Domain
Advanced Manufacturing
Making things that still work after Tuesday, at a scale that would have impressed Brunel.
From digital twins to toolpath poetry. Manufacturing at its best is not nostalgia — it is the quiet art of making the difficult thing repeatable.
Britain makes less than it once did and considerably better than it tells anyone. The difficulty is rarely the first article. A capable team in a decent workshop can produce one of almost anything. The difficulty is the fiftieth, made on a different machine, by someone on a night shift, from a powder lot that arrived late.
That gap — between a part that exists and a process that can be trusted — is where most manufacturing innovation actually lives, and it is chronically underfunded. Universities are rewarded for the novel process. Customers pay for the qualified one. The interesting work is the translation, and it usually needs a factory-shaped room rather than a seminar.
The other shift is data. Every plant now has more sensors than it has people who trust them. The useful projects are not the ones that add a dashboard; they are the ones that let a quality engineer act on a number, defend the decision, and get the same answer next week.
- Additive, subtractive, and the argument between them
- Factory data that means something on the shop floor
- Supply-chain resilience that is not a slogan
- Process qualification, not just process demonstration
- Metrology that survives a customer audit
What sits inside
The sub-disciplines people actually search for.
Additive manufacturing
Powder-bed metal, wire-arc, and polymer systems — together with the heat treatment, machining and inspection that consume most of the budget and all of the schedule.
Machining, tooling and forming
Five-axis work, fixturing, tool life, and the composites lay-up problems that look solved in a render and less so on a mandrel.
Automation and robotics
Cells for tasks that are dull, dangerous or too precise for a long shift. Increasingly compliant, increasingly asked to work near people.
Metrology and inspection
CT, structured light, on-machine probing, and the awkward question of whether the measurement is better than the tolerance it polices.
Digital manufacturing
Process monitoring, traceability and models that predict a defect early enough for someone to do something about it.
Where the UK stands
Without the boosterism, and without the gloom.
- The High Value Manufacturing Catapult's seven centres exist precisely for the translation problem — the AMRC in Sheffield, the MTC in Coventry, the NCC in Bristol, WMG, CPI, Nuclear AMRC and the AFRC.
- The supply base is dominated by capable SMEs who cannot afford a failed qualification, which makes de-risked collaboration unusually valuable to them.
- Aerospace and automotive still set the tolerance culture, and increasingly the energy sector is inheriting it.
- The binding constraint is rarely capital equipment. It is people who can commission, qualify and argue for a process.
Work that gets commissioned
The shape of a package in this domain.
- First-article production and parameter development, off the customer's critical path
- Process capability studies against a named specification
- Inspection method development, including the measurement-uncertainty argument
- Failure analysis on parts that passed every test except the last one
- Factory data pilots scoped to one line and one decision
Where projects come unstuck
Reliably, and usually in week nine.
- Budgeting the printer and forgetting the post-processing, which is usually the larger number
- Qualifying a process on one machine, then discovering that machines have personalities
- A demonstration cell that has never met a night shift, a tired operator, or a Friday
- Treating metrology as a footnote until a customer asks how the tolerance is policed