4.5 Improving Quality of Design

There are a number of techniques that can improve quality of design. Here we will deal with three of them. Quality Function of Deployment (QFD):

The key purpose of quality function deployment (QFD) is to try to ensure that the eventual design of a product or service actually meets the needs of its customers. It is a technique that was developed in Japan at Mitsubishi's Kobe shipyard and used extensively by Toyota, and its suppliers. It is also known as the 'house of quality' (because of its shape) and the 'voice of the customer' (because of its purpose). The technique tries to capture what the customer needs and how it might be achieved.

Figure 4-4 shows an example of quality function deployment being used in the design of a new information system product.. The matrix contains various sections:
The Whats. Or 'customer requirements', is the list of competitive factors which customers find significant.

The competitive scores indicate the relative performance of the product, in this case on a 1 to 5 scale. Also indicated are the performance of two competitor products.

4.5 Improving Quality of Design

A QFD matrix for an information system product

4.5 Improving Quality of Design

The hows, or 'design characteristics' of the product, are the various 'dimensions' of the design which will operationalize customer requirements within the product or service.

The central matrix (sometimes called the 'relationship matrix') represents a view of the interrelationship between the whats and the hows. This is often based on value judgements made by the design team.

The bottom box of the matrix is a technical assessment of the product. This contains the absolute importance of each design characteristics. The triangular 'roof' of the 'house' captures any information the team has about the correlations (positive or negative) between the various design characteristics.

Although the details of QFD may vary between its difference variants, the principle id generally common, namely to identify the customer requirements for a product or service (together with their relative importance) and to relate them to the design characteristics which translate those requirements into practice. In fact, this principle can be continued by making the hows from one stage become the whats of the next (see Fig.4-5).

4.5 Improving Quality of Design

Figure 4-5
QFD matrices can be linked with the 'hows' of one matrix forming the 'whats' of the next

QFD matrices can be linked with the 'hows' of one matrix forming the 'whats' of the next

4.5 Improving Quality of Design

Value engineering

The purpose of value engineering is to try to reduce costs, and prevent any unnecessary costs before producing the product or service. Simply put, it tries to eliminate any coats that do not contribute to the value of the product or service. ('Value analysis' the name given to the same process when it concerned with cost reduction after the product or service has been introduced.) The team may attempt to reduce the number of components or use cheaper materials, or simply processes.

Value engineering requires innovative and critical thinking, but it is also carried out using a formal procedure. The procedure examines the purpose of the product or service, its basic functions and its secondary functions.
Team members would then propose ways to improve the secondary functions by combining, revising or eliminating them. All ideas would then be checked for feasibility, acceptability, vulnerability and their contribution to the value and purpose of the product or service.

Taguchi methods

The main purpose of Taguchi methods, as advocated by Genichi Taguchi, is to test the robustness of a design. The basis of the idea is that the product or service should still perform in extreme conditions.

4.5 Improving Quality of Design

The task is then to achieve a design which can cope with all these uncertainties. The major problem designers face is that the number of design factors which they could vary to try to cope with the uncertainties, when taken together, is very large. Performing all the investigations ( or experiments, as they are called in the Taguchi technique) to try to find a combination of design factors which gives an optimum design can be a lengthy process.

The Taguchi procedure is a statistical procedure for carrying out relatively few experiments while still being able to determine the best combination of design factors. Here 'best' means the lowest cost and the highest degree of uniformity.

4.5 Improving Quality of Design