How do NFC checkpoints prove that a round was done?
They prove that a device read that tag at that time, which is more than a signature on a sheet and less than a person being there. The value is in the sequence: a round with times attached shows what was visited, in what order, and what was skipped.
What a checkpoint records
A tag fixed at the point to be visited, a phone or a dedicated reader carried by the person doing the round, and one record per tap: which tag, at what time, on which device.
Plant rooms, basements and stairwells have no coverage, so the application has to queue records and send them later. That is the difference between a system that works on a real round and a demonstration in an office, and the offline page sets out what a tap can do with no network at all.
The reader matters less than people expect. A phone is enough, and a dedicated reader earns its place where gloves, cold or a policy against personal phones make a phone impractical.
What it proves, and what it does not
It proves a device read that tag at that time. Not that a person was attentive, not that anything was inspected, and not that the tag was in its place, since a tag can be copied or moved, which the copied checkpoint page covers in full.
Compared with a paper sheet signed at the end of a shift, that is a real improvement: the times are recorded as they happen rather than reconstructed, and a round done in four minutes is visible as such.
Compared with the claim sometimes made for these systems, that they prove the work was done, it is not. What turns a checkpoint into evidence of work is what the round captures alongside the tap: a reading, a photograph, an answer to a question at that point.
Designing the round itself
Put the tag at the thing that has to be checked rather than at the door of the room it is in. A tag by the entrance proves somebody opened a door.
Fix it the way the maintenance page describes: the same place on every similar asset, marked, on metal where the surface is metal, and mechanically fixed wherever an adhesive would be the weak point.
Expect a replacement cycle and plan it, since a missing checkpoint turns into a gap in a record that somebody will have to explain, and the replacement page covers keeping a population alive.
Write down where every checkpoint is and what it corresponds to. A round documented only in the application is a round nobody can audit, which the documentation page treats as the thing a deployment owes its successor.
What the record is worth
For managing the work, a lot: routes that are too long, points that are always skipped, shifts where the round happens in the last ten minutes. That is the return that justifies the system in most sites.
For a dispute, it is evidence of what was recorded, which is not the same as evidence of what happened. Treat it accordingly, and say so to the people relying on it.
And be clear about the second thing it creates, a record of where a named person was through a shift, which the monitoring page takes as its subject.
Frequently asked questions
Does a checkpoint prove somebody was there?
It proves a device read that tag at that time. A person carrying a device is an assumption the system makes rather than something it establishes.
What happens where there is no signal?
The records queue on the device and go up later. Any system for rounds has to assume basements and plant rooms have no coverage.
Where should the tag go?
At the thing being checked, not at the door. A tag by the entrance only proves that a door was approached.
Is a phone good enough?
Usually yes. A dedicated reader helps with gloves, cold, or a policy that keeps personal phones out of the workflow.