Martin Arvidsson
Chalmers University of Technology, Sweden
Ida Gremyr
Chalmers University of Technology, Sweden
Torben Hasenkamp
Chalmers University of Technology, Sweden
Download article
Published in: 10th QMOD Conference. Quality Management and Organiqatinal Development. Our Dreams of Excellence; 18-20 June; 2007 in Helsingborg; Sweden
Linköping Electronic Conference Proceedings 26:5, p.
Published: 2008-02-15
ISBN:
ISSN: 1650-3686 (print), 1650-3740 (online)
During the past two decades quality improvement efforts have been moved upstream to the design stages of products and processes. This is confirmed by initiatives often run under the heading of Design for Six Sigma (DFSS) (Chowdhury; 2002 and Tennant; 2002); Variation Risk Management (Thornton; 2004) or Robust Design Methodology (Arvidsson and Gremyr; 2005). In line with this trend there has been a particular emphasis on the development of statistical techniques to create products and processes insensitive to sources of variation or noise factors; respectively (see e.g. Hunter (1985); Box and Jones (1992); and Arvidsson; Kammerlind et al. (2001)). These techniques are a part of Robust Design Methodology (RDM) that Arvidsson and Gremyr (2005) in their review article define as “…systematic efforts to achieve insensitivity to noise factors. These efforts are applied in all stages of a product design process.” To simplify matters; in the following; ‘product’ refers to a ‘product or process’.
Despite extensive research on techniques useful for creating robust designs; i.e. designs that are insensitive to noise factors; the industrial use and knowledge of RDM is still low; see Gremyr; Arvidsson et al. (2003). Arvidsson and Gremyr (2005) argue that one reason for this limited use of RDM might be an inattention to the overall framework of RDM in favour of development of applicable statistical methods. Gremyr and Johansson (2005) interview companies applying the methodology and suggest possible reasons underlying its low level of use. One reason is the lack of support for the use of tools applicable in RDM; which in turn implies a need for a structured framework.
The purpose of this paper is to operationalize RDM in terms of principles; practices and techniques; this way of operationalization was previously applied by Dean and Bowen (1994) with respect to the concept of “Total Quality”. In this paper we characterize RDM by means of three principles: awareness of variation; insensitivity to noise factors and continuous applicability; meaning that RDM efforts are appropriate in all stages of design. These principles in turn are implemented through a set of practices; which are the focus of this paper. Earlier research forms the basis of these practices and will be presented briefly in the following section.
Robust design methodology; variation; noise factors; control factors; P-diagram; quadratic loss function
