the shell of the building, then the space inside, then the specific asset — gives a response time that means something. Detection only at the strong room gives you a notification and a loss.

The other half of the job is not crying wolf. A system that false-alarms gets switched off by the people it protects, and that is the most common way an intrusion system fails. Detector choice and siting are what prevent it: a PIR pointed at a roller shutter that moves in the wind, or at a heater, or at a gap under a door where a cat gets in, will alarm every night until somebody isolates the zone. We survey against those, use dual-technology detectors where the environment warrants it, and set entry and exit routes so ordinary use of the building does not trip it.

It also has to be usable. Arming and disarming has to be quick and obvious for whoever locks up, part-arming has to work for a night shift in one wing, and the alarm has to go somewhere — a monitoring centre, a keyholder’s phone, the security desk, or all three. We commission the whole chain including the notification path, and train the people who will use it daily.