To remain competitive and improve sales income, companies must continually evolve new product features and new processes to produce such features. One of the keys to achieving continuous improvement is getting customers involved as early as possible in the product or service development process. Quality Function Deployment (QFD) is a technique that can be used to assure that a product or service is designed and realised to exceed customer expectations.
In QFD, customer attributes and technical characteristics, representing the voice of customer (VOC) and the voice of engineer (VOE), respectively, are linked via a relationship matrix. Using an appropriate relationship measurement system, the relationships can be accurately measured, and the voice of the customer can be numerically mapped onto the voice of the engineer successfully. Thus, a suitable relationship system will lead to accurate interpretation of the information provided in the HOQ.
To better understand relationships in the relationship matrix, a 5-level relationship system is proposed. Two issues regarding any relationship system are important to note, namely, the relationship strength and the numerical value. There are altogether five levels (excluding non-relationships) to describe the possible relationship strength between each customer attribute and each technical characteristic (see Table 1).
Theoretically, a relationship matrix constructed using the traditional relationship system can be easily converted into another matrix with five relationship levels. This is provided that the relationships have been reasonably well developed. "Weak" relationships measured by the traditional system could be replaced by "Weak" or "Some"; "Moderate" relationships could be replaced by "Some", "Moderate", or "Strong"; and "Strong" relationship could be replaced by "Strong" or "Extreme" in the refined relationship system.
However, the 5-level relationship system provides for better discrimination. More levels of relationships allow for greater freedom and possibility in differentiating relationship strengths. This should lead to more accurate portrayal of the customer attribute/technical characteristic relationships. In addition, the proposed 5-level relationship system adopts more reasonable numerical value settings for the relationship strengths. Although the range is still the same (from "1" to "9"), the difference between two adjacent values is not as large as in the traditional system. Thus, the problem of less accurate discrimination that could occur in the traditional system is partly avoided.
To illustrate the proposed 5-level relationship system, an example is presented here which is adapted from a case study on education. In this case study, QFD was used to successfully help improve advising and teaching in engineering education.
The following customer requirements were obtained by asking what are requirements needed in order to have a class that is valuable for the students:
The following design requirements were obtained by asking how the customer requirements can be accomplished:
Based on the customer requirements and design requirements, a house of quality was established. After performing the relationships conversion using the proposed 5-level relationship system, the relationship matrix is shown in Figure 1. Research is underway on techniques to apply the methodology to a wide spectrum of industrial applications.
Table 1. Proposed 5-level relationship system
Figure 1. Relationship matrix for a well taught class
using 5-level relationship system.
