Custom poka-yoke fixture design: a practical guide
A poka-yoke fixture makes the wrong way physically impossible or immediately obvious: a part that only fits one way, a sensor that won't let the cycle start, a gauge that rejects a bad part. Here's how a custom one goes from defect to production.
The short answer
Start with the specific defect and where it happens. Choose the simplest mechanism that prevents it (shape, pin, stop, sensor or interlock). Prototype quickly, often with 3D-printed parts, test it with the operators on the real line, then build the production version in materials that stand up to the shift count. Good error-proofing costs little and pays back the first time it stops a bad batch.
Types of poka-yoke
- Contact (physical): locating pins, asymmetric nests, go/no-go features. The part only fits one way.
- Fixed-value: the fixture counts or confirms a set number of actions, such as all four fasteners present.
- Motion-step: sequence is enforced, so step 3 can't start until step 2 is confirmed.
- Warning vs control: a warning alerts the operator; a control stops the process. Controls are stronger; warnings are cheaper and easier to add.
The design process
- Define the defect. What goes wrong, how often, what it costs, and at which step. Photos and a sample good and bad part help.
- Watch the operation. The best fixtures fit how people actually work, including left-handed operators, gloves and the end of a long shift.
- Pick the mechanism. Physical prevention first; add sensors only when geometry can't do the job.
- Prototype. 3D-printed nests and pins let operators try it in days, not weeks.
- Validate. Try to defeat it. Run known-bad parts. Check cycle time didn't suffer.
- Build for production. Aluminium, steel, UHMW or durable printed materials depending on wear, with replaceable wear points.
- Document. Drawings, a spare-parts list and a one-page operator instruction.
Worked example
A bracket that can be installed rotated 180° causes a 1% escape rate on 2,000 parts a week: 20 bad assemblies, each costing $50 in rework or scrap, so $1,000 a week. An asymmetric locating pin that makes the wrong orientation physically impossible usually costs a small fraction of that to design and build.
What to have ready
- Part drawings or CAD, or physical sample parts (good and defective).
- Photos or a short video of the current operation.
- Cycle time target, shift count and any cleanroom, ESD or food-safety requirements.
- How the fixture mounts: bench, conveyor, existing tooling plate.
Common questions
What's the difference between a jig, a fixture and a poka-yoke?
A fixture holds a part in position; a jig also guides the tool. A poka-yoke is any feature that prevents or flags a mistake, and it's often built into a jig or fixture.
Can poka-yoke fixtures be 3D printed?
Prototypes almost always are, and many production fixtures are too, using durable materials. High-wear surfaces are often made from metal or replaceable inserts.
Do you need CAD files to start?
They help, but sample parts, measurements and photos are often enough to begin. Parts can be measured or scanned.