Where person-location deployments fall short — and the fix.
A person-location system is judged on one day you hope never comes. These are the ways installations quietly fail that test, all of them found long before an emergency if you go looking.
| What goes wrong | How you notice | The fix |
|---|---|---|
| Readers placed to a map, not to travel routes | Last known location resolves to a shaft station rather than a working place | Walk the routes. Readers belong where people change level or direction, not spread evenly across a plan |
| Tag carried on a badge | Headcount reconciles but location does not | Tie the tag to something nobody works without — the cap lamp — so carrying it is not a daily decision |
| Backbone sharing a supply with everything else | The system is unavailable during exactly the event it exists for | Give readers and backbone their own protected supply, and prove it in an outage drill rather than on paper |
| Readers not time-synchronised | Reconstructed movement makes no sense after the fact | Synchronise the network. Without a common clock the order of passes is guesswork |
| Tag health invisible | A flat tag looks identical to a person who never went underground | Treat tag battery and last-seen as a first-class alarm, not a line in a monthly maintenance report |
| Issuance not tied to access control | Contractors appear as employees, or not at all | Reconcile the tag register against the access system daily; a stale register is worse than none |
| Emergency query is an administrator’s task | In a drill, people reach for the phone instead | The per-person query must be one action performed by whoever is in the control room at 03:00 |
| Never drilled against a physical roll-call | The data exists and nobody trusts it under pressure | Run the drill, compare the system’s answer to the roll-call, and fix the difference — that comparison is the only real commissioning test |
Two of these decide whether the system works when it is needed. Power is the first: a person-location system that depends on the same supply as everything else is unavailable in precisely the scenarios — fire, inrush, power failure — that make it necessary. The drill is the second, and it is the only honest commissioning test: until the control room’s answer has been compared against a physical roll-call, nobody knows whether the coverage matches the way people actually move.