QM BenchQuality engineering tools

Problem solving and improvement

Quality tool selector: which method fits your problem?

Choose the area and the problem. You get the methods to start with, the supporting methods with the reason for each, the first three steps and the KPI that shows success. Most methods open directly in a free QM Bench tool.

Production

High scrap and rework

Scrap or rework is above target, often spread over several defect types.

First steps

  1. Collect 4 weeks of scrap data by defect type, station and shift.
  2. Pareto by cost, not only by count.
  3. Pick the top defect, check the gauge, then analyse causes with data.

Measure success by

Scrap rate (%) and scrap cost per month.

Full guide: High scrap and rework

All problems by area

Production

Quality

Suppliers

Launch and engineering

Maintenance

Logistics

Customer

Management system

People and safety

Five rules for choosing a method

  1. Contain first, analyse second. Protect the customer before you look for the root cause.
  2. Trust the data only after the measurement check. Gage R&R or attribute agreement before charts and experiments.
  3. Match the effort to the problem. A 5 Why on the shop floor, an A3 for a team problem, DMAIC for chronic losses.
  4. Prove the root cause. Switch it on and off, or confirm it with a test, before you invest in actions.
  5. Lock the result. Update the PFMEA, control plan, work instruction and audit checklist, or the problem comes back.

Frequently asked questions

Which problem-solving tool should I use?

Start from the problem, not the tool. Customer complaints need an 8D, internal improvements fit an A3, chronic problems with many causes need data (Pareto, multi-vari, DOE), and repeated human errors need error-proofing. The selector gives the methods to start with for each typical problem.

What is the difference between 8D, A3 and DMAIC?

8D is the customer complaint format with containment and verified root cause. A3 is a one-page PDCA story for internal problems. DMAIC is a Six Sigma project for chronic problems where the cause is hidden in data and takes weeks to months. All three follow the PDCA logic.

Should I always use a fishbone diagram and 5 Why?

They are good for structuring ideas, but they only list and explain candidate causes. Verify the cause with data or by switching it on and off. For problems driven by several interacting factors, use multi-vari studies and designed experiments.

Why check the measurement system first?

If the gauge or the visual inspection cannot see the problem reliably, every chart, Cpk and experiment is built on noise. A Gage R&R or attribute agreement study takes a day and can save weeks of wrong analysis.