UDesign Process User's Guide

The UDesign Process, adapted from Clausing[1] and Pugh[2], is made up of five phases: Definition, Requirements, Benchmarking, Concepts, and the Go/No-Go Review. Figure 1 depicts the fundamental components, along with some of the characteristics of each phase.

This User's Guide is intended to help the team master the "mechanics" of a project. "Session" numbers listed below do not, in most cases, refer to the exact number of times a team will convene on a given project. Rather, they represent logical breaks in the flow of a typical project.


Figure 2 - Customer Wants

This first spreadsheet (Figure 2) is used to state the team's mission, and make a list of customers and wants. Customers are also known as clients, stakeholders, or constituents. They care about the task you are performing. Develop a sharp, one-sentence statement of the mission of this UDesign project or subproject. In making the customer list, use first and last names of real people whenever possible. Make a complete list of all possible external and internal customers. Consider converting internal customers into partners. In other words, get a buy-in from them to join the team as a member or consultant. Members are responsible for decisions, while consultants only provide information. Partners may provide serious constraints that will limit the scope of the possible solutions.

Once the first-pass list of customers is done, continue the session to brainstorm the wants of each customer. Role-play each customer. Enter the results as you go alongside each customer's name. Duplications are fine. Just be as complete as possible. Conflicting wants are common, and are particularly useful to sharpen the decisions to be made. Continue to add wants, and even more customers until the team is satisfied. Consensus is sought for each team decision. Understanding the conflicting opinions is necessary. Make the decisions and keep moving forward. Speed is especially useful early on. Iterations are expected and welcomed, but not too many sub-iterations.

This information is critical to the success of a design project. To be effective, it must be organized in priority order. A recommended method for ordering the wants is shown on the right hand side of Figure 2.

List the wants. Give each want a score based on the importance of the customer to the project (rank from Column 3), and the importance of that want to that customer (use values from Row 9). The formula for the score is shown in the example calculations in the spreadsheet.


Copy and paste the wants and scores into columns 12 and 13 (use "paste special" to only transfer values). Use the "sort" function in Excel to put the wants in order of their score. Add the column into "SUM". Use the last column to calculate the percent importance. Then transfer the results to the next sheet of the spreadsheet, as shown in Figure 3. Now the team and its management can see the relative importance of the wants. This also begins the process of linking the priorities back to individual customer representatives. When the situation changes based on new information, the customer priorities must be re-confirmed and re-linked.


Figure 3 Top Ten Wants

This should mark the end of the first session. Save the spreadsheet with a name that includes "wants" and the current date. Make printouts for the team. Action items will probably include some customer contact (market research) to verify the team opinions expressed in the first spreadsheet.


Next is the beginning of the Competitive Benchmarking phase of the process. A listing of "competitors" is keyed in. It is especially good to include competitive entries that represent "best practices", the best-known performance in each of the separate customer wants. Continue the listing as long as necessary. Then rate the competitors against the customer wants. This may take outside homework and multiple sessions. Don't get stopped here. It is important for the team to exercise judgment so they can keep moving forward rapidly. Find out how the competitors measure themselves and how they advertise. Homework to verify the competitive benchmarking will almost always be welcomed, and will be a continuing process.

The next task is to brainstorm ways to measure the customer wants. Taking each customer want from left to right, consider all sorts of ways to measure that want. Key them into the spreadsheet in arbitrary order, making a column of metrics. The metrics should be thought of as the engineering quantities that represent the "yardstick" of measurement for that want. Try to think of at least 3 or 4 ways of measuring each want. Some metrics will be useful in measuring multiple wants. List each such metric only once. Continue until consensus is reached or you run out of ideas. Keep in mind that the metrics will be collected into a set of engineering specifications.

The next task is to judge how strongly correlated each metric is to the customer wants. Taking each customer want in turn, insert a number in the cells in each spreadsheet column according to the legend shown at the bottom of the spreadsheet. Each number corresponds to a multiplying factor as shown. In effect, the customer wants will be transformed into a set of measures that directly correlate with them. When the process is completed, the software can calculate the points associated with each metric. The one with the most points will be the most important metric. Copy the portion of the screen including all the metrics (all the way down). Copy (& paste special) the column labeled "%" to sort by. Sort this 2-column matrix. The result will be a re-ordering of the quality metrics in order of importance from top to bottom. The percentages are shown. Save the spreadsheet under a different name. Now go through the metrics one-by-one to verify the interactions by row. See if they make sense. The process works best when the number of metrics is small, so try to cut the list down. Especially look at the bottom of the list to see if some higher ones cover some of the lowest ones. If possible, eliminate metrics until the number of total metrics is about 10-15. But don't eliminate any metric that is strongly correlated to a want.

Now, a target value or range should be developed for each metric. Based on your benchmarking, each competitor can be compared with respect to the metrics. When consensus is reached, go out and celebrate! You have reached a major milestone in the process. You have developed an excellent team understanding of your customers, their wants, the competition, and how to measure the customer wants. Save the spreadsheet!


Figure 4 Concept Selection

The next session begins with the final spreadsheet. As before, the mission statement is included for constant reference. The benchmark refers to the incumbent or current concept. Metrics, percent importance, and target values are copied over from the previous spreadsheet, in order of importance.

At the top of the page is a cell to describe each concept. Each "concept" should be thought of as a complete answer to the customer wants, but the concepts list often becomes a list of features to be considered for inclusion. This part of the process is probably the most "creative," so constraints here are unwelcome. Let the ideas flow freely without criticism, in a classical brainstorming session.

When you re-convene, you will be entering one of the most powerful elements of the process. The Pugh Concept Selection Process ensures that only robust concepts are selected. It saves the team from taking risks on weak concepts. When completed, this part enables the entire team to uniformly defend their decisions to each of their constituencies. It is the final step before the Go/No-Go Review.

Start in the upper left corner of the spreadsheet with concept 1 and quality metric 1. Ask yourselves if concept 1 is better, worse, or the same as the benchmark in terms of quality metric number 1. Insert a b (better), w (worse), or s (same), respectively. If you don't know, put a u (unknown). Continue to evaluate concept 1 against the benchmark by continuing down the column marked concept 1. When you finish concept 1, do the same thing with concept 2, 3 and so forth until you fill all cells of the spreadsheet. Sum up the number of entries of each type at the bottom of the spreadsheet.

Now comes some serious decision-making. Seek to identify weak concepts at this point. If a concept has so many minuses that it would be weak even if all the knowledge gaps (U) were resolved, set it aside. This shows the power of the process to save time by not spending resources on weak concepts. Many knowledge gaps will probably remain. Action items should be assigned to fill the most serious knowledge gaps. Take a look at the most attractive concept. During the next iteration, make that concept the benchmark. Proceed carefully to re-order concepts and re-compare all remaining ones to the most likely winners. Add detail to the remaining concepts to be able to better differentiate them. Continue iterating and doing homework as necessary. Try hard to end up with one best concept to take to the Go/No-Go Review. (Save the Spreadsheets!)

Page 1/7

[1] Clausing, D. P., Total Quality Development, ASME Press, 1994.

[2] Pugh, S., Total Design, Addison-Wesley, 1990.