In short: Hospital equipment tracking fails far more often on infrastructure than on software. The tags cost a few pounds each; the coverage that has to hear them, across old buildings, basements, lift lobbies and car parks, is the part nobody costs properly, and it decides whether the system reports a location or a gap.
Key Takeaways
- Nurses hunt, trusts over-buy — when a ward cannot find a pump it borrows, hoards or rents another, so one shortage bills twice, once in lost clinical time and once on a rental invoice.
- Tags are cheap, coverage is not — a BLE or UWB tag costs a few pounds, but the gateways that hear it need power and a network drop in every corridor, lift lobby, plant room and basement store on the site.
- Accuracy tiers decide the budget — room-level tracking for recalls needs a fraction of the infrastructure that sub-metre workflow tracking does, and buying the wrong tier is the most expensive mistake in the project.
In a nutshell
What is hospital equipment tracking?
Hospital equipment tracking means tagging mobile clinical assets, then using a wireless system to report where each one is against a single asset register. The assets are the ones that walk: infusion pumps and syringe drivers, telemetry units, bladder scanners, pressure-relieving mattresses, beds, wheelchairs, ultrasound trolleys. Vendors sell it as RTLS, a real-time location system, and trusts buy it for three reasons. Staff stop hunting. The rental and over-purchase bill falls. And when the MHRA publishes a Field Safety Notice, the medical physics team can answer the only question that matters, which is where every affected device is at this moment.
The business case is usually sound. The reason so many of these projects stall after the pilot ward is that the location layer underneath them is treated as a software purchase when it is a building services problem.
Why can staff never find the pumps?
Because mobile equipment behaves exactly as you would expect it to when nobody can see it. A pump follows a patient to theatre, comes back on a different ward, is put in a bay because the ward is busy, and is then unavailable to the equipment library that owns it. The ward that needs one next does not search the estate; it borrows from a neighbour, hides two in a store cupboard for next time, or raises a rental. Every one of those responses is rational for the ward and expensive for the trust.
Lord Carter's 2016 review of operational productivity in English NHS acute hospitals put a number on the general problem, identifying around £5bn of annual efficiencies available by 2020/21 and pointing directly at inventory and asset management as part of it. The Department of Health's response was Scan4Safety, which ran across six demonstrator trusts between 2016 and 2019: Derby, Leeds, North Tees and Hartlepool, Plymouth, Royal Cornwall and Salisbury. The published evidence from that programme reported millions of pounds of benefit per trust and tens of thousands of hours of clinical time released back to patient care. GS1 and PEPPOL adoption were then written into NHS information standards as DCB1077 and DCB1078.
So the policy direction has been settled for the better part of a decade. What has not been settled, trust by trust, is how the location data actually gets out of a building that was in many cases finished before the NHS existed.
Why do RTLS rollouts stall after the pilot ward?
A pilot ward is the easiest environment in the hospital. It is one floor, one corridor, recently refurbished, with structured cabling and decent WLAN because the electronic patient record demanded it. Put thirty tags and eight gateways in and the system works beautifully, which is precisely why the demonstration is always done there.
Then the scope moves to the rest of the estate, and the estate answers back. The 2023/24 ERIC return put the NHS England maintenance backlog at around £13.8bn, and a meaningful share of the estate predates 1948. That means solid masonry, lead-lined imaging suites, lift shafts, service tunnels, plant rooms, basement stores, modular decant buildings in the car park, and the loading bay where the equipment library actually receives and dispatches. Those are not edge cases. They are where equipment spends a large part of its life, and they are the places where there is no spare network port, no accessible containment, and an asbestos survey between you and the ceiling void.
The cost per gateway then stops being the price of the gateway. A BLE receiver might be £150. Getting a PoE drop to it in a listed Victorian corridor, out of hours, with an infection control permit and a sixteen-week lead time on the containment works, is routinely several times that. Multiply by a few hundred locations and the infrastructure line dwarfs the software licence that started the conversation. We have watched this arithmetic kill more than one otherwise sensible project.
How accurate does the tracking actually need to be?
This is the question that decides the budget, and it should be answered per asset class rather than for the whole trust.
Room-level accuracy, which tells you a device is in Ward 12 or in the decontamination unit, is enough for recall management, maintenance scheduling, utilisation reporting and finding the equipment library's missing stock. It is achievable with passive RFID at choke points, BLE gateways at ward and department granularity, or infrared where a hard room boundary matters. Gateway density is modest.
Sub-metre accuracy, which tells you a pump is in bay 3 rather than bay 4, is what workflow automation and hands-free clinical documentation need. It generally means ultra-wideband, several anchors with overlapping coverage in every tracked space, and cabling density closer to a lighting circuit than a data network.
The two differ by an order of magnitude in installed cost. Trusts that specify sub-metre accuracy across the estate because the demonstration looked impressive tend to discover the number at business case stage and stop. Trusts that specify room-level for the whole estate and sub-metre for the handful of areas that earn it, typically theatres, ED and sterile services, get something built.
What does the tracking layer need from the site?
Three things, and they are all unglamorous.
First, continuous coverage rather than good coverage. A tracking system is only as useful as its worst blind spot, because that is where the equipment that nobody can find will be. A system that covers 85% of the estate does not deliver 85% of the benefit; it delivers a register that staff learn not to trust, which is worth close to nothing.
Second, backhaul for the gateways. Every receiver, of whatever radio type, has to get its observations back to the location engine. In a new-build wing that is a patch panel. In a 1930s block, a car park modular unit, or a community site two miles away that shares the same equipment pool, it is the whole problem.
Third, power. PoE solves this where structured cabling exists and creates the difficulty where it does not, which is why battery-powered gateways with a cellular uplink keep reappearing in the more difficult parts of these designs.
Where does private 5G fit?
It belongs in the backhaul and coverage layer rather than the location engine, and the distinction is worth holding onto. A private 5G network does not give you sub-metre asset location on its own. What it gives you is a single licensed radio layer, under the trust's control, that reaches the parts of the estate structured cabling has never reached, and that gateways, handhelds and increasingly the tags themselves can attach to with a SIM instead of a network port.
That matters in four specific situations. Where cabling a corridor costs more than the equipment it would serve, a cellular gateway uplink removes the containment works entirely. Where the estate is dispersed, with car parks, ambulance bays, modular wards, community clinics and an off-site equipment library all holding trust assets, one network covers outdoor and inter-building space that WLAN was never designed for. Where the same infrastructure can carry porters' handhelds, bed management terminals and the tracking traffic together, the business case stops belonging to one project. And 3GPP Release 17 RedCap devices are now bringing cellular into the power and cost envelope that battery-powered tags and low-cost gateways require, which was the honest objection to cellular in this role three years ago.
Where private 5G is the wrong answer is equally clear. A single modern building with good Wi-Fi 6 coverage and spare ports should use what it has. A trust that needs sub-metre accuracy in theatres needs UWB anchors there whatever else it does. The sensible pattern we keep arriving at is a hybrid: use the existing WLAN where it is already good, use private 5G for the awkward estate and the outdoor and inter-site gaps, and reserve high-density UWB for the small number of spaces that justify it.
What to settle before buying tags
Four things, in this order, and all of them before the procurement.
Survey the estate for coverage rather than floor area. Walk the routes the equipment actually takes, including the lift, the service corridor, the loading bay and the car park, and record where a tag would go silent. That map, not the ward count, sizes the project.
Set an accuracy tier per asset class and write it into the specification. Pumps and mattresses almost always want room level. Anything driving an automated workflow wants sub-metre. Mixing them is fine; pretending one tier fits everything is not.
Cost the installation, not the hardware. Ask any vendor for an installed cost per gateway in a pre-1948 building with no spare containment, out of hours, and treat the answer as the real unit price.
Name the owner of the asset register. An RTLS reports locations; it does not reconcile them against a register that estates, medical physics and the equipment library each maintain differently. Trusts that fix the register first get value from the tags in months. Trusts that buy tags first spend the first year doing data cleansing they could have done without a single gateway.
If you are scoping equipment tracking across a dispersed NHS site and want the coverage layer costed before the tags are ordered, get in touch. We survey the estate, the outdoor space and the satellite sites together, and we will tell you where your existing WLAN is already good enough. Read more on our health and social care sector page.