SUMMARY
Six Sigma Roadmap Step 4A looks into the valuable improvement tools and techniques that support the DMAIC process. It involves prioritizing, analyzing, and implementing improvements. These activities in turn need to do the following:
In choosing the tools, you need to make your objectives clear and decide the technique that will most likely meet your needs. Keep the objectives simple and match the complexity of the tool with your situation. Adapt the method to your needs and if you find that a tool is not working, stop!
The DMAIC cycle is not purely a linear activity; rather it’s an iterative activity.
The Define phase clarifies the problem, goal, and process. It helps you look into the problem, the goal, the customer and the process that your organization is investigating. It starts with the so-called Six Sigma Project Charter. The Charter may comprise the Problem Statement, the Goal Statement, the Constraints and Assumptions, the Team Members and Responsibilities, Team Guidelines, and Preliminary Project Plan.
It is in this stage, too, that you will use the Voice of the Customer strategy. Your process improvement teams will have to balance the ideal of having a thorough understanding of the customers.
The final and essential step in DMAIC is developing a “picture” of the process. This tells you to put the problem in context, refine its scope and analysis, uncover obvious root causes, clarify inputs, roles, supplier/customer relationships, and target what to measure.
The Measure phase validates or refines the problem and begins the search for root causes. It measures two issues: the extent of the problem (baseline measure) and the key data to help narrow the problem to its root causes. The previous chapter discussed thoroughly the strategies to accurate measurement. Your organization must be able to establish a solid, repeatable measure confirming and often clarifying the problem or opportunity. Having done this will ready your organization to the Analyze phase.
The Analyze phase challenges the organization to become a process detective. You get down to the root causes of a problem or opportunity. You are asked to be on the lookout for the “usual suspects.”
In doing this, you represent Analyze as a cycle driven by generating and evaluating “hypotheses” or “educated guesses” on the cause of the problem.
The root cause hypothesis/analysis cycle looks like this.
You can enter the cycle either at point (a), that is, looking at the process and the data to identify possible causes; or point (b), where you start with a suspected cause and work to validate or refute it through analysis.
As shown in the diagram, you have two key sources of input to determine the true causes of the problem: (1) data analysis, or the use of already collected data or new data gathered; (2) process analysis, or a deeper investigation into and understanding of how work is done to identify problem areas.
There is a danger however in working with processes. People may work on the same business process differently, or using a different cycle. The causes of variations in a business process are referred to as the “5Ms and 1Ps.”
* Material – the consumables or raw inputs.
* Method – procedures, processes, work instructions
* Machine – equipment like computers
* Measures – techniques used for assessment
* Mother Nature – the environment where the work is done, or what affects any of the other variables
* People
Often the best way to learn from your data is to literally “see” them. For different types of data, you also have different types of diagrams.
The Pareto Chart or Pareto Analysis stratifies data into groups from largest to smallest. This bar chart helps you identify the most common occurrences or causes of a problem so that you can prioritize them. You can use Pareto only with discrete or category data. It is based on the “80/20 Rule:” 80% of the costs or pain in an organization is created by just 20% of the problems.
The Histogram or Frequency Plot shows the range and depth of variation in a group of data, or population. Technically, a Histogram translates continuous data, while a Frequency Plot displays discrete “count” data, such as numbers of defect. Both show data on an X and Y axis.
Run Chart or Time Series Plot shows the variation in a process, product, or other factor over time. The Run Chart (a.k.a. “Trend Chart” or “Line Graph”) and the Control Chart show things change, which can monitor ongoing activity or performance. A Run Chart structure shows a horizontal X axis as time or sequence of occurrence from left to right. The vertical Y axis represents any continuous or count measure.
The Scatter or Correlation Diagram shows the correlation between two factors that vary by count or on a continuum. It shows potential causal relationships between one factor and another, or between the suspected causes of a problem. You may see a positive, a negative, or a curvilinear correlation. When there is no correlation, the points will literally be scattered all around the chart like a cloud. When this happens, it means that a change in one factor has no effect in the other.
To be able to evaluate your Analysis phase and be able to proceed to the Improve phase, you must:
a. verify the cause through logical analysis
b. check the cause through observation
c. confirm your suspicions with people who know
d. apply the confidence test
The Improve phase takes care of the possible actions or ideas that will help you address the root cause of your problem, identify the workable potential solutions, those that will most likely bring your organization to achieve your goal with the least cost and disruption, and evaluate your chosen solution for its effectiveness so that you can implement it.
The ideas you come up with during the Improve phase are like raw materials. To be of value to the organization, they need to be refined. Usually, you may start with a Solution Statement, a clear expression of a proposed improvement. It becomes the project objective once you’ve chosen a solution to implement.
In this midpoint of the Improve phase, you are now ready to actually do something. To launch solutions successfully, you need to focus on the “four Ps:” Planning, Piloting, and Problem Prevention. And to close the Improve phase, you must capture data in such a way that they track the impact of the changes as they take effect.
COMMENTARY
Clearly, we can see that DMAIC is a methodology that should be applied to all processes. Its value lies in the fact that it helps organizations identify and manage linkages in a process-based quality management system.
The use of AutoRec story in all stages of the cycle is helpful in the sense that it gives continuity to the analysis and implementation of the Six Sigma projects. It is helpful in visualizing what ought to be done and not to be done. Many managers would rather hear practical examples than read theoretical explanations.
Although the DMAIC method is an amended version of the P-D-C-A originally conceptualized by Deming in the 1960s, it broke down the P-D-C-A model into a more detailed steps and processes. Today, what DMAIC is to Six Sigma, PDCA is to the ISO Quality Management System. In general sense, both are almost similar.
