Who Watches the Alerts

Every lone worker system ends at somebody looking at a screen or answering a phone. Who that is determines whether the system works at three in the morning.

When who watches the alerts also depends on reliable work records, task time tracking software can support time, attendance and workload review without being treated as the emergency response itself. Compare any workflow with ICO employment guidance and keep alarm ownership, escalation and dispatch responsibilities explicit.

In-house, during office hours

A named person or team watching alerts while the office is open.

Cheap, and they know the workers, the sites and the clients — which makes their assessment of an alert far better than a stranger's.

Useless outside those hours, which is when most lone working happens.

Viable where the lone working genuinely coincides with office hours and nowhere else.

In-house, on a duty rota

A duty phone carried by a rota of managers.

Covers all hours, keeps the local knowledge, and costs somebody's evenings.

Fails in predictable ways: the phone not handed over, the person asleep, holiday gaps, and the slow erosion as the rota becomes a burden nobody wants.

Workable in a committed organisation with enough people to share it, and a common cause of a system quietly failing where there are not.

A monitoring centre

A third party answering alerts around the clock.

Never asleep, never on holiday, and trained in escalation.

Knows nothing about your workers or sites unless you tell them, which means the quality of the information you supply is the quality of the response.

And it is a contract, with response times that should be specified rather than assumed.

The question that decides

When does your lone working actually happen?

Plot the hours worked alone across a month.

If a meaningful share sits outside office hours, in-house daytime monitoring is not an answer whatever else is true.

Most organisations have never plotted this and assume their pattern matches their office.

What a monitoring centre needs from you

Current worker list, with numbers.

Site and client information, including access.

The escalation tree, kept current.

Standing instructions: when to call emergency services without asking you first.

And your out-of-date data is their wrong decision, which makes the quarterly review of what they hold a safety task rather than an administrative one.

The hybrid that usually works

In-house during the day, where local knowledge is valuable.

Monitoring centre outside those hours.

With one escalation tree, not two, because two diverge within months and the divergence is discovered during an incident.

What the operator needs to decide well

The worker's name, role and normal pattern.

Where they were going and when they were expected.

Whether this worker has raised alarms before, and what happened.

Any medical information they agreed to share.

And the address, in full, with access details.

An operator with all of this decides in seconds. One with a name and a coordinate spends five minutes finding out what they are dealing with, and those five minutes are the difference the whole system is buying.

The quality question nobody asks

How many alarms does the monitoring centre handle per operator per hour?

And what happens at their peak?

A centre monitoring many contracts has a busy period, and yours may coincide with it — end of shift for a large sector is the same hour everywhere.

Ask for the answer time distribution rather than the average, because the average hides the tail and the tail is what you are buying protection against.

Contracting for it

A contracted answer time, measured and reported.

Standing instructions: the circumstances in which they call emergency services without contacting you first. This should be broad rather than narrow, and it should be written.

What happens when your escalation tree is unreachable, which is the situation that eventually occurs and which most contracts do not address.

Reporting: monthly alarm log with timings, so you can see the performance rather than trusting it.

And a route to review incidents together, because their operator's perspective on what went wrong is information you cannot get otherwise.

The in-house pitfalls specifically

The duty phone that does not get handed over, which is the commonest single failure of rota-based monitoring.

Holiday and sickness gaps, which are predictable and usually unplanned for.

Erosion: the rota starts with eight people and ends with three doing it every third night, at which point it fails quietly.

And the manager who silences the phone at a family event, which is human and is exactly why an always-available backstop matters.

Reviewing the arrangement

Quarterly: answer times, missed alarms, what the operator could not find out.

Annually: does the monitoring coverage still match when people actually work alone?

Rotas change, services extend their hours, and a monitoring arrangement set three years ago may now cover half the exposure.

That comparison takes an hour and is never done.

The hybrid in practice

One tree, one set of instructions, two sets of eyes at different hours.

Handover is the risk: the moment the in-house arrangement stops and the external one starts.

State the times precisely and make the transition automatic rather than a message somebody sends.

And test across the boundary, with an alarm raised ten minutes before and ten minutes after the changeover, which is the test nobody runs and the gap that exists.