Buying decision
RTLS with tags, or on smartphones
A total cost comparison between a beacon-and-tag location system and one that reuses the Wi-Fi and phones already in place.
When one indoor location system is compared with another, the conversation usually starts and ends with accuracy. It is the number on the first slide and the one everybody remembers.
It is also the metric that worst predicts whether the project will go anywhere.
What decides that is not how many centimetres of error the system has, but what it costs to reach that accuracy and what it costs to keep it running every day. This article breaks that sum down.
The two architectures
Beacons and tags
The classic approach, used by UWB systems and most Bluetooth angle-of-arrival systems.
Fixed anchors are installed across the space to be covered: devices that must be mounted, powered and connected. And an active tag is issued to every person or asset to be located: a battery-powered device that must be handed out, charged and replaced.
It delivers centimetre accuracy. It is the right answer in a factory, an automated warehouse or an operating theatre.
On existing infrastructure
The alternative installs nothing. It uses the Wi-Fi access points already mounted as the measurement reference, and the corporate phone the person already carries as the measuring device.
It delivers metres, not centimetres. And that is the whole trade-off.
The sum nobody quite does
Let us put approximate numbers on a concrete case: a metro line with twelve stations and seventy officers to cover. These are orders of magnitude, not a quotation: they shift with vendor, country and year. What does not shift is the proportion between line items.
Cost of entry
| Item | With beacons | On existing infrastructure | |---|---|---| | Anchors or beacons | Dozens per station, from several hundred to a few thousand euros each | None | | Installation and cabling | Per installed point, plus power and conduit | None | | Night-time work windows | Required: you cannot drill a platform in service | Not applicable | | Permits and works coordination | Required | Not applicable | | Tags | One per officer, plus spares | None | | Terminals | — | The ones already in service | | Platform and licences | Yes | Yes |
The line item that surprises people is not the hardware: it is the work window. Installing on a network in service means working through the night, with permits, with safety coordination, and at a pace set by the operator’s calendar rather than the installer’s. That is what turns weeks into months.
Cost of running it, year on year
This is where they really diverge, and the part that rarely makes the initial comparison.
With tags, a new process appears that did not exist before and that somebody has to sustain every single day:
- Issue the tag at the start of the shift and collect it at the end.
- Charge it. Seventy tags are seventy batteries to manage.
- Replace those lost, dropped or failed. Annual replacement rates in field environments are not negligible.
- Audit that each tag is where it should be and assigned to who it should be.
- Replace non-removable batteries at end of life.
That process costs somebody’s hours, and it has a silent failure mode: the day the tag is not issued, the system does not report missing information; it reports that the person is not there. A system where people sometimes vanish without warning stops being consulted.
On existing infrastructure, that process does not exist. The phone is already issued, already charged and already replaced, because it was the work tool before the project started. Location adds no new object to the officer’s day.
Cost of growing
| | With beacons | On existing infrastructure | |---|---|---| | Adding a station | A full installation project | Check the APs and load the topology | | Adding twenty officers | Twenty more tags and their management | Install the app on twenty phones | | Replicating on another line | Another infrastructure project | Survey and calibration | | Changing network vendor | Not applicable, it is a closed system | Any FTM-capable AP works |
A beacon system scales linearly in cost: every new square metre costs what the last one did. One that leans on existing infrastructure scales almost flat, because the main spend was already made for another reason.
When fine accuracy genuinely pays
It would be dishonest to present this as if one architecture always won. There are cases where UWB is simply correct:
- When you must distinguish two positions less than a metre apart: which machine, which shelf, which bay.
- When what you are locating are objects that carry no phone.
- When the space is small and controlled, and installation does not compete with a running operation.
- When a regulatory requirement fixes an accuracy that metres cannot meet.
The right question is not which is better, but what resolution the decision made with that data actually needs.
And in operational security across a transport network, that decision is: where do I send support and which entrance do they use. For that, knowing the zone and the direction of travel is enough. Knowing the tile adds nothing, and paying for it in civil works and in the daily management of seventy devices definitely subtracts.
The one-line summary
A beacon system gives you centimetres in exchange for an infrastructure project and a permanent operational process. One on existing infrastructure gives you metres in exchange for almost nothing.
If your operational decision is made in metres, the second option is substantially cheaper to buy, far cheaper to maintain and dramatically faster to deploy.