Part one of this article, which includes a "Scoring Guidelines" table, can be found on page 27 of the April issue of Quality Digest. It is also on the Web at www.qualitydigest.com/april02/html/sixsig.html

Due to the size and complexity of the charts, they are not included here. To read the entire article complete with charts, click here to download the PDF file of the article .

Designing product and process

The second stage in the product development process is designing the product and process. The evaluation here is based on a total of 500 points and is divided into several subsections. Each subsection carries its own requirements and its own weight of points.

 The first subsection is selecting product or process concept. The requirements are to create or establish alternative product design and manufacturing process concepts, and to derive best alternatives for development. There are six questions, worth 10 points each, that will facilitate the decision and the process. (See Table 5 by clicking here to download the PDF file of the article ) The questions are designed to promote an open discussion about "newness" without fear of intimidation or retaliation. The most critical characteristic of the process isn't the numerical scheme but the ability to differentiate the product or process differences in a manner that's appropriate to the customer, the organization at large and the regulatory bodies. A minimum score of 35 is expected, but a good score is anything higher than 45.

 When selecting the new concept for product, process or both, the engineer must also consider concurrent product and process designs. This is imperative in our modern world, and this stage of product development should address it. The requirements are very simple but very hard to implement. Specifically, we're interested in design and model products and processes concurrently using low-cost tolerances and inexpensive materials. (We can do that with parameter and tolerance design, as part of the development phase, with the sole purpose of creating robust designs. That's why, in design for Six Sigma, we must focus on Y = f[x, n] rather than the traditional Y = f[x].)

 In this subsection, there are 14 basic questions, worth 10 points each, which will facilitate the decision and the process. The questions are designed to promote an understanding of concurrent engineering and the application ramification in the design process. (See Table 6 by clicking here to download the PDF file of the article .) This is the stage in which much engineering discussion is geared toward alternative analysis and optimizing testing possibilities. The important characteristic of this particular review process isn't the numerical scheme but the ability to express the differences in a manner that's appropriate to the customer, organization at large and the regulatory bodies. The basis for this analysis is focused, as appropriate, on trade-off and many other tools and methodologies. A minimum score of 85 is expected, and questions five through nine should have a minimum value of eight points each. A good score is anything higher than 115.

 The third subsection in evaluating the design product and process is the approach (i.e., methodology or process) that allows the engineer to identify and prevent failure. The requirement here is to improve product and process through reduction of potential failure modes and functional variability. (See Table 7 by clicking here to download the PDF file of the article.) Usually in this category, there are four core questions, worth 10 points each, which guide the evaluation process. The numerical scheme isn't as important as recognizing and discussing potential failures and eliminating them from the design. The questions should facilitate the process and will focus the discussion to priority items. A minimum score of 25 is expected, and question two should have a value of 10 points. A good score is anything higher than 30.

 The fourth component in evaluating the design product and process is the optimization function in the presence of noise. In design for Six Sigma, this is the most important characteristic. The DMAIC model focuses on fixing problems, whereas design for Six Sigma focuses on prevention and robustness. Robustness is indeed the design's focal point if we're really serious about improvement.

 The traditional model of Y = f(x) is no longer appropriate. We must focus on the Y = f(x, n), which means that the customer functionality (Y) must be satisfied with engineering requirements (x) but in the presence of noise (n). Therefore, the requirement at this stage is to optimize product and manufacturing/assembly process functions by testing in the presence of anticipated sources of variation (i.e., noise). There are six questions, worth 10 points each, and they should serve as the springboard of ideas for sound evaluation. (See Table 8 by clicking here to download the PDF file of the article.) A minimum score of 35 is expected, and questions four and five should have minimum values of nine points each. A good score is anything higher than 45.

 The fifth component of designing product and process is the issue of tolerance design—perhaps one of the most misunderstood concepts in any design endeavor. Tolerance design isn't the same as tolerancing; major differences exist between the two. Tolerance design forces the engineer to think in terms of modern systems—i.e., a holistic, top-to-bottom approach.

 The requirement for tolerance design is to adjust product/process tolerances and materials to achieve a desired performance, with cost-benefit trade-offs factored in. Key characteristics for manufacturing control and continued variability reduction must also be identified. There are four questions, worth 10 points each, which deal with this subsection. (See Table 9 by clicking here to download the PDF file of the article.)

 The sixth subsection of designing the product or process deals with finalizing process/control plans. The requirement here is to concur with process tooling, gages and control plans. There are nine questions that should guide the evaluation process, worth 10 points each. (See Table 10 by clicking here to download the PDF file of the article.) A minimum score of 60 is expected. Question three must have a value of 10 points, and questions five through nine should each have a minimum value of nine points. A good score is anything higher than 75.

 The seventh subsection of designing the product or process is design verification. The requirement for this substage is to integrate and verify design and manufacturing process functions with production-like hardware and/or software. There are seven questions, worth 10 points each, which may facilitate the understanding and decision making. (See Table 11 by clicking here to download the PDF file of the article.) A minimum score of 40 is expected, and a good score is anything higher than 55.

 The third stage in the product development process is to verify product and process. The evaluation here is based on a total of 100 points and is divided into two subsections. Each carries its own requirements and weight of points.

 The first subsection deals with design/manufacturing confirmation. The requirement here is to confirm manufacturing and assembly process capability to achieve design intent. Remember that the intent is always driven by the customer's functionality. Therefore, if the intent is not met, functionality is not met; moreover, the customer isn't satisfied. There are six questions, worth 10 points each, which focus on this intent. (See Table 12 by clicking here to download the PDF file of the article.) A minimum score of 35 is expected, and question three must have a value of 10 points. A good score is anything higher than 45.

 The second subsection of verifying product and process deals with launch and mass production confirmation. Obviously, if your organization doesn't deal with this, it's not appropriate for evaluation purposes. However, if this subsection is

relevant to your organization, remember that the requirement here is to launch the product, then ramp up and confirm that mass production delivers function, cost, quality and performance objectives. To facilitate this, there are four questions worth 10 points each. (See Table 13 by clicking here to download the PDF file of the article.) A minimum score of 25 is expected. A good score is anything higher than 30.

Managing the program

 The fourth and final stage in the product development process is to manage the program. The evaluation here is based on a total of 150 points and is divided into three subsections, each carrying its own requirements and weight of points.

 The first subsection requirement is to establish and maintain a highly effective team, for both product and process, that has a shared vision. Without this shared vision, everyone will pull his or her own way, and failure will result. There are nine questions, worth 10 points each, that focus on the team effort. (See Table 14 by clicking here to download the PDF file of the article.) A minimum score of 70 is expected. A good score is anything higher than 80.

 The second subsection of the fourth stage deals with establishing a program information center. The requirement is to maintain and use this program information center to understand global programs of applicable, social and institutional knowledge. How sad that even major corporations continue to repeat the same steps to a repetitive problem because no one takes the time to document the information appropriately. In this subsection, we focus on four questions worth 10 points each. (See Table 15 by clicking here to download the PDF file of the article.) A minimum score of 30 is expected, and questions three and four must have minimum values of 10 points each. A good score is anything higher than 35.

 The third subsection of managing the program deals with updating corporate memory. We all talk about "things learned," but unfortunately very few of us, if anyone, systematically document these learned things so that they can be used again directly or as a surrogate data for other problems.

 The requirement here is to update the corporate knowledge database with technical, institutional and social lessons learned. To do that, the focus is on two basic questions, worth 10 points each. (See Table 16 by clicking here to download the PDF file of the article.) A minimum score of 15 is expected, and question two has a minimum value of nine points. A good score is anything higher than 15.

Design review timing

 As mentioned earlier, the actual timing is based on organizational and product milestones that are realistic and attainable within the constraints of the organization's internal and external forces.

 It must be emphasized, however, that in any evaluation the three components of approach, deployment and results are kept separately, and vigilance is necessary to keep them under control in each product development cycle. They're all important.

Summary sheet

 For the convenience of the practicing engineer, the summary sheet on page 51 can be used to log the design review process as well as the results.

About the author

 D. H. Stamatis, Ph.D., CQE, CMfgE, is president of Contemporary Consultants. He specializes in quality science and is the author of 20 books and many articles about quality. This article is based on his recent work in Six Sigma and Beyond, a seven-volume resource on the subject. More about design for Six Sigma may be found in volume six of that work, which is published by CRC Press. E-mail Stamatis at dstamatis@qualitydigest.com . Letters to the editor about this article can be sent to letters@qualitydigest.com .