Standard methods

DMAIC: Six Sigma projects from Define to Control, with the evidence.

DMAIC is the data-⁠driven Six Sigma method for process variation whose cause is unknown: Define, Measure, Analyse, Improve, Control. With SolveR the project runs in one record from Define to Control, so the figure you show your customer has the data, the confirmed cause and a control plan with an owner behind it.

DMAIC in SolveR is set up by agreement: on the discovery call we show how it would run on your project and tell you when you can start using it.

When to use DMAIC, and when PDCA or 8D

DMAIC pays off when a measurable characteristic of a process varies or scrap will not come down, and the cause is unknown and has to be shown by data that still needs collecting. A project takes weeks to months, so you do not start one for a problem the team can solve in a few days.

  • PDCA when you have an idea and want to test it in one cycle.
  • 8D for a customer complaint where the customer expects the answer in one document.
  • Kaizen for a team improvement in one area, where the waste can be seen on the spot.

Torque variation on a screw joint, today and with SolveR

Typical case: the torque on a screw joint at one assembly station varies, and some joints fail at the final check.

TodayWith SolveR
The charter, the data, the fishbone and the control plan sit in separate files, and the pack is rebuilt for every phase review.The project runs in one record from Define to Control, with the target and the project boundary set at the start.
Process capability is calculated from readings nobody checked, and the measurement system was not assessed again when the gauge changed.The baseline has its source in the final check records, and next to it the record shows whether the measurement system was checked.
The fishbone lists many factors, one correlation is enough to call the cause "confirmed", and the action is already running.Recalibrating the tool changed nothing, so that cause is ruled out. A log of the air pressure shows that pressure on the shared line drops when the neighbouring station runs too, and that is the confirmed cause.
The control plan is updated after the fact, the reaction plan has no owner, and the project is closed and never checked again.The actions, a separate air supply and an electronic tool, go in only once the cause is confirmed. The record holds the control plan and reaction plan with an owner, and the updated PFMEA.

The case is an illustration written for this page.

What you get from a project

A project record that runs by phase from Define to Control: the baseline and target with their source, the cause with the measurement that confirmed it, actions with their trial, and a control plan and reaction plan with an owner. When the customer asks where a figure comes from, the path to it is in one record.

An illustration of how a DMAIC project record is built, not a screen of the product.

The control plan and the reaction plan are outputs of the method. The team and the process owner draw them up, and the record keeps them together with the project that justified them.

Why the FMEA so often stays empty: Poka-⁠Yoke for Knowledge: Why Your FMEA Stays Empty.

The DMAIC phases

  1. DefineProblem, target, project boundary, team
  2. MeasureA baseline with a source, a checked measurement system
  3. AnalyseA cause confirmed by data
  4. ImproveActions on the confirmed cause, a trial
  5. ControlControl plan and reaction plan, the process owner

One knowledge base for every method

A closed DMAIC project stays in the knowledge base with its confirmed cause, its actions and its control plan. When similar variation shows up at another station or line, the team starts from what the project already proved instead of collecting the same data again. Every method writes into the same base, and searching it is what the Brain page is about.

Data stays in the EU, never used to train AI models: Security & Architecture.

Why a closed case pays off only once it can be found: Factories forget what they already solved.

Questions about DMAIC

What is DMAIC?

DMAIC is the structured problem solving cycle of Six Sigma, run in five phases: Define, Measure, Analyse, Improve and Control. It is used on an existing process when the cause of a problem is not obvious and has to be found in the data.

When should you use DMAIC instead of PDCA or 8D?

Pick DMAIC when the cause is not obvious and you will not find it without analysing data. PDCA is quicker for everyday improvements, when you have an idea you want to test. 8D is the method for a customer complaint, where the customer expects the answer in one document.

How long does a DMAIC project take?

Usually weeks to months, depending on how complex the process is and how much data has to be collected. Projects with several influencing factors take longer. The typical duration of every method is in the selection table on the How it Works page.

Do you need a Six Sigma certification to run DMAIC?

Not necessarily, but it helps. The Analyse phase calls for statistical methods, so more complex projects are usually led by someone with Six Sigma training, such as a Green Belt or Black Belt.

Does SolveR replace our statistics software?

No. You calculate process capability, measurement system analysis and statistical tests in the software you already use. In a Six Sigma project SolveR is the project and knowledge layer: it holds the baseline, the confirmed cause, the actions and the control plan with their sources, and the closed project stays in the knowledge base.

Other methods

Every team picks the method that fits the problem, and one knowledge base stands behind all of them.

Last checked: 9 October 2026. Prepared by Galileon, the company that makes SolveR.

Bring a process that varies, and the data you already have on it, to the call.