A new hesitant fuzzy QFD approach: An application to computer workstation selection

Highlights

• A new hesitant fuzzy MCDM method based on linguistic term sets is proposed for QFD.

• The proposed method is relatively more efficient than the existing QFD approaches.

• It can aggregate the linguistic assessments of more than one decision maker.

• The relations between CRs&DRs and correlations among DRs via HFLTS are considered.

• A sensitivity analysis for robustness check is realized.

Abstract

Computer workstation selection is a multiple criteria decision making problem that is generally based on vague linguistic assessments, which represent human judgments and their hesitancy. In this paper, a new fuzzy quality function deployment (QFD) approach is used to effectively determine the design requirements (DRs) of a computer workstation. Hesitant fuzzy linguistic term sets (HFLTS) are innovatively employed to capture the hesitancy of the experts in this approach. More precisely, the proposed new QFD approach is the first study that determines the importance of customer requirements (CRs), the relations between CRs and DRs and the correlations among DRs via HFLTS. Additionally, HFLTS based Analytic Hierarchy Process (AHP) and Technique for Order Performance by Similarity to Ideal Solution (TOPSIS) methods are utilized in the computational steps to select the best computer workstation. A real industrial application is carried out to validate the implementation of the proposed approach.

Introduction

A workstation is a customized computer that is designed for specific scientific or technical application. Increasing competition and technological innovation in the industry and business world in general brings about new developments in the workstation design. However, workstations are usually designed arbitrarily with little consideration to the specific needs and requirements of their users. Considering additional benefits of tailor-made workstations that are customized for specific uses and needs, a customer-driven approach in workstation design would benefit companies. Such an approach would not only capture customers’ perspectives, but also raise the overall level of their satisfaction level. Quality function deployment (QFD) is a customer-driven tool that is widely used for product planning purposes. It can be beneficial to reach higher levels in customer satisfaction [1], [2]. Good design requires consideration of design aspects that clients want and expect. To address this, QFD uses a matrix called House of Quality (HOQ) [3] that translates Customer Needs or Requirements (CRs) into engineering characteristics or Design Requirements (DRs). The HOQ is constructed with the importance weights of each of the CRs, as well as the correlation matrix among DRs and the relationship matrix between CRs and DRs [1], [2], [3], [4], [5].

The importance levels of CRs, functional relationships among CRs and DRs, and the assessments of alternatives based on DRs are difficult to express precisely. Although crisp data are needed to design workstations, experts usually prefer to provide their evaluations in linguistic terms. The fuzzy set theory lets these linguistic assessments be incorporated into numerical analyses. The ordinary fuzzy sets have been recently extended to Type 2 fuzzy sets, hesitant fuzzy sets, intuitionistic fuzzy sets, non-stationary fuzzy sets and fuzzy multisets [6]. Hesitant fuzzy sets (HFS), which are developed by Torra [7], allow more than one value for defining the membership value of an element, enabling an expert better express his/her assessment [8]. In this paper, we prefer to use hesitant linguistic term sets (HFLTS) in the development of a new fuzzy QFD approach since HFLTS enable the integration of various linguistic evaluations assigned by experts as an inclusive linguistic interval. HFLTS have been used in several papers in the literature [9], [10], [11], [12], [13], [14], [15], [16].

Main features of the proposed hesitant fuzzy QFD approach are the use of HFLTS in the pairwise comparisons among CRs; the use of relations between CRs and DRs; the use of correlations among DRs, and the evaluation of alternatives. The weights of the CRs are determined by a hierarchical and pairwise comparison-based approach while the alternatives are ranked by using a hesitant fuzzy TOPSIS method. Besides, we propose a new approach taking the hesitant correlations among DRs into account in the HOQ operations. To the best of our knowledge, there is no QFD study based on hesitant fuzzy sets in the literature and this study is different from the other existing approaches since it considers the experts’ hesitancies in each phase of the QFD approach.

The remainder of this paper is structured as the following; Section 2 presents basic concepts of QFD and a literature review of fuzzy QFD methodology. In Section 3, the main concepts of HFS and HFTLS are given. Section 4 gives the proposed decision making approach which is based on hesitant fuzzy QFD. In Section 5, a case study is provided to demonstrate the applicability of the proposed method. The last section concludes the paper and gives some perspectives.