Where to place guard tour patrol checkpoints

Put the checkpoints in the wrong places and you get perfect scan compliance covering nothing. The guard taps every tag, the report comes back green, and the fence line nobody walked is where the hole gets cut. Placement is the whole system.

What is a checkpoint actually proving?

Four things, and they are not the same thing.

• The guard was physically at that spot.
• The guard was there at a recorded time.
• Someone had a reason to look at that specific thing.
• You can answer a client who says the site was never patrolled.

The first two are automatic. The third is the one placement controls, and it is the one that matters on the night something happens.

A tag by the front door is easy to install and proves nothing about the perimeter. A tag at the far northeast corner cannot be reached without walking the fence. Same hardware, completely different evidence. Ask of every proposed location: what does a scan here prove that a scan at the next one over does not?

Then pair each one with an observation. The loading dock checkpoint is not a presence marker. It is the reason someone checks that the overhead door is down, the trailer is where it should be, and the man door beside it is latched. Write that expectation into the post orders, because a checkpoint with no stated observation becomes tag-tapping within a month.

Where do exterior checkpoints belong?

Exterior work usually eats most of the patrol, and this is where the difference between real and theatrical perimeter security gets decided.

Start with geometry. Four corner checkpoints on a rectangular property mean the guard cannot complete the tour without covering the whole fence line. That single decision does more than the next ten placements combined.

Then work through the fabric of the site.

• Gates and fence access points. Every one, plus any section that has been breached before. History repeats at the same weak spot.
• Camera blind spots. The places the system cannot see are exactly the places that need a person. Pull the camera coverage map and put checkpoints in the gaps.
• Every exterior door. Doors that should be locked, emergency exits that get propped for a cigarette break, the loading dock.
• Ground-floor windows and roof access. Roof ladders and mechanical penthouse doors get skipped in most routes.
• Parking areas and dumpster enclosures. After hours, this is where the incidents actually occur.
• Utility gear, generators and outbuildings. Unglamorous, and precisely what copper thieves and vandals go for.

If you are building this list from scratch, a site security assessment gives you the vulnerability list first. Placement is easier when it is answering a survey rather than guessing.

What gets missed inside the building?

Interiors get shortchanged, and a large share of losses happen indoors. Give the inside the same discipline.

High-value rooms. Server rooms, cash handling, pharmacy or controlled substance storage, high-value inventory. These earn a checkpoint each, and the scan should come with a look at the door and frame, not a wave in the corridor.

Doors that are supposed to be locked. Stairwell doors, restricted zones, after-hours areas. The value here is repetition. A guard who checks the same twelve doors nightly notices the one that is different.

Safety equipment. Extinguisher stations, AEDs, first aid cabinets, marked exits. Two jobs at once: you confirm the gear is present and unobstructed, and the guard builds a map of it in their head. At 3am during a real medical call, that map is the whole difference.

Building systems. Electrical rooms, HVAC, elevator machine rooms, telecom closets. Half the value here is not security at all. Guards find water leaks, failed equipment and open panels long before the facilities team does, and that finding is worth reporting through the same incident report flow as anything else.

NFC, QR code or GPS?

Different physics, different evidence, different failure modes. Match the technology to what you need to prove.

Type	Proves presence	Best for	Fails when
NFC tag	Strongly, by physics	Exterior, high-accountability, weather	Tag damaged, removed or painted over
QR code	Weakly	Interior, well-lit, low fraud risk	Photographed once, scanned anywhere after
GPS zone	Approximately	Large open sites, no hardware to mount	Indoors, urban canyon, roof versus ground floor

The NFC case rests on range. The NFC Forum specifies the technology at 13.56 MHz with a typical range of up to 2cm. Two centimeters is not a policy. It is a physical constraint: the phone was touching the tag, and the tag is bolted to the wall you claim was checked.

That is why NFC is the default for anything you might have to defend. Mount tags at chest height, on something structural, protected from a snowplow and a bored teenager.

QR codes are cheap and replaceable, which is exactly their weakness. A code is an image, and an image travels. One photo on a phone and the accountability is gone. They are reasonable inside a building where lighting is consistent and the fraud risk is low, and a poor choice on a dark perimeter.

GPS zones need no hardware, which makes them attractive for large lots, yards and acreage where mounting tags is impractical. Know the limit. GPS.gov puts smartphones at about a 4.9 meter radius under open sky, degrading near buildings, and a guard can drive through a zone without ever looking at anything. Treat it as coverage, not proof. Configuring geofences without the false alarms covers the tuning.

Most sites end up mixed. NFC on the perimeter and at every high-value room, QR inside where it is cheap to maintain, GPS across the back acreage.

How long should the route take?

Longer than the desk estimate. Always.

Design the sequence to flow: outer perimeter, then building exterior, then inside, without doubling back. Group nearby checkpoints so an area gets covered once, properly, instead of the guard crossing the lot four times. Backtracking is where the time goes and where the resentment starts.

Then walk it yourself. At patrol pace, stopping at every checkpoint to actually look at the thing you are asking guards to look at, using the stairs they will use and waiting for the elevator they will wait for. Do it at night if the post runs at night.

Add a buffer on top of that walk. The honest number is whatever your own site produces in bad weather with one unexpected conversation. A 45-minute route in daylight is an hour in February.

Vary the pattern. A route run at the same times in the same order teaches anyone watching exactly when the back lot is empty. Alternate the sequence, stagger the start times, and keep a separate high-risk route for the hours that need it. Building routes that repeat on schedule without becoming predictable goes into the mechanics, and night patrol covers what changes after dark.

What goes wrong most often?

Too many checkpoints. Thirty tags on a 45-minute route turns the patrol into a foot race. Ten checkpoints a guard can stop and think at will catch more than thirty they sprint past. Count is not coverage.

Impossible timing. A route that cannot be finished on time gets finished dishonestly. Guards skip tags, scan without looking, or work out how to game it. That is a design failure, and it is on the person who set the window.

Placement that never changes. Construction goes up, a tenant moves out, a dumpster relocates, incidents cluster somewhere new. Review the map quarterly against the last quarter's incidents, or you are protecting last year's site. The patrol data itself will tell you: checkpoint completion by location shows which ones nobody reaches and which take three times as long as planned.