Guard Patrol Systems for Housing Societies: QR Checkpoints & Proof of Patrol

Guard patrol with QR checkpoints

Almost every society pays for night patrols. Very few can prove they happened.

The traditional instrument is a register in the guard cabin where the guard writes a time against each round. Everyone involved knows its weakness: it can be filled in at 6am for the whole night in about ninety seconds, and it usually is. The society is not buying patrols, it is buying a signature.

QR checkpoint patrols replace the signature with evidence. A tag is fixed at each point on the route; the guard scans it as they pass; the scan carries a timestamp that cannot be back-dated. The committee gets a record of which points were covered, when, and by whom — and, more usefully, which were not.

How the system works

Checkpoints. A QR tag at each location on the route — basement lift lobby, terrace door, pump room, rear gate, each wing entrance. Weatherproof, fixed where it cannot be photographed and reused from the cabin. Ten to fifteen points is typical for a mid-size society.

Schedule slots. The rounds the society expects, by time. A common pattern is four night rounds at 22:00, 00:30, 03:00 and 05:00, plus two day rounds. Slots are what make a missed round detectable — without an expected schedule, "no scans since midnight" looks the same as "no patrol scheduled".

Rounds. A guard starts a round, scans checkpoints as they go, and ends it. A round that is started but never ended within a reasonable window is marked abandoned, which is a distinct signal from a round that was never started at all.

Scans. Each scan records the checkpoint, the guard, and the server-side timestamp. This last part matters — if the time comes from the guard's phone, the whole thing is defeated by changing the phone clock.

Incidents. A guard who finds something on a round — a lift door jammed open, a car alarm going, a light out in the basement — reports it against the round, with a severity. This is the part that turns patrol from a compliance ritual into an operations input.

Designing a route that gets walked

The route design decides whether the system is useful or just theatre.

• Put checkpoints where the risk is, not where it is convenient. The rear gate and the basement are where things happen. The main lobby, which the guard passes anyway, proves nothing.
• Order them so the route is a sensible walk. If the sequence forces backtracking, guards will reorder it, and your data becomes hard to read.
• Fix tags out of easy reach and out of camera-line from the cabin. The point of a physical tag is that presence is required.
• Do not over-instrument. Twenty-five checkpoints on a 15-minute round means the guard is scanning, not looking. Ten good points beat twenty-five rushed ones.
• Include at least one point that requires opening a door — terrace, pump room. These are the places an intruder uses and the places a guard skips.

Reading the data honestly

The value is in patterns over weeks, not in individual rounds.

That last row is the one committees miss. A patrol system that produces perfect rounds and zero incidents is not a well-run society; it is a system where nobody feels safe reporting anything.

Rolling it out with the security agency

This is a labour-relations exercise as much as a technical one. Guards correctly read patrol monitoring as surveillance of their work, and if it is introduced as a gotcha you will get compliance-shaped behaviour rather than better security.

• Talk to the agency supervisor before the guards. Patrol data will surface agency staffing problems; the supervisor needs to be a participant, not a defendant.
• Frame it as protection for the guard too. When a resident claims nobody patrolled the basement on the night their car was scratched, the scan log defends a guard who did their job.
• Handle the phone question. Decide whether guards scan on a society-provided device or their own. A shared cabin device is usually cleaner and avoids expecting staff to use personal data.
• Run two weeks in observation mode, with no consequences attached, and fix the route problems that surface.
• Agree what happens on a missed round — supervisor notified, logged, discussed at the monthly security review. Write it down.

What to report to the general body

At the AGM, three numbers make the case: rounds completed against rounds scheduled, incidents raised and how many were closed, and the median time from incident to resolution.

This is also the evidence base for the harder conversation about whether the society needs a third night guard. "It feels unsafe" loses that argument every year. "We scheduled 120 night rounds last quarter, 31 were not completed, and all 31 were on the shift where one guard covers both gates" wins it.

How this works on Plinth

Checkpoints and schedule slots are configured by the admin and can be deactivated when a route changes, rather than deleted — so historical rounds stay readable.

Guards start a round, scan checkpoints, report incidents with a severity, and end the round. Scans are timestamped server-side. Rounds left open past their window are marked abandoned. A day log view shows the committee every round, scan and incident for a given date, and incidents carry through to resolution rather than disappearing at the end of the shift.

Everything is scoped to your society, and the record is append-only — a completed round cannot be quietly edited afterwards.

Frequently asked questions

Can a guard cheat by photographing the QR codes? They can photograph them, which is why tag placement matters — points behind doors that must be opened, and out of line-of-sight from the cabin. The scan-interval pattern also gives it away: a full route completed in ninety seconds is visible in the data.

Do we need special hardware? No. A phone camera reads the tags. NFC tags are an alternative and are harder to copy, but they cost more and need NFC-capable devices.

What if there is no mobile signal in the basement? This is the most common practical problem. Choose a system that queues scans offline and syncs when the guard is back in coverage, and confirm the queued scan keeps its original scan time.

Does this replace CCTV? No. CCTV tells you what happened at a fixed point. Patrol tells you a human was present across the whole property. Societies with serious security needs run both.

Who owns the data if we change security agencies? The society, if the system is the society's. Avoid agency-supplied patrol apps where the history leaves with the contract — you lose your baseline exactly when you most need it.

Step-by-step guides

• Admin: Set Up Guard Patrol

Related: Emergency SOS panic button · society security guidelines checklist · visitor gate management