Lean Six Sigma
Lean Six Sigma
What Is Lean Six Sigma?
- Six Sigma = less variation, better quality; Lean = less waste, more speed
- Data-driven improvement in five DMAIC steps
- Sustaining the gains (Control) is harder than improving
What Lean Six Sigma really changes is how people talk about improvement. 'We have had a lot of defects lately' becomes 'a 3.2% dimensional tolerance deviation rate, KRW 6.2 million (≈ US$4,640) in monthly losses,' and the search for causes moves from 'Isn't it operator skill?' to 'The difference in means between Machine 2 and Machine 3 is significant.' Because you find the input variables (X) that drive the result (Y) and lock them in as control items, the improvement stays in place even when the owner changes. The real value of this methodology is that it breaks the pattern in which the guesswork of the loudest voice in the meeting wins.
The test for choosing a tool is whether you know the cause. If flow is blocked, inventory and waiting times are long, and the cause is plain to see, Lean tools (rearranging flow, shortening changeovers) can finish the job in a few weeks. When the cause is hidden and variation is the problem, bring in DMAIC. If you are designing a new product or process, use DMADV (Design for Six Sigma) instead of DMAIC. For reference, 3.4 DPMO is the Six Sigma level, while many Korean manufacturing processes actually run at 3 to 4 sigma (4 sigma is about 6,210 DPMO). Aiming for Six Sigma from day one leaves projects stuck.
On the operations side, Green Belt projects typically run 4–6 months and Black Belt projects 6 months or more. The numbers hold up only in organizations where Finance agrees on the savings calculation when projects are selected, and checks 3–6 months after closing that the savings actually reached the P&L. If project leaders calculate and report their own savings, the amounts inflate and trust collapses.
The most common misstep is setting the number of belt holders or projects as a KPI. Teams then dress up problems whose answers they already know as projects and squeeze them into the DMAIC template, and the shop floor sees it as 'one more report to write.' The second is skipping Measure. When you analyze data collected without a measurement system analysis (Gage R&R), a problem blamed on the equipment often turns out to be caused by the gauge and measurement method. The third is neglecting Control: if you add control charts but set no rules for what to do when a process drifts, the defect rate returns to its original level within six months.
John KimInstructor John Kim, author of B2B Marketing Design and CEO of Nanum Consulting
Ara LimSpecialist in global leadership, AI, and business communicationCreated in 1986 by Motorola engineer Bill Smith and rolled out company-wide at GE by Jack Welch in 1995, it spread as a leading management innovation method. It later combined with Lean, which comes from the Toyota Production System, to become Lean Six Sigma.
Consider a hypothetical case: the injection molding line at Company A, a second-tier automotive parts supplier (240 employees, KRW 78 billion (≈ US$58,357,000) in annual sales). The dimensional tolerance deviation rate on one connector item had climbed to 3.2%, bringing in two customer complaints per quarter. Production blamed the mold and quality blamed the material lot, and the two sides had been deadlocked for six months. Manager Kim of the quality team led a four-month Green Belt project.
- At the kickoff meeting, Manager Kim deleted the phrase 'reduce defects' and set a firm target: 'deviation rate from 3.2% to 1.0%, annual savings of KRW 94 million (≈ US$70,300).' She first got Deputy Manager Lee in Finance to sign off on how the savings would be calculated.
- Questioning the measurement system first, they ran a Gage R&R and found operator-to-operator variation accounted for 28% of total variation. Only after replacing the gauge and standardizing measurement positions did they recollect two weeks of process data.
- Splitting the tests by machine, shift and material lot showed that material lot differences were not significant, and the cause narrowed to mold temperature variation on Machine No. 3. Foreman Park on the production team said, 'We knew the heater band was worn out long ago, but we had no data.'
- They ran the three factors of mold temperature, holding pressure and cooling time through a design of experiments to fix the settings, and wrote those settings straight into the standard work instructions and the equipment inspection checklist. Heater bands are now replaced every six months.
- After putting in a control chart and a weekly check cycle and monitoring for three months, the team cut the deviation rate to 0.9%, had zero customer complaints, and booked KRW 68 million (≈ US$50,900) in savings that the finance team recognized as real.