DMAIC is the five-phase project cycle of Six Sigma: Define the problem and goal, Measure the current process with real data, Analyze root causes statistically, Improve by implementing and validating solutions, and Control by locking in the gains with monitoring. A typical DMAIC project runs three to six months.

Manufacturing Glossary

Lean Six Sigma in Manufacturing

Q: What is Lean Six Sigma?

Lean Six Sigma is an improvement methodology that combines two disciplines: Lean (eliminating waste and improving flow, rooted in the Toyota Production System) and Six Sigma (reducing variation using statistical analysis, developed at Motorola and popularized by GE). Lean makes the process fast; Six Sigma makes it consistent.

Q: Apa itu Lean Six Sigma dalam konteks pabrik di Indonesia?

Di pabrik Indonesia — terutama supplier otomotif — Lean Six Sigma biasanya berjalan berdampingan dengan aktivitas kaizen dan requirement improvement dari OEM. Proyek DMAIC dipakai untuk masalah kronis yang tidak selesai dengan kaizen biasa: reject rate yang naik-turun, variasi cycle time antar shift, atau scrap yang penyebabnya tidak jelas. Kuncinya adalah data — fase Measure membutuhkan data produksi dan kualitas yang akurat.

Q: How does digital production data change Lean Six Sigma?

The Measure phase traditionally consumes 30-50% of a DMAIC project — teams manually collect samples for weeks. With MES and digital checksheets already capturing every cycle, downtime event, and inspection result, the baseline exists on day one. The Control phase also transforms: instead of relying on manual audits to sustain gains, live dashboards and automatic alerts detect drift the day it starts.

By Mahasys Multi Zenith · Manufacturing software for Indonesian plants · Last updated June 2026

Lean Six Sigma combines two improvement disciplines: Lean — eliminating waste and improving flow, rooted in the Toyota Production System — and Six Sigma — reducing variation through statistical analysis, developed at Motorola in the 1980s. Lean makes a process fast; Six Sigma makes it consistent. Together they attack the two ways a factory loses money: doing unnecessary things, and doing necessary things unpredictably.

Two parents, one methodology

Aspect	Lean	Six Sigma
Origin	Toyota Production System (1950s)	Motorola (1986), scaled by GE (1995)
Enemy	Waste (muda) — anything the customer wouldn't pay for	Variation — inconsistency in output quality
Primary question	"Why does this step exist?"	"Why isn't this step repeatable?"
Core tools	Value stream mapping, kanban, 5S, SMED, standard work	SPC, capability analysis, DOE, hypothesis testing
Typical cadence	Continuous, daily kaizen	Project-based (3–6 month DMAIC)

The combination matters because each covers the other's blind spot. A pure-Lean plant can flow smoothly while producing inconsistent quality; a pure-Six-Sigma plant can make statistically perfect parts through a wasteful, slow process. Manufacturing excellence needs both — which is why most automotive supplier improvement programs, and most OEM supplier-development requirements, draw on both toolkits.

DMAIC: the project engine

Define. One problem, one metric, one goal. "Reduce paint defects on line 3 from 4.2% to under 1.5% by Q4." Vague charters kill projects before they start.
Measure. Establish the real baseline with data — not opinions, not last year's report. This phase traditionally eats 30–50% of project time because the data has to be collected manually.
Analyze. Find root causes statistically: which variables actually correlate with the defect? Pareto analysis, fishbone diagrams, hypothesis tests, regression.
Improve. Implement the fix, validate with data that it works, and confirm no side effects.
Control. Lock in the gain: updated standard work, control charts, alerts when the metric drifts. This is where most projects quietly fail — the improvement decays once attention moves on.

The belt system, briefly

Yellow Belt — basic awareness; participates in projects.
Green Belt — leads small projects part-time alongside their day job; the workhorse level in most plants.
Black Belt — leads complex cross-functional projects full-time; strong statistics.
Master Black Belt — trains belts, selects projects, owns the program.

A caution from experience: belts are a means, not an end. Plants that measure success by "number of people certified" get certificates. Plants that measure by "chronic problems permanently closed" get results.

Where digital data changes the game

Two DMAIC phases are transformed when a plant already has digital production data:

Measure — from weeks to hours. With an MES logging every cycle and downtime event, and digital checksheets logging every inspection, the baseline exists before the project starts. Teams skip straight to Analyze with months of clean historical data.
Control — from audits to alerts. The traditional Control plan relies on periodic manual audits, and gains decay between them. With live dashboards, the improved metric is watched continuously and drift triggers an alert the day it starts — not at next quarter's review.

The reverse is also true and worth saying plainly: Lean Six Sigma on paper-based data is slow and fragile. If your defect records are handwritten and your downtime reasons are guesses, the Measure phase becomes archaeology and the Analyze phase becomes statistics on noise. Digitalize the data capture first; the improvement methodology gets sharper for free.

How it fits with kaizen and jishuken

In an automotive supplier context, three improvement modes coexist:

Daily kaizen — small, continuous improvements by operators and supervisors. The cultural baseline.
Jishuken — intensive workshops, often with OEM or supplier-association facilitation, that develop people while improving a focused area.
Lean Six Sigma projects — for chronic, data-heavy problems that survive kaizen: fluctuating reject rates, unexplained scrap, variation between shifts.

Use the smallest tool that solves the problem. If a two-day kaizen fixes it, don't charter a six-month DMAIC.

Frequently asked questions

What is Lean Six Sigma? +

An improvement methodology combining Lean (eliminating waste and improving flow, from the Toyota Production System) and Six Sigma (reducing variation using statistics, from Motorola/GE). Lean makes the process fast; Six Sigma makes it consistent.

What is DMAIC? +

The five-phase Six Sigma project cycle: Define, Measure, Analyze, Improve, Control. A typical DMAIC project runs three to six months and targets a chronic, measurable problem.

Apa itu Lean Six Sigma dalam konteks pabrik di Indonesia? +

Biasanya berjalan berdampingan dengan kaizen dan requirement improvement dari OEM. Proyek DMAIC dipakai untuk masalah kronis yang tidak selesai dengan kaizen biasa — reject rate naik-turun, variasi antar shift, scrap yang penyebabnya tidak jelas. Kuncinya data: fase Measure membutuhkan data produksi dan kualitas yang akurat.

How does digital production data change Lean Six Sigma? +

Measure shrinks from weeks to hours because MES and digital checksheets already hold the baseline. Control shifts from periodic manual audits to live dashboards with drift alerts — gains stop decaying between reviews.