Skip to content

On-board layer

Which train, and which carriage

Station-level location solves half the shift. The other half happens inside a moving train, and that calls for a different kind of answer: not a coordinate, but an identity.

The gap station location leaves

Metrosafe’s base layer ranges against the Wi-Fi access points in stations. It works well where those access points are: platform, concourse, corridors. The moment an officer boards a train, three things change at once and none of them help.

  • The reference moves. The access points stand still and the officer does not. Between stations there is nothing to range against.
  • The vehicle is a metal box. A metro car attenuates and reflects: it is among the worst possible environments for a radio measurement taken from outside.
  • And even with a good measurement, you would not know the thing that matters. An officer seven metres from the platform edge could be on the platform, boarding, or already inside a departing train.

That gap matters because a large share of patrol work happens exactly there: on board, between stations, where the officer is most alone and where an incident is hardest to reach.

One beacon per car, and the problem changes shape

The on-board layer fits one BLE beacon per carriage. Each beacon broadcasts its own identifier; the officer’s corporate phone reads it and the platform publishes the result.

ANDÉN TRAIN 4207 CAR 1 CAR 2 CAR 3 CAR 4 Car 3 · at station

What matters is that the answer stops being a coordinate with an error radius and becomes a discrete identification: car 3 of train 4207, currently at a given station. There is no room for doubt: either the phone sees car 3’s beacon or it does not.

What the service publishes

Train
identifier of the consist in service
Car
carriage number within the consist
Station
where it is, or which it is approaching
Direction
of travel, where operations provide it
Timestamp
and the quality of the reading

The two layers coexist without competing: in station the Wi-Fi measurement leads, giving the zone with metric accuracy; on board the beacon leads, giving train and car unambiguously. The platform picks whichever is trustworthy at that moment, and says which.

What changes in the control room

SituationBase layer onlyWith the on-board layer
Officer patrolling between stationsLast known position at the station they leftTrain 4207, car 3, approaching the next station
Alert raised from inside a trainYou know the line, not the trainYou know the train and the car: you can wait on the right platform, at the right door
Coverage reportingCoverage per stationCoverage per train and per car, which is how the service is actually contracted
Post-incident reviewWho was in the stationWho was on that train, in which car, and from which stop

The operational change people mention most: being able to go and meet the train. Knowing the officer is in car 3 of the service arriving in two minutes turns a search into an appointment.

What fitting it involves

Here is why this layer does not break Metrosafe’s central argument: the beacons go in the rolling stock, not in the infrastructure. Not a platform, not a concourse, not a metre of tunnel is touched.

Where
inside the car, high up and protected
How many
one per carriage
Fixing
adhesive; no drilling, no bracket
Power
own battery; no take-off from the vehicle
Cabling
none
When fitted
during a depot visit, not in traffic
Maintenance
battery replacement folded into the car’s inspection cycle
  • No electrical type approval needed: the beacon connects to nothing on the train.
  • No interference with on-board systems: it transmits in the BLE band, at low power, and does not receive.
  • Reversible: removing a beacon leaves the car as it was.
  • Incremental: you can start with one line, one class of stock, or the trains on a single shift.

And a consequence that often decides the project: with the on-board layer you do not have to survey the tunnels unless you want to. Between stations, car identity replaces the coordinate, so the radio survey can be limited to whichever stations matter.

The same privacy standard

The on-board layer changes nothing about data processing: it remains personnel location with a purpose bounded to operational safety. It is in fact less intrusive than metric ranging, because what gets published is coarse-grained: a car, not a point.

  • Collected during the shift and at the configured rate, not permanently.
  • The purpose is not mixed with performance or punctuality monitoring.
  • Prior information to staff and their legal representatives, as with the base layer.
  • Bounded retention, anonymisation for analytical uses, and logging of who queries the history.

Full detail in how worker data is handled.

When it is worth adding

Not in every case. It pays off when at least one of these holds:

  • A meaningful share of patrol work happens on board and not only in stations.
  • You need to evidence coverage per train for a service contract or for the asset owner.
  • You want to avoid the full tunnel radio survey and keep mapping to stations.
  • Response to an on-board alert is slow today because nobody knows which train to look in.

If the service is purely platform and concourse, the base layer alone covers it and adding beacons is spend without return. We say so before selling, not after.

Shall we try it on one class of stock?

An on-board pilot fits into one train type and a handful of units. It is the cheapest way to see whether it changes your operation.

We reply within one business day.