Multi-layered HOQ approach for quality function deployment

• Main Author: Yeo, Ya Yun.
• Other Authors: Leong Kah Fai
• Format: Final Year Project
• Language: English
• Published: 2011
• Subjects: DRNTU::Engineering::Industrial engineering::Quality engineering
• Online Access: http://hdl.handle.net/10356/46069
• Institution: Nanyang Technological University

The House of Quality (HOQ), a fundamental tool used in the first phase of the Quality Function Deployment (QFD) process, reflects customer desires and tastes and translates them into technical specifications. Researchers realise the importance of a HOQ matrix that is able to support good decision making to prevent carrying over errors to the subsequent stages. Thus it is critical that the HOQ matrix has precise input information and an effective building process to ensure robust output results. One of the factors that contribute to an accurate HOQ matrix is the relationship matrix. The relationship matrix influences the final scores of each customer requirement and technical specification which is the output of the HOQ. This report tackles the problems of a large relationship matrix that is common in many HOQ matrices. The proposed framework introduces a two-and-a-half dimensional relationship matrix utilizing different abstraction layers. It enables engineers to build the matrix based on abstract layers which will contain lesser analysis factors. An algorithm completes the process by taking the values from them and calculates the scores in the final relationship layer. The purpose of this framework is to prevent information overload for the engineers and allow a clear structuring of information in the HOQ. A case study applying the framework using a Microsoft Excel programme illustrates the workings of the framework. A small number of participants did not show significant differences between the proposed and conventional frameworks of HOQ due to the weak p-values associated with the two sample unpaired t-test. The increase in information carrying capacity and the simpler analysis required were listed as some benefits of the framework while the risk of information loss and certain inconsistencies in the relationship matrix were limitations faced. Suggestions for future work that look into the limitations of the framework have been provided to enable more desirable results. Finally, it is believed that the introduction of this framework has created a new possibility of enhancing the HOQ matrix beyond its two dimensional capability.