In short: UK pharmaceutical manufacturing now lives or dies on the audit trail — and the regulatory regime that governs it, from GMP Annex 1 to ALCOA+ data integrity, assumes a level of continuous, trustworthy connectivity that cleanrooms were never designed to carry. Private 5G is the quiet way to close that gap without putting cables through a sterile envelope.
Key Takeaways
- The cleanroom is the worst place to run a cable — every penetration through a sterile envelope is a contamination risk and a validation cost, which is exactly why wireless-first connectivity is gaining ground in GMP manufacturing.
- Data integrity is now a network problem — ALCOA+ expectations and the revised Annex 1 mean batch records, environmental monitoring and equipment telemetry all have to reach the system of record contemporaneously and without gaps.
- Serialisation and track-and-trace need density — FMD-style unique identifiers, aggregation scanning and cold-chain telemetry put thousands of read events on the network per shift, a load that contended Wi-Fi struggles to guarantee.
In a nutshell
The Medicines and Healthcare products Regulatory Agency does not inspect a UK pharmaceutical plant the way a fire officer inspects a warehouse. The inspector is looking for one thing above all others: can you prove, beyond reasonable doubt, that every batch was made the way your licence says it was made, and that the record of how it was made has not been altered, lost or quietly reconstructed after the fact? That single question — the integrity of the record — has reshaped how modern medicines factories are built, instrumented and connected. We have spent the last few years watching the connectivity layer move from an afterthought to a regulated dependency, and the cleanroom is where the tension is sharpest.
This post sets out why pharmaceutical manufacturing has become a continuous-data operation, why the cleanroom is uniquely hostile to the conventional ways of carrying that data, and where private 5G fits — not as a novelty, but as the most practical answer to a regulatory problem the sector already has.
The record is the product
In most manufacturing, the physical output is the thing you sell and the data is a useful by-product. In regulated pharmaceutical manufacturing it is closer to the reverse. A vial of sterile injectable is worthless — legally unsellable — without the batch record that accompanies it: the environmental monitoring data for the room it was filled in, the calibration status of the equipment, the operator sign-offs, the in-process checks, the deviation log. The European GMP framework that the UK continues to follow (EudraLex Volume 4) treats that documentation as part of the product, not a description of it.
The principle that governs the record is captured in the industry shorthand ALCOA+: data must be Attributable, Legible, Contemporaneous, Original and Accurate, with the later additions of Complete, Consistent, Enduring and Available. The word that quietly carries the most operational weight is contemporaneous. It means the record has to be made at the time the thing happened — not written up at the end of the shift from memory or transcribed from a paper note found in a gowning lobby. Contemporaneous recording at scale is only possible if the instrument that takes the reading can reach the system that stores it, reliably, the moment the reading is taken. That is a connectivity requirement dressed up as a compliance one.
Annex 1 raised the bar on the room itself
The revised GMP Annex 1, governing the manufacture of sterile medicinal products, came fully into force in 2023 and pushed the sector further down the road of continuous monitoring. It formalised the expectation of a Contamination Control Strategy — a documented, holistic view of every way contamination could enter the process — and leaned heavily on continuous environmental monitoring of grade A and B zones, particle counting, and the kind of real-time data that lets a site demonstrate control rather than merely sample for it.
Continuous monitoring means continuous data. A modern aseptic filling line carries particle counters, differential pressure sensors, temperature and humidity probes, airflow visualisation, and increasingly camera-based systems watching for interventions at the point of fill. Each of these is a stream that has to land in a validated historian without dropouts, because a gap in grade A monitoring data during a fill is not a minor IT incident — it is a question the inspector will ask, and a potential reason to reject the batch. The room got smarter and far more instrumented; the network underneath it has to keep up.
Why the cleanroom resists the obvious answers
Here is the awkward part. The two conventional ways of carrying that data — structured cabling and Wi-Fi — both fight the cleanroom.
Cabling fights it physically. A cleanroom is a sealed, pressure-controlled envelope with smooth, cleanable surfaces and a tightly validated air regime. Every cable that crosses the boundary is a penetration that has to be sealed, cleaned and, crucially, validated — and every change to it later means re-validation. Trunking and cable trays are dirt traps in an environment whose whole purpose is the absence of dirt traps. Adding a sensor or moving a camera in a hard-wired cleanroom is rarely a quick job; it is a change-control process with documentation attached. The result is that instrumentation tends to get frozen at design time, which is the opposite of what a continuous-improvement, continuous-monitoring regime wants.
Wi-Fi fights it electromagnetically and administratively. Cleanrooms are full of stainless steel — walls, benches, isolators, RABS (restricted access barrier systems) — which is close to a worst case for 2.4 and 5 GHz propagation. Signals reflect, fade and drop in ways that are hard to predict and harder to validate. And in a regulated environment, the unmanaged, best-effort nature of contended Wi-Fi is itself a problem: when the network is shared with everything else on site and offers no guarantee that a grade A particle reading will get through ahead of someone's software update, you cannot easily argue that your monitoring data is complete and available in the ALCOA+ sense.
Where private 5G changes the calculation
Private 5G addresses the cleanroom problem from a different direction. Because it is a licensed or shared-spectrum cellular network that the site controls, it offers three things the regulated environment specifically values.
First, deterministic behaviour. A private 5G network can be engineered to prioritise traffic — network slicing and quality-of-service mean that environmental monitoring and equipment telemetry can be given guaranteed treatment, separated from camera streams, separated again from general site IT. That separation is exactly the kind of documented, designed control that a Contamination Control Strategy and a data-integrity argument are built on. You are no longer hoping the reading gets through; you are designing the path it takes.
Second, coverage through the steel. Cellular systems are built to penetrate and to hand over cleanly between cells as a device moves. A small number of well-placed indoor radios can blanket a filling suite, the gowning lobbies, the warehouse and the QC labs with one continuous network — through the stainless, around the isolators — rather than the patchwork of access points and dead spots that Wi-Fi tends to produce in those rooms.
Third, fewer penetrations. This is the underrated one. A wireless-first cleanroom needs power to its radios and very little else crossing the envelope. Sensors, mobile devices, scanners and cameras connect over the air. Reconfiguring a line — adding a particle counter here, moving a camera there — becomes a change to a managed wireless network rather than a re-cabling job through a validated boundary. For a sector where the cost of change is dominated by validation effort, that shift is material.
Serialisation, aggregation and the cold chain
Annex 1 and data integrity are the regulatory drivers, but they are not the only load on a pharma plant's network. Falsified-medicines controls require serialisation: a unique identifier on each saleable pack, scanned and verified as it moves through pack-out, aggregation into cases and pallets, and dispatch. A busy packing hall generates a continuous stream of read events — thousands per shift on a fast line — each of which has to be captured, checked against a database and committed. Aggregation, where the system learns which units are in which case in which pallet, is unforgiving of a missed scan; a broken parent-child relationship at the pallet level can hold a shipment.
Add the cold chain — temperature-logged storage and transport for biologics and vaccines, where an excursion can write off a high-value batch — and you have a second, dense, latency-sensitive workload sitting alongside the manufacturing data. UK sites from large originator plants such as GSK at Barnard Castle and AstraZeneca at Macclesfield through to the growing base of contract manufacturers (CDMOs) running multi-product, fast-changeover operations all carry some version of this combined load. It is exactly the kind of high-density, mixed-criticality traffic that private 5G was designed to keep separate and predictable.
The validation question, answered honestly
We should be candid about the trade-off, because the sector rightly is. Introducing any new system into a GMP environment carries a validation burden, and a private network is no exception — it has to be qualified, documented and managed under the same computerised-systems and GAMP 5 expectations as anything else that touches the data. There is no version of this where you install a network on Friday and start making sterile product on Monday. Anyone who tells a quality director otherwise is not worth listening to.
The honest case for private 5G is not that it avoids validation. It is that it reduces the recurring validation cost — the change control, the re-cabling, the disturbance to a sealed envelope every time the process improves — by moving connectivity off the physical fabric of the cleanroom and onto a managed, controllable wireless layer that was built for determinism. You validate the network once, properly, and then you change what runs on it far more freely than you could ever change a hard-wired room.
What we conclude
UK pharmaceutical manufacturing has, over the last decade, turned the integrity and availability of its data into a first-order regulatory obligation, and the revised Annex 1 has made continuous monitoring of the cleanroom itself a baseline expectation. That combination puts a demand on connectivity that the two obvious answers — cabling and Wi-Fi — both struggle to meet inside a sterile envelope, one for physical reasons and one for electromagnetic and administrative ones.
We conclude that private 5G is not a futuristic addition to the pharma plant but a sober response to a problem the sector already has: how to carry trustworthy, contemporaneous, complete data out of the hardest room in the building to wire, without putting more penetrations through walls whose entire purpose is to keep contamination out. The medicines are the visible product. The record is the regulated one — and increasingly, the network is what makes the record defensible.