Coverage, and How to Survey It

Everything in a lone worker system depends on a signal existing at the moment it is needed, and that is the assumption least often tested.

When coverage, and how to survey it also depends on reliable work records, contractor attendance tracking can support time, attendance and workload review without being treated as the emergency response itself. Compare any workflow with ISO standards catalogue and keep alarm ownership, escalation and dispatch responsibilities explicit.

Why it matters more here than elsewhere

In most systems a coverage gap is an inconvenience.

Here it is the difference between an alarm and silence, at exactly the locations where serious incidents occur.

And it is invisible until tested, because an operator's coverage map describes outdoor reception at street level and says nothing about a basement, a plant room, a lift lobby or a house with thick walls.

How to survey

Walk the actual locations with the actual device.

Not a phone, unless the phone is the device.

Record three states — full, marginal, none — and the position within the building.

Two people, an afternoon per site.

The output is a map that the assessment, the schedulers and the workers all use, which is more than most safety documents achieve.

What to test at each location

Can an alarm be raised?

Does it arrive, and how long does it take?

Does a location come with it, and is that location right?

And does a two-way call work, since an alarm that arrives without the ability to speak leaves the operator guessing.

The last two fail more often than the first, which is why raising a test alarm beats looking at a signal bar.

What to do with the gaps

Small gap, low hazard: a check-in either side of the period spent in it.

Gap coinciding with a hazard: the task moves, gets paired, or gets different technology.

Genuinely remote work: satellite, which is the only real option.

And a stated rule for entering a known no-signal area — tell somebody, agree a time, confirm on exit.

That rule is the oldest control in this subject and covers what no device will.

Store-and-forward, and what it does not solve

Many devices queue an alarm and send it when signal returns.

Useful for a record and useless for a response, since the alarm arrives after the person has left the gap, by which time they have either recovered or not.

Check whether the timestamp reflects the alarm or the delivery, because a delayed alarm presented as current produces a response to the wrong place at the wrong time.

Multi-network SIMs

A SIM able to use more than one network covers more ground than one tied to a single operator.

A specification item worth asking about, and not standard.

In rural areas it is the difference between a device that works and one that does not, and the cost difference is small enough that it should not be a decision.

Re-surveying

When a site changes: a refurbishment, new plant, a new building on the same footprint.

When the network changes, which happens without announcement.

Annually for the critical locations.

And immediately after any incident where an alarm did not arrive, which is the survey that should have happened earlier.

Publishing it

To the workers, because they need to know where they are out of contact.

To the schedulers, because it should influence who goes where and when.

And into the assessment, because a known coverage gap that is not recorded is a foreseeable failure that was not addressed — which is precisely the phrase that appears in enforcement findings.

The client premises problem

In home-visiting work the locations are private houses, and surveying them is impractical.

Which means coverage is unknown at exactly the locations with the highest incident rate.

The practical substitute is recording it as it is discovered: a field on the visit record for whether there was signal, completed once and carried forward.

After a few months this produces a coverage map of the round, built by the people who go there, at no cost.

Testing after a device change

A new device model changes everything measured here.

Different radio, different antenna, different network agreement.

Which means a coverage map produced for the old estate does not transfer, and a supplier change with the same map is an assumption nobody checked.

Re-survey the critical locations before cutover, not afterwards.