Where to Start

Most lone worker programmes start with a supplier demonstration. The order that works starts with two lists and a stopwatch.

When where to start also depends on reliable work records, the complete explanation can support time, attendance and workload review without being treated as the emergency response itself. Compare any workflow with HSE lone-working guidance and keep alarm ownership, escalation and dispatch responsibilities explicit.

Weeks one to four

List the situations, not the job titles. Every shift pattern where somebody is alone, including intermittently.

Plot the hours: when does lone working actually happen, across a month?

Most organisations find both lists surprising — more people than the policy covers, and more of it outside office hours than assumed.

Then the risk assessment, including the question of whether any of these tasks should be done alone at all.

Weeks five to eight

Build the escalation tree and test it.

Unannounced, at two in the morning, with a real alarm from a real location.

Time every step and write down what broke.

This costs nothing and will produce more improvement than any device, because whatever you buy ends at this chain.

Fix what the test found, then test again.

Weeks nine to twelve

Introduce check-in arrangements, which need no hardware — a call, a message, a shared sheet.

Write the withdrawal policy and say it out loud.

Walk the difficult locations with a phone and record where there is no signal.

Then, if a device is still indicated, trial one with the coverage map and the response times already known.

What to avoid

Buying before the assessment, which is how an organisation ends up with devices for people whose real problem is a two-worker requirement nobody funds.

Starting with the easiest team, which tells you nothing.

Announcing a programme before the escalation tree is tested, because the first real alarm will test it instead.

What a quarter should produce

A list of lone working situations by shift.

A risk assessment, dated, including any task that should not be done alone.

A tested escalation tree with timings.

Check-in arrangements in use.

A coverage map of the actual locations.

And a purchase decision made on evidence — which is sometimes that the arrangements are the programme and the device adds little.

The order and why it is that order

The assessment comes first because it determines what is needed, and because it is the document whose absence is itself a finding.

The escalation tree comes second because everything ends there, and because testing it is free and immediately informative.

Check-ins come third because they need no equipment and deliver protection while procurement takes its months.

Equipment comes last because it is the only step that requires money, and because by then you know what to buy.

Organisations that reverse this end up with devices that do not fit the work and a chain that does not answer them.

What the first test will probably find

Nobody covering the hours the work actually happens.

At least one wrong number.

A responder unclear about whether they may call emergency services.

And a total time that surprises everybody in the room.

All four are free to fix, which is why this step comes before a budget request rather than after one.

Getting the assessment done when nobody has time

Two hours with the people who do the work beats two weeks of a manager writing it alone.

Take the shift list, go situation by situation, and write what could go wrong and what happens if it does.

Record the controls you already have, because most organisations have more than they think and fewer that are tested than they believe.

And write down the tasks that should not be done alone, which is the output that matters most and takes the least time to produce.

The first quarter's deliverables, restated

A dated assessment covering situations rather than job titles.

A tested escalation tree with timings written down.

Check-in arrangements in use, with somebody named to receive them.

A coverage map from walking the difficult locations.

A withdrawal policy that has been said out loud by somebody senior.

And a purchase decision, if one is still needed, informed by all of the above.

Who should do it

Somebody with enough standing to change a rota, because the findings will require that.

Not solely a health and safety adviser without operational authority, whose assessment will be accurate and unimplemented.

And with the workers involved throughout, since the shift list and the coverage map come from them.

What to do when the answer is uncomfortable

Sometimes the honest finding is that a task should not be done alone and the staffing to avoid it does not exist.

Write it down anyway.

Then either fund it, change the task, or stop doing it — and if none of those is possible immediately, record the interim controls and a date.

An assessment that quietly omits the finding because it is inconvenient is the document that appears in an enforcement file.