Start withthe problem.
Define what actually needs to exist before deciding how to build it.
The first question is not how something will be manufactured. It is what the product has to be — for whom, against what problem, at what cost, under what constraints. A product defined only as a shape is a product that will be redefined later, expensively.
- What is being created, and what it has to do
- Who it is for, and the sector it serves
- The problem it solves and how it will be used
- Desired features and what matters most among them
Turn the idea intoquestions we can answer.
Convert intent into technical direction, and name what is still unknown.
An idea becomes a project when its uncertainties are written down. This stage is about establishing what the product requires technically, what remains unresolved, and which of those unknowns have to be settled before anything can be committed to.
- Design refinement against real constraints
- Mechanical requirements and how the product holds together
- Material considerations and what properties actually matter
- Component requirements, and what must be specified rather than assumed
OutputDefined technical direction
Make ittangible.
Physical proof, and an honest account of what is still undefined.
Prototypes answer questions that drawings cannot, but they answer them one at a time. A concept visual shows what a product might look like. A functional prototype lets behaviour and experience be evaluated. Neither, on its own, makes a product ready to manufacture — readiness is a separate question about how much has actually been defined.
- Concept visual — what might it look like
- Functional prototype — can the intended behaviour and experience be evaluated
- Production-ready direction — what still has to be defined before manufacturing decisions can responsibly be made
- Specifications, materials and components
Prove theprocess.
An initial run exposes what prototypes cannot.
A prototype is usually built once, carefully, by someone who understands it. Production is built repeatedly, at pace, by people working from documentation. A limited first run is where the difference between those two things becomes visible — in assembly, in consistency, in the things nobody thought to write down.
- Repeatability — does the second one match the first
- Assembly sequence and whether it can be followed as written
- The production process itself, observed rather than assumed
- Quality criteria applied to real output
Quantity, tooling, cost, lead time and factory requirements depend on the specific product and the manufacturing path it takes. There is no standard run and no standard timeline.
OutputValidated production direction
Build forrepeatability.
From a proven process to production that holds over time.
Repeatable manufacturing is not a phase that begins after development ends. It is the accumulated result of every decision taken earlier — what was specified, what was left open, what was approved and what was assumed. Products stay consistent when those decisions were made deliberately.
- Approved specifications as the reference everything measures against
- Production planning against volume and timing
- Quality expectations agreed before, not negotiated after
- Repeat orders and what has to stay identical between them