Study on Quality Prediction Technology of Manufacturing Supply Chain

Introduction

21st century is the quality century, in which the traditional “vertical integration” single enterprise quality control mode has changed into “horizontal integration” supply chain quality management mode in manufacturing industry (Feigenbaum, 1983; Mettler et al., 2012). Modern manufacturing supply chain is a complex system, with human resources, capital, materials, energy and information as input, and the tangible products and intangible services as output (Huang, 2003; Carmen & Manoj, 2005; Kart et al., 2010), any link in the chain with a problem can affect the whole supply chain’s quality and finally the products quality (Zhang et al., 2011). Although supply chain quality control is a hot topic among researchers (Kuei et al., 2001; Thomas, 2008), supply chain quality prediction is actually an important but unsolved problem in manufacturing industry.

In 1995, CENTRIM of Brighton University finished the project “Total quality management in the supply chain” with Bath University as a beginning of supply chain quality prediction. Beamon and Ware (1998) worked out a process quality model for the analysis improvement and system control of supply chain. Duan and Tong (1999) studied on quality assurance system and quality prediction methodologies of supply chain. Also Zhang and Huang (2003) studied on supply chain quality prevention decision problem under asymmetric information. Furthermore, Zhou (2005) and Xu (2008) researched on quality management of supply chain in their maters’ theses within supply chain quality prediction technology.

Nowadays, there are more and more researches on supply chain quality prediction (Tsai & Wang, 2004), some of which are applied in manufacturing industry (Hans et al., 2000; Tang, 2004). However, there are still many problems need to be solved, mainly as follows:

• •

  Key quality characteristics extraction of the product realization process under supply chain environment;

• •

  Index data acquisition of dynamic supply chain quality prediction;

• •

  Nonlinear problem in mathematical modeling of supply chain quality prediction.

To overcome the shortcomings, an approach of predicting the quality satisfaction degree of manufacturing supply chain in the next time period is proposed in this paper. Quality prediction models based on customer satisfaction and process control are built, and technologies used in quality prediction are studied. While firstly the prediction index system is established, then the adaptive variable step back propagation method (ABPM) is used in quality prediction, and finally an example is given to illustrate the approach, which provides a new direction for the quality prediction technology research.

Modeling Of Manufacturing Supply Chain Quality Prediction

The 3D Model

Supply chain quality prediction should not only focus on the quality of “interface” (customer satisfaction), but should also attach great importance to the quality of “process” (process control). Customer satisfaction should be combined well together with process quality control in supply chain quality prediction (Min & Zhou, 2002; Goknur & Turan, 2010). According to the characteristics of supply chain quality prediction, based on systematically thinking, the 3D model of supply chain quality prediction is built up with three parts—quality prediction related parties, quality prediction contents and quality prediction technologies, as shown in Figure 1.

Figure 1.

The 3D model of manufacturing supply chain quality prediction