It tells the controller that an exit is legitimate, so the door opening does not raise a forced-door alarm.

The controller cannot otherwise distinguish someone leaving from someone forcing the door - both look like an opening with no valid credential.

On a door with a mechanical handle, the person leaves regardless of the electronics. The device's job is purely to suppress the alarm for a short shunt period.

The event is still logged, so the audit trail shows an exit rather than nothing at all.

On a door held by a magnetic lock, the same signal may be configured to release the lock, because there is no mechanical way out.

Which of those two behaviours applies is a configuration decision and should be recorded on the door schedule, not left to whoever commissioned it.

A switch in the hardware is more precise because it responds to an actual attempt to open; an infrared sensor is easier to retrofit and covers any exit method.

Handle, lever and panic-bar switches operate only when the mechanism moves, so they do not fire when someone merely walks past the door.

That precision matters on doors in circulation routes, where an overhead sensor shunting the alarm every time someone passes leaves the door effectively unmonitored.

Infrared sensors need no modification to the door or the lockset, which is why retrofits favour them.

They also cover exits that do not use the handle - a door pushed open on a free-swing closer, for instance.

Many doors use both: a hardware switch for precision and a sensor as a backstop, with the controller accepting either.

Yes, and it is a well-known attack - through a letterplate, a gap under the door, or an adjacent glazed panel.

Anything that can put warmth or movement into the sensor's field from the insecure side will shunt the alarm and, on a magnetic lock, may release the door.

The defences are positional. Mount the sensor so its field does not reach the door line from outside, use the masking supplied with it to trim the pattern, and avoid siting it directly above a letterplate.

Doors with glazing beside them deserve particular attention, because a sensor visible through the glass can be targeted.

Where the risk is high, a hardware switch removes the attack entirely because it requires the mechanism to move.

Test it deliberately at commissioning: try to trigger the sensor from the insecure side with whatever the door's geometry allows.

Long enough for a person to open the door and pass through, and no longer - typically a few seconds, set from observed traffic.

Too short and the alarm fires while somebody is still in the doorway, which trains staff to ignore it.

Too long and the door can be propped open within the shunt window without any alarm at all, which is the more serious failure.

Doors used for deliveries or trolleys need longer than a personnel door, and that should be set per door rather than globally.

Held-open detection with its own longer timer covers the case where the door genuinely stays open, and a local sounder before the alarm usually solves propping without involving anyone.

Review the settings after a few weeks in service. The values chosen at commissioning rarely match how the door is actually used.

Only where it releases the lock, and even then it is not the emergency provision - that is a separate device wired independently.

On a door with a mechanical handle, escape does not depend on the electronics at all and the sensor is purely an alarm-management device.

On a magnetically locked door where the sensor releases the lock, it has become part of the means of escape, and its reliability and power supply matter accordingly.

Even in that case, local fire rules will normally still require an emergency release that cuts lock power directly, independent of the controller and the sensor.

A sensor that has failed, been masked or lost power must never leave people unable to leave.

Establish the requirement with the authority having jurisdiction rather than inferring it from a previous project in another country.

Yes. An exit that leaves no record is a gap in the audit trail and hides patterns worth seeing.

Logging exits shows how a door is actually used, which is the information needed to set the shunt and held-open timers sensibly.

It also distinguishes a genuine exit from a forced door in the record, which is exactly what an investigation needs.

Repeated exit activations with no corresponding entry can indicate a door being propped or a sensor triggering spuriously.

The volume is high on busy doors, so check the event storage and retention settings can cope before enabling it everywhere.

Where storage is constrained, log exits on the doors that matter rather than turning the feature off entirely.

It stops sensing - so the door's behaviour depends on the lock type and the standby supply, and the emergency release remains the guarantee.

An unpowered sensor cannot shunt an alarm or release a lock. On a mechanically handled door that changes nothing about escape.

On a magnetically locked door, fail-safe operation means the lock releases when power is lost, so people can still leave.

Standby batteries in the panel are what keep the door secure during a short outage, and they have to cover the locks as well as the electronics.

This is precisely why the emergency release is wired to break lock power directly. It is the one device that must work when everything else has stopped.

Test the whole sequence during commissioning: mains off, battery in circuit, then battery removed, checking the door behaves as the fire strategy requires at each stage.