The £10M Production Facility That Proved We Could Compete

The Japanese equipment was supposed to arrive in March 2021. Then April. Then May. By the time the containers finally showed up in September—six months late—we'd already built the Production Hall and were just... waiting. The building sat ready, HVAC commissioned, utilities in place, while somewhere in the global shipping chaos our £10M bet on UK manufacturing automation was stuck on a boat.

I'd spent nearly a decade at the business running capital projects and technology investments. We were a print company—sophisticated finishing equipment, the kind of manufacturing that requires precision and coordination. When COVID hit and the government put out the call for UK PPE manufacturers, the owners started exploring it in March 2020. By June, when I got pulled into the project, we were already committed to building a medical manufacturing business from scratch.

The first time I visited the site post-lockdown, I was sketching factory layouts in a notebook. We had no clean rooms, no medical-grade equipment, no idea how to make masks at scale. We just knew we needed to start.

Building Whilst Learning

The initial approach was pure pragmatism—get machines, build rooms to fit them, figure it out as we go. We were airfreighting equipment from China and constructing clean rooms around whatever arrived. It worked, in the sense that we got masks made, but it was chaotic. Every room was slightly different. Every machine had its own quirks. Training was bespoke. HVAC was one-off.

We built Hall A, then Hall B, and somewhere between the two I realised we were creating a nightmare for ourselves. If every room was unique, we'd never scale. So we stopped, and designed properly.

We worked it out in CAD. A standard configuration: seven machines per hall, standardised HVAC, repeatable layout. Once we had the template, Halls C through F became carbon copies. About 28 people per hall—3 operators per machine, 3 QC staff. The design work upfront meant we could build and commission rooms in parallel. Get one right, then replicate it.

This wasn't something I invented. I'd been reading Bent Flyvbjerg's research on major projects in How Big Things Get Done, and this was exactly the school building story he tells. Standardise the design, then replicate it. Remove variation, remove risk. It felt obvious once we did it, but it took us two attempts to realise we needed it.

By September 2020, we were adding an industrial laundry. October brought a high-density warehouse. November meant a full training team and an incubator cell for new staff. We were running 24/7, pumping out masks, and the modular approach was working.

The Decision to Automate

By December 2020, we had a problem that wasn't obvious to everyone else. We were running 24/7, hitting our volume targets, keeping pace with the government contract. But we could see the structural issue: we couldn't compete with Far East labour costs using the same equipment they used. We had Chinese machines requiring three operators each. They had the same machines with labour costs a fraction of ours.

The only way UK manufacturing could compete was efficiency through automation. Not just slightly better—dramatically better. We needed equipment that could produce the same output with a quarter of the staff.

So we made the decision to invest £10M in a Production Hall with fundamentally different machines. It wasn't just an upgrade—it was a bet that there would be a market for high-quality European PPE beyond COVID, and that automation could make UK manufacturing viable long-term.

We identified two vendors. The Japanese machine produced standard TypeIIR surgical masks—tried and tested technology that had been running in Japan throughout their COVID response. Proven, reliable, known output. The German supplier had experience with FFP2 production and had previously made envelope window insertion machines for the mailing industry. We were designing masks with transparent panels, and their background seemed like a good fit for what we needed.

Two vendors, two continents, two different technologies. We placed the orders and started building the Production Hall.

The Reality of £10M Projects During a Pandemic

The building was ready by early 2021. The structure, the clean room specifications, the utilities, the HVAC—everything was commissioned and waiting. And then we waited. And waited.

Global shipping was completely overwhelmed. Container costs had gone through the roof. Equipment that should have arrived in March was delayed to April, then May, then June. We had a building we couldn't use and equipment we couldn't access. There was nothing to do but wait and hope the shipments would eventually arrive.

They did, in September 2021, nine months after we'd ordered them. The high-end production equipment was impressive—when it finally showed up—these were sophisticated machines, nothing like the varying states of the early Chinese equipment we'd been dealing with. But even quality equipment needs adaptation. The Japanese machines were spec'd for 200V three-phase power—Japanese electrical standard—which meant we were doing voltage conversions and rewiring before we could even test them. Servo motors came with different software versions than we'd been told to expect.

Not to mention, we were doing all of this whilst running a medical manufacturing business in the same building. We had clean room protocols, contamination control, quality standards that couldn't slip whilst we were also managing a construction site installing heavy equipment. The disciplines required were completely opposite—medical-grade cleanliness versus construction dust and noise—and we had to maintain both simultaneously. It required obsessive organisation and separation of zones.

The coordination challenge wasn't like smaller projects I'd run. We had electrical contractors, equipment installers, HVAC specialists, automation engineers, all needing to work in sequence but also in parallel because we couldn't afford delays. Two international equipment suppliers who'd never worked together, pandemic travel restrictions preventing site visits, everything happening over video calls and emailed technical drawings.

What We Built

By late 2021, the Production Hall was operational. Two lines of new equipment, seven people per line. Fourteen people total, producing the same output as two of our modular halls—which would have required 56 people.

Four times the labour efficiency. The Japanese line ran TypeIIR surgical masks at the volume and quality we needed. The automation worked. The bet had paid off technically.

I handed the ramp-up over to the operations team and moved on to other projects in the business. The Production Hall was running. We'd proven the model.

What Happened Next

After the government contract ended, the British government decided to offshore all PPE procurement again. Domestic manufacturing, regardless of capability or efficiency, wasn't part of their long-term strategy. The business was mothballed.

It was heartbreaking. We'd built something that worked. We'd proven you could compete with Far East efficiency if you invested in the right automation. We'd shown you could stand up medical-grade manufacturing in the UK in months, not years. Yet none of it mattered because the purchasing decision was made on cost alone, without factoring in supply chain resilience or domestic capability.

The team was redeployed to other parts of the business. I left in 2023.

What I Learnt About Major Projects

Flyvbjerg was right about modularity in How Big Things Get Done, the standardised halls—C through F—were the reason we could scale as fast as we did. Every time we built a room, we were building reference-class experience for the next one. Our estimates got better. Our installation times shortened. Our problems became predictable. That's the lesson from his schools research, and it held true for clean rooms just as much as classrooms. But we only got there by designing it properly in CAD after the chaos of Halls A and B. The design work upfront was what made the speed possible later.

Flyvbjerg's other big finding in How Big Things Get Done—that most major projects fail because of optimism bias—that one caught us too. We were absolutely full of optimism. We believed we could install equipment on the same timeline as our previous machines, even though these were new technologies from international vendors we'd never worked with. The ramp-up took longer than we expected because we were learning new systems, not replicating known ones. The modular approach de-risked the building programme, but it couldn't de-risk the novel technology.

What surprised me most wasn't the scale of the project—I'd run complex capital investments before—it was the external dependencies I couldn't engineer away. I could plan the building programme, manage the contractors, and coordinate the utilities. But I couldn't control global shipping. I couldn't speed up international freight. I couldn't make pandemic travel restrictions disappear so suppliers could do site visits.

The discipline I learnt from years of capital projects—plan thoroughly, then adapt constantly—didn't change, but the adaptation required was different. In previous projects, when something went wrong, I could usually find a way around it. Different supplier, different approach, throw resources at it. With shipping delays and international logistics chaos, there was no workaround. You just had to wait and keep everything else ready so you could move fast when the equipment finally arrived.

If I were to do this again—a similar major capital project, not this specific one—I'd insist on more site visits from suppliers during the design phase. Except I couldn't have done that in 2020-2021 because the world was locked down. I'd insist on earlier equipment delivery to allow testing time. Except global shipping was overwhelmed and nothing was moving on schedule. The honest answer is there isn't much I'd do differently on the project itself. We built it well, we built it fast, we used modularity to de-risk what we could control.

What I couldn't control was the policy environment we were building into.

What It Meant

There were a lot of negative stories during COVID; about PPE suppliers ripping off the government, delivering substandard equipment, profiteering from the crisis. We weren't that. We made a quality product in volume. We invested £10M in long-term manufacturing capability when we could have just run cheap equipment and maximised short-term profit. We made a genuine contribution to the UK's COVID response.

Success isn't just delivering the project. It's delivering into a context that values what you've built. The Production Hall delivered on time—well, as much as any project facing nine-month shipping delays can be "on time"—on spec, and the equipment worked. The efficiency model proved out. By traditional project metrics, it was a success.

But we delivered manufacturing capability into a policy environment that chose offshore procurement as soon as the immediate crisis passed. That's not a project management failure, that's a lesson about which risks you can engineer away and which ones you can't.

The capability we built was real. Fourteen people doing the work of 56, medical-grade quality, fast scaling through modularity. That all happened. That all worked. The fact that it's not being used now doesn't erase what we proved was possible.

I wish the story had a different ending. I wish there was still a market for UK-manufactured PPE that valued efficiency and quality over lowest cost. I wish the Production Hall was still running.

But major projects aren't just about schedule and budget and technical delivery. They're about understanding the context you're building into, knowing which dependencies are within your control and which aren't, and being honest about what success means when external forces determine whether your work continues or stops.

We built something impressive. It just doesn't exist anymore. And somehow, both of those things are true.