Design for Manufacture and Assembly, Explained.

Design for manufacture and assembly rests on a plain idea. The process that builds a product should be designed at the same time as the product, not after it.

I first studied this as a mechanical engineering discipline. I have spent the past decade applying it well outside a factory. It is the origin of how I work, so it is worth setting out properly instead of offering it as a credential.

The failure it removes

In sequential development, design finishes and manufacturing inherits. Drawings complete. Intent fixed. The people who have to build it are then asked how.

They find the things that are always found at that point.

A tolerance that needs a process nobody has. An assembly order that needs a hand to reach somewhere a hand does not fit. A fastener specified in four sizes where one would do. A part that cannot be tested until it is enclosed, so a failure means scrapping everything around it.

Each of these is fixable. Each fix reopens a design decision everybody treated as closed. That means redrawing, reapproving, retooling, requalifying.

The cost of a change climbs steeply the later it is made. Sequential development guarantees the changes arrive late.

What changes in practice

Part count becomes a target. Fewer parts means fewer suppliers, fewer purchase orders, fewer inspection points, fewer operations, fewer ways to fail. Cutting part count is usually the biggest saving available, and it is only available early.

Assembly order gets designed, not discovered. Which face goes down. What is reachable when. Whether a component can go in the wrong way round. Design a part so it physically cannot be fitted backwards and you have removed a quality check permanently.

Testability goes in from the start. If a subassembly cannot be tested before it is enclosed, every later failure destroys the value added after that point.

Tolerances get argued against real process capability. A tolerance is a cost. Specifying one tighter than the function needs buys nothing.

The people who build it sit in the room during design. This is the organisational change and it is the hard one. It means manufacturing, quality, sourcing and design working together at a stage where designers usually want to be left alone.

The honest cost

Concurrency is not free. The trade should be stated plainly.

It costs more upfront. Coordination time. Meetings that feel premature. Early decisions made on less information than a sequential process would have gathered by then. It needs manufacturing input before there is much of a design to comment on, which some engineers find a waste of an afternoon.

It also asks functions with different incentives to reconcile early. Where sourcing is measured on unit price and design is measured on schedule, that reconciliation is a management problem before it is a technical one.

What you buy is pieces that fit by construction instead of by late correction.

Whether the trade is worth it depends on how expensive change is in your programme. Soft tooling and low volumes make sequential development survivable. Hard tooling, high volumes or regulated qualification make late change the dominant cost, and concurrency pays for itself before the first production run.

Why I use it outside manufacturing

The idea travels. That is why I built a strategy method on it.

A commercial plan has the same structure as a product. A proposition. A process that delivers it. Economics that have to reconcile. Design those one after another and they will not fit, for exactly the reasons a sequentially designed product does not fit.

That is my whole method. It came from a factory discipline, not from a strategy framework.

Common questions

What is design for manufacture and assembly? It is designing a product and the process that manufactures and assembles it at the same time, so manufacturability shapes the design instead of being discovered after it is fixed. The usual measures are lower part count, a simpler assembly sequence, and tolerances matched to real process capability.

Why does design for manufacture reduce cost? Mostly by removing parts, operations and inspection steps before they are designed in, and by preventing the late engineering changes that force retooling and requalification. The saving comes from timing as much as from the design.

When should design for manufacture start? At concept stage, alongside the first functional design work, because the decisions with the biggest manufacturing consequences are made earliest. Introduce it after specification freeze and it stops being a design method. It becomes a rework exercise.

A test for your programme

Ask who from manufacturing has been to a design review in the last month. Then ask whether they were there to comment or to be told.

The answer tells you which process you are running, whatever the process document says.

The Concurrent Method sets out how this discipline extends from the factory to a commercial strategy.

---

Dainu Devis is a business strategist, serial entrepreneur and founder of Sharktech Global. His method started as an engineering specialisation in concurrent product and process design. For the past decade he has used it as an operator and adviser across manufacturing, retail, construction and critical infrastructure, working with boards, technology companies, startups and industrial operators. Go-to-market delivery runs through Divine Lab Worx, the go-to-market consultancy arm of Sharktech Global, at divinelabworx.com.