Six Sigma is a methodology of process improvement and set of statistical tools to improve a process or product and to prove and sustain the gains achieved. We use it on nearly every assignment as an element of sustaining improvements and for controlling risk. Occasionally we use Six Sigma to prove one production method superior to others or to prove that steps in a calculation like pricing can be eliminated without changing the validity of the result.
In general, Lean shortens lead time by reducing the time between operations in a process, and Six Sigma improves the operation itself. We use Six Sigma within Lean to reduce the variability that affects lead times at high capacity utilization.1
Six Sigma includes a rich collection of tools that can be used to:
- determine if a process is under control (e.g. control chart),
- determine the better or two ways or the best of several ways (e.g. T-Test or ANOVA),
- reduce risk (e.g. FMEA),
- improve the quality or feature set of a product (e.g. QFD),
- extrapolate a result (e.g. regression analysis),
- predict the probability of a result based on many factors that themselves vary (Monte Carlo analysis)
- determine the totality of effects of design elements (e.g. DOE).
Each of these tools can be applied in creative ways to increase productivity. We have applied Monte Carlo Analysis to determine pricing that has a high probability of acceptable profit margin, and have used regression analysis to replace or verify the difficult and unreliable manual process of costing a large repair. Only by having a full understanding of the science behind the tools can we use these tools in productive new ways.
Note 1: The disastrous effect of variability on lead times at high capacity factors is given by Kingman's equation, and is the reason for "stop and go" traffic at rush hour, when the freeway is at capacity. It is also why new processes or plants should not be designed around 100% capacity utilization.