How These Programmes Fail

Lone worker systems fail in recognisable ways, and almost none of them is the hardware.

When how these programmes fail also depends on reliable work records, employee time tracking tools 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.

The stale escalation tree

Symptom: a test call reaches somebody who left the organisation.

Correction: quarterly review, and a line in the leaver checklist.

Office-hours monitoring for out-of-hours work

Symptom: the alarm log shows activity at hours nobody was watching.

Correction: plot when lone working happens, then match the monitoring to it.

Devices left in vehicles

Symptom: alarms raised from the van, or positions that never move with the worker.

Correction: ask why. It is usually comfort, or a belief that it tracks them, and both have answers.

False alarms tolerated

Symptom: response times lengthening with no decision to lengthen them.

Correction: measure the rate, tune sensitivity, fix the wearing position.

No cancel path

Symptom: accidental activations going unreported, so the responder's picture is wrong.

Correction: a five-second stand-down that needs no phone call.

Withdrawals never happening

Symptom: zero recorded in a year.

Correction: the policy is not believed. Say it again, record the next one as a success, and make it visibly fine.

The intermittent lone workers uncovered

Symptom: an incident involving someone the policy did not consider a lone worker.

Correction: assess by shift and situation rather than by job title.

The two-worker requirement unfunded

Symptom: flagged visits attended by one person when staffing is short.

Correction: a scheduling rule that cannot be overridden by shortage, and an escalation when it cannot be met.

Testing that never happens

Symptom: nobody can say how long a response takes.

Correction: an unannounced test, at night, this quarter.

The common thread

Every one is a process or a staffing decision rather than a technology problem, which is why replacing the supplier fixes none of them — and why the fix is usually free.

How to find which ones you have

One afternoon, five questions.

When was the escalation tree last tested, and what were the timings?

How many withdrawals were recorded in the last year?

What proportion of alarms are false, and does the responder still take each seriously?

How many devices are unaccounted for?

Which situations in the rota are not covered by the policy?

Every failure in this note shows up in one of those answers, and an organisation that cannot answer them has not found the failures rather than not having any.

The order to fix them in

The escalation tree first, because everything else ends there and because it is free.

Then the coverage hours, since monitoring at the wrong times protects nobody.

Then the false alarm rate, because it is silently degrading the response.

Then the intermittent lone workers, which is a scoping fix rather than a spending one.

And the device estate last, which matters least and feels most urgent.

What none of them requires

A new supplier.

A larger budget, in most cases.

Or a new product, which is what will be proposed when the failures become visible.

Each is a process or staffing decision, and the work is in making the decision rather than in implementing it.

The failure that hides the others

Nobody owning the arrangement.

Because all of the above are found by somebody looking, and an unowned programme has nobody whose job that is.

Which is why the ownership note sits next to this one, and why naming a person is the single most effective intervention available after the escalation tree.

The failures in the sectors

Home visits: the flag that did not travel to a worker from another organisation.

Retail: a fixed alarm nobody has tested and a cash routine that has not changed in years.

Remote work: a system that treats no signal as an alarm, in an area where no signal is normal.

Nights: a responder list staffed for daytime.

Each is the general failure wearing local clothing, and each is found by the same five questions applied to that setting.

Recovering a programme that has already failed

Do not start by replacing anything.

Test the tree, publish the response time, fix the worst finding, and tell the workforce what you fixed.

Then classify a month of alerts and tune the thresholds.

Then re-scope the assessment to situations rather than titles.

One quarter of that converts an estate of ignored devices into a working arrangement, and it costs nothing beyond attention — which is the honest summary of this whole subject.