Q: My employer recently introduced a strategy of outsourcing product manufacturing processes to China and service processes to India. My team is currently working on several processes related to the outsourced products and services. Should we continue working on improving the processes or abandon the efforts entirely and move on to something else?
A: Undoubtedly, the answer depends on several considerations:
- The level of control your company intends to maintain over the companies in China and India.
- Contractual obligations that have been established.
- The current performance of the processes being outsourced.
Consideration one (the “in-your face” approach): Some organizations look for more than just a deliverable from their suppliers. For example, they might want a say in what process is used, who works on the process and how changes to the process are effected. In this situation, you might want to continue working on the processes in an attempt to simplify, streamline and mistake-proof them.
Transferring responsibility for executing a less than pristine process halfway around the world might create insurmountable complications down the road. In addition, process complexity, the availability of training and training documentation, and the overall education level of the workforce must factor into your decision regarding whether you should continue.
Consideration two (the “fire-and-forget” approach): Some organizations will stipulate contractually the requirements or technical specifications related to the deliverables with no regard for the process required to produce those deliverables. All the organization cares about is that the right product is delivered on time with the specified quality characteristics.
Essentially, the organization has intentionally abdicated authority and responsibility for the process that produces the deliverables it requires. This might sound harsh, but it might be the best business decision. However, when making such a decision, the organization must understand that potentially serious consequences can occur due to a lack of oversight or minimal involvement.
Consideration three (the “hear no evil, see no evil” approach): Some organizations have jumped on the outsourcing bandwagon and established strategic outsourcing initiatives. Consequently, they are outsourcing many of their internal processes to Asia for purely financial reasons. Process performance or capability is not likely a consideration. As a result of such parochial vision, some companies have had to insource.
Insourcing carries its own set of problems, such as re-establishing the knowledge base (for example, job cuts), repurchasing capital equipment (equipment was sold off or shipped overseas) and redesigning the outsourced process. So what initially might have appeared to produce short-term financial gains turns out to be an operational disaster in the long run.
I hope these three considerations help to make a decision and also aid you in influencing key decision makers in your organization regarding outsourcing initiatives.
T. M. Kubiak
TK Performance Improvement Solutions
Q: Is Six Sigma an appropriate solution for every problem? If not, then how do we choose a project that would benefit from Six Sigma?
Ajay Sahore, firstname.lastname@example.org
A: The short answer to the first questions is, “No.” It is not always necessary to assemble a large team to solve a problem because some problems have straightforward solutions.
Consider this scenario: A circuit board manufacturer has excessive rework at the end of the assembly line. A checksheet is created to tally the failure modes. Some of the failures are a result of incorrect wiring. This failure mode can be eliminated by using color coded lead wires and connectors. The operator simply attaches the wire to the connector of the same color. Many assembly errors can be eliminated with simple fool-proofing techniques or poka-yoke fixtures.
Some problems require more sophisticated tools to reach a solution. I once participated in a team that was trying to fix a leaky irrigation valve. The problem had plagued the production line for years. Approximately 0.5% of the valves leaked. This forced the production line to add 100% inspection using a high-pressure water test. We used a designed experiment to identify the factors contributing to porosity (the root cause of the leaks). Minor adjustments to the molding process completely eliminated the problem, and after a few weeks we discontinued the 100% inspection.
Because water testing was the slowest operation on the line, eliminating 100% inspection increased the capacity of the line by more than 25%, and scrap costs became virtually negligible. To solve this problem, we applied design of experiments (DOE), a tool in the Six Sigma toolbox, but we did not use a formal Six Sigma approach.
How do you choose a project for Six Sigma implementation? Begin by identifying potential projects. Assess each opportunity in terms of the potential payback in financial terms. Consider whether the project is aligned with the strategic direction of the company. Even if the payback is low, management might choose to support a project if it strengthens the company’s competitive position by improving customer satisfaction.
Consider the complexity of the problem. If it can be solved quickly using simple tools and an informal approach, then you don’t need a Six Sigma project.
Be sure to carefully define the scope of the project. If its scope is too broad, the likelihood of success will be greatly diminished. Many practitioners suggest creating a project selection matrix, on which various proposals are “scored” relative to pre-established evaluation criteria. The proposals with the highest overall scores get management support, including allocation of resources and a budget for capital expenditures.
For more information:
Implementing Six Sigma, second edition. Forrest Breyfogle. Available from ASQ Quality Press or through the website www.smartersolutions.com.
Master Black Belt and principal consultant
Q: It appears that many teachers (certainly compared to professionals in other disciplines) don’t seem to recognize the usefulness of Total Quality Management (TQM) tools in public education as avenues for increasing student achievement. What management strategies have been successful in overcoming this barrier?
A: Educational author Richard Elmore describes the predominant climate in which public school teachers view themselves as independent contractors. In other words, teachers are not open to substantive collaboration or change.
One frequently used strategy for implementing TQM practices is board of education or superintendent dictates. But such approaches often leave gaps as a result of resistance and low morale.
Conversely, participation based methods have shown success in certain contexts. Such strategies use a research based approach. These begin with district or schoolwide conversations or other inputs about what classroom practices comprise quality education and how to control for variance.
One approach uses a series of interviews with key educational leaders, focus groups with teachers, and broad based districtwide surveys to identify and then rank essential quality practices or outputs that teachers are most likely to adopt.
This participation based approach is not as clean as a dictate, but it builds support along the way without the typical gaps of resistance.
coordinator of assessment
Sheboygan Area School District, WI
Q: In evaluating our statistical process control (SPC) for a customer, we examine product impurity by evaluating the upper limit of process capability. In our reports we see a different result for overall process capability, Cpk, and Cpu, the upper limit. Which one is appropriate and why do we see a difference?
Yu Chen, email@example.com
A: This is an interesting question because you are only interested in the Cpu, or the upper limit of process capability. The definition of Cpk is the minimum (Cpu, Cpl) in which
USL and LSL are the upper and lower specification limits.
It could be that whatever software is used to calculate Cpk, Cpu, and Cpl is automatically calculating Cpl and using that as Cpk. In this case, with a one-sided limit, the Cpu is the real indicator of process capability for this type of data.
For more information:
I. Elaine Allen, Ph.D.
Arthur M. Blank
Center for Entrepreneurship
associate professor of Statistics & Entrepreneurship