The short answer
Testing a lone-worker alarm properly means verifying three links end to end, not just that the device beeps: coverage at the exact place you work, that a real person or monitoring centre receives and acknowledges the signal, and how long it actually takes from pressing the alarm to verified help. Run a functional test before every shift and a full live drill at least annually, recording every result with a timestamp.
Key takeaways
- A test that only proves the device 'works' gives false confidence: you must verify coverage, acknowledgement and real response time across the whole chain, not just the unit on the bench.
- Carrier coverage maps are not confirmation. Verify signal on site at every actual work location, including underground and confined spaces, and treat intermittent coverage as a dead zone with a written fallback.
- Exercise every trigger type separately — manual SOS, man-down/no-motion, timed check-in and pre-alarm cancel — not only the panic button.
- A signal is 'received' only once the monitoring centre or supervisor explicitly acknowledges it; devices should repeat the alarm until receipt is confirmed.
- Measure true time from activation to verified help; published benchmarks such as a 30-second connection under BS 8484 and an 80%-in-10-seconds answer target give a realistic yardstick.
- Functional test before each shift, an annual inspection by a competent person, and a full live assistance-chain drill at least yearly and after any change or repair.
Why a shallow test is false comfort
A lone-worker alarm is a chain of dependencies, not a standalone gadget: the trigger, the transmission channel (cellular, radio or satellite), the receiving side that must acknowledge the signal, and the person or service that reaches the worker within a target time. If even one link is untested, a unit that 'works' on the desk will not protect a person in the field.
The UK Health and Safety Executive, in its guidance on protecting lone workers, stresses that employers must provide training, supervision, monitoring and support and keep in touch with those working alone. In practice that means a test must reproduce a real scenario, not stop at the device's built-in self-check. Separate two levels: the functional check of the unit itself and the end-to-end check that includes the controller and whoever physically responds.
A common mistake is treating a test call that 'went through' and was logged as proof of readiness. A test call proves transmission; it does not prove that the controller answered, recognised the worker, located them and knew what to do next. Only a live exercise answers those questions.
- The three points of failure: transmission, acknowledgement, physical response
- A functional unit test does not replace an end-to-end test with the controller
- The test must reproduce the real scenario, including location and time of day
Layer 1 — coverage and connectivity
Coverage is verified on site, not from a carrier map. Maps show broad signal zones and miss basements, tunnels, lift shafts, areas behind metal structures and rural locations with variable signal. The German rule DGUV Regel 112-139 states that before a personal emergency system is used, the radio path to the receiving centre must be checked against local and operational conditions, for example by field measurements, and that repeat measurements may be needed after structural changes.
In practice, treat intermittent coverage as a dead zone rather than 'generally works'. For those points, define the fallback in advance: satellite connectivity, a fixed call-back window or a buddy check procedure. Guidance for fleets working beyond cellular reach recommends confirming coverage for each high-risk site before dispatch and never assuming it from operator maps.
Check connectivity at the point of entry into tunnels and confined spaces separately from the surface: outdoor signal does not guarantee reception underground. Multi-network SIMs reduce dependence on any single operator but do not remove the need to measure at the specific location.
- Radio measurements or test calls at every real work location, including underground and confined spaces
- Re-measure after structural changes, relocations or network upgrades
- Intermittent coverage is planned as a dead zone with a written fallback
- The fallback (satellite, call-back window, buddy check) is agreed before the shift starts
Layer 2 — acknowledgement of the signal
A signal counts as delivered only once the receiving side acknowledges it. Under the German pre-standard DIN VDE V 0825-11, a personal emergency device must repeat the alarm until acknowledgement of receipt is shown, and the radio link between device and receiver is checked regularly, with a technical alarm raised on failure.
Test each trigger type rather than only the SOS button. Besides a manual call, most devices carry a man-down or no-motion alarm (fired when the worker stays still and does not reset it in time), a timed check-in alarm and a resettable pre-alarm that precedes an automatic call. Each scenario has different conditions and timing and must be exercised on its own.
Confirm that the controller sees exactly what they should: the worker's identity, location, alarm type and a working voice channel where speech is part of the solution. Test voice quality, not just that the connection was established, and record the outcome with a timestamp.
- The alarm repeats until acknowledgement of receipt is shown
- All triggers are tested: manual SOS, man-down, timed check-in and pre-alarm cancel
- The controller sees identity, location, alarm type and hears the voice channel
- Radio-link monitoring between device and receiver runs regularly
Layer 3 — time from alarm to help
The decisive metric is not transmission time but the time from activation to the start of real help. Measure it with a stopwatch in a live drill, not 'by feel'. Recognised schemes give comparison points: under the UK standard BS 8484 a device or app should connect to the Alarm Receiving Centre within 30 seconds of a red-alert press, and centres publish answer targets such as responding to 80% of calls within 10 seconds.
Separate 'time to answer' (an operator picks up), 'time to verify' (the operator assesses seriousness) and 'time to help' (someone physically intervenes). The German DGUV Regel 112-139 includes a table of maximum reaction times: a voluntary personal alarm within about 2–30 seconds depending on configuration, automatic alarms (position, rest, loss) within about 90 seconds, and a technical alarm within about 10 minutes.
These figures are benchmarking targets for conformance and provider audit, not a universal legal limit. Set your own realistic target from the risk assessment — the hazard, distance to the nearest help and how long help remains effective — write it down and check it is actually met at each drill.
- Measure three intervals: to answer, to verification, to physical help
- BS 8484 (UK): ARC connection within 30 seconds; centre targets often 80% answered in 10 seconds
- DGUV 112-139 (Germany): a table of maximum reaction times by alarm type
- Your target comes from the risk assessment and is verified in a live drill
Frequency, responsibility and records
Different checks run at different intervals. A functional test with a visual inspection is carried out before work begins — DGUV Regel 112-139 states systems must be checked by a functional test and inspection before the start of work. The worker typically does this when collecting the device and returns it immediately if a fault is found.
Deeper inspection falls to a competent person: in Germany the rule requires a check before first commissioning, after repairs and thereafter as needed but at least annually. A full live drill of the assistance chain, including real responders reaching the 'casualty', takes place before first commissioning and then at least once a year, and it should be repeated after any change to procedures or equipment.
Document everything: every personal alarm, every technical alarm and every test with a timestamp, the name of the tester and the result. The log lets an auditor or supervisor see that the system is genuinely exercised rather than merely installed. Keep the reports until the next inspection and make them available to the enforcing authority on request.
- Functional test and inspection before every shift, done by the worker
- Full inspection by a competent person before commissioning, after repair and at least annually
- Live assistance-chain drill before first use and at least annually
- A log of all alarms, technical alarms and tests with timestamps and results
Put it into practice
Lone-worker alarm test protocol: three-layer check sheet
Use this sheet for every end-to-end check and keep it in the log. Items 1–3 are the worker's functional test before each shift; the rest belong to the scheduled inspection and the live drill run by the responsible person.
- Functional test of every unit before issue: button, indicators, sound and screen respond correctly.
- Each trigger type tested separately: manual SOS, man-down/no-motion, timed check-in and pre-alarm cancel.
- Battery and signal level sufficient for a full shift; charge or replace before issue if uncertain.
- Coverage confirmed on site at every actual work location, including underground and confined spaces — not assumed from carrier maps.
- Points with intermittent reception flagged as dead zones with a written fallback (satellite, call-back window, buddy check).
- Radio path or connectivity re-checked after structural changes, moves or network upgrades.
- Acknowledgement confirmed: the device repeats the alarm until the controller explicitly confirms receipt.
- Receiving side verified: worker identity, location, alarm type shown and the voice channel working.
- Time from activation to the controller's answer measured by stopwatch and compared with your target (e.g. connection within 30 seconds).
- Live assistance-chain drill run at a real remote or dead-zone site, with time to physical help recorded.
- Annual inspection by a competent person performed; re-test after any repair, maintenance or software update.
- Every test, alarm and technical alarm logged with timestamp, tester and result.
Questions people ask
How often should a lone-worker alarm be tested?
A functional test with a visual inspection is done by the worker before every shift when collecting the unit — for example, DGUV Regel 112-139 in Germany requires systems to be checked by a functional test and inspection before work starts. A full inspection by a competent person takes place before first commissioning, after repairs and at least annually. A live drill of the whole assistance chain, with real responders reaching a 'casualty', is held before first use and then at least once a year.
What does coverage verification actually involve?
It means confirming that a real signal reaches the receiving side at the exact location where the worker operates — not reading it off a carrier map. You make test calls or radio field measurements at each actual work point, including basements, tunnels and confined spaces, where reception differs from the surface. Intermittent areas are treated as dead zones with a written fallback such as satellite connectivity, a call-back window or a buddy check. Re-measure after structural changes or network upgrades.
What is the difference between a device self-test and an end-to-end test?
A self-test or functional check proves the unit itself: the button, indicators, battery and that a signal leaves the device. An end-to-end test proves the whole chain: the signal reaches the monitoring centre or supervisor, they explicitly acknowledge it, see the identity, location and alarm type, and real people begin help within the target time. Because a unit can pass its own test while the receiving side fails to respond, the two checks are not interchangeable and both belong in the programme.
Who should perform the formal inspection of a lone-worker alarm system?
A competent person with the relevant training, experience and current professional activity to assess the equipment. In the German model under DGUV Regel 112-139 this can be a manufacturer's service engineer; the system is inspected before first commissioning, after repairs and thereafter as needed but at least annually. The day-to-day functional test before each shift is done by the worker, who returns the unit immediately if a fault is found. All results are documented in an inspection report kept until the next inspection.
What should I do if a test alarm is not acknowledged by the controller?
Treat the check as failed and do not let the worker go on lone working with that system until it is resolved. First rule out the device — battery and settings — by repeating the test in a known-good-signal area. Then investigate the receiving side: queue, channel failure or incorrect configuration. Fix the cause, repeat the end-to-end test and record both the result and the corrective action in the log. Until it is fixed, move the worker to the fallback plan, such as buddy supervision or another channel.
Sources and further reading
Sources were checked when this page was generated. Confirm changing dates, rules and prices with the original publisher.
- BS 8484: Lone Worker Safety Standard ExplainedSoloProtect
- BS 8484 Explained — A Practical Guide to the Lone Worker StandardPeopleSafe
- Are you looking for a lone worker monitoring station?First County Monitoring
- Lone worker safety for utility fleets: A checklist for the sites beyond cellular reachGeotab
- What does DIN VDE V 0825-11 say? Requirements for lone worker protectionCalima
- DGUV Regel 112-139: Einsatz von Personen-Notsignal-Anlagen — Funktionstest, Reaktionszeiten, Prüfungenumwelt-online / DGUV
- Protecting lone workers: How to manage the risks of working alone (INDG73)Health and Safety Executive (HSE)