Impact of Information Sharing on Statistical Quality Control

Impact of Information Sharing on Statistical Quality Control

Fugee Tsung

   By:

       Garima Verma

       PGDIE 42

       A-31

Abstract—With recent advances in information technology (IT), the re- search on and practice of information sharing is now having a significant impact on many aspects of supply chains. Nevertheless, few investigations focus on the impact of information sharing on product and process quality. Furthermore, it is still not clear how and what information should be shared or used, and how to quantify the benefits of information sharing in terms of quality improvement. In this research, a “matching problem” is used to demonstrate the impact of information sharing on quality. We quantify and compare the impact of different information-sharing strategies on process and product quality, and suggest that real-time information sharing may lead to dramatic quality improvement for an assembly process, the example here being a two-stage supply chain. The proposed approach to evaluate information sharing in terms of quality improvement can be extended to a more complex supply chain.

I. INTRODUCTION

In past annual benchmarking meetings of some U.S. automotive companies, it has been observed that the quality of the individual parts they produce usually equals the quality of Japanese parts, but that assembled American automobiles often undermine rather than showcase the quality of their parts. One accepted reason for the high quality of assembled Japanese automobiles is the well-established cooperative relationship that exists between Japanese manufactures and their suppliers, who they generally regard as extended factories. The Japanese case suggests that supply-chain cooperation may be a critical factor for quality improvement and greater competitiveness in the global marketplace.Supply chain management (SCM) has received an enormous amount of attention in both industry and academic circles. One recent interest in SCM is in incorporating in- formation flow among various members of a supply chain.

A. Matching Problem

In this paper, a “matching problem” is used to illustrate the impact of supply-chain information sharing on quality. We specifically focus on the process capability and quality improvement of two components that are used in fuel injectors, in an actual automotive assembly facility: bodies from a tier-1 supplier and needles from a tier-2 supplier. Thetier-1 supplier sorts components such that a needle from the upstream manufacturer can be matched with the body that they have produced in order to satisfy the tolerance requirements. If a needle is slightly too large for the body, the assembly may be sticky or have reliability problems.

B. Quality Measure

Process capability indices such as C_p  and C_pk  have been widely used as a measure of quality, with Cp measuring potential process performance and C_pk  measuring actual process performance [12], [17]. There are many manufacturers in the U.S. and Japan who require sup- pliers to produce items with C_p and C_pk of more than 1.0. However, we will show that greater process capability for a single process may in fact lead to worse overall assembly quality.

C. Key Results and Contributions

In this paper, we quantify the impact of various information sharing strategies on process and product quality in a two-stage supply chain framework. Our key results can be summarized as follows:

•         Without information sharing, although an individual process may be controlled to have greater process capability and dimensional quality, that enhancement may in fact lead to poor assembly matching.

•         Controlling the process based on information sharing will lead to better assembly matching, even though the capability and dimensional quality of an individual process may be adversely affected.

•         As the process data are not fully utilized by merely sharing simple descriptive statistics, real-time information sharing may have a greater impact in quality improvement than non real-time information sharing.

•         For the purpose of benchmarking, the lower bounds of quality loss for both real-time and non real-time information sharing pro- cesses are derived.

II. PROCESS WITHOUT AND WITH INFORMATION SHARING

A. Without Information Sharing

           Needle and body manufacturers, like many conventional tier-1 and tier-2 suppliers, have little communication with each other, and there- fore represent a situation of no information sharing. To match two sets of components to form assemblies without information sharing, Glover developed an algorithm involving measuring and sorting all components to maximize the matching number. Lee, Hausman, and Gutierrez suggested grouping the components into different classes for operational convenience.

B. With One-Way Information Sharing

For a process with information sharing, we will match the components to form an assembly in the same sorting and matching approach. However, an advantage is derived from the proper control of the component distribution based on feed-forward information before matching. If one-way information sharing is possible, which means that the down- stream manufacturers can obtain information from the upstream manufacturers, but not the other way around, there is still a chance to improve the quality of assembly matching. There have been some studies to determine the optimal machine setting, i.e., the optimal process mean of a single process, based on upstream information.

C. With Two-Way Information Sharing

If two-way information sharing is possible, which means that the upstream and downstream manufacturers can obtain information from each other, there will be a greater chance to improve the quality of assembly matching. Lee et al and Gutierrez et al deal with the cooperation of settings for multiple processes for assembly operations, which is only achievable with two-way information sharing. Different criteria may lead to different suggestions for both machine settings. Our study is based on the criterion of maximizing long-run expected matching.

III. PROCESS WITH REAL-TIME INFORMATION SHARING

Process data are not fully utilized by merely sharing simple descriptive statistics, e.g., the mean and standard deviation values. With recent advances in IT, real-time process data sharing has become common, and this has great potential for further quality improvement. The bounds are commonly established on the basis of engineering judgment, taking into consideration the cost of adjustment and the cost of a mismatch.

IV. INFORMATION QUANTIFICATION WITH EXAMPLES

 In this section, we investigate the impact and quantify the value of information sharing in terms of improvement in quality loss. The difference between these two bounds indicates the possible quality improvement we may make by real-time information sharing. Here, we simulate 30 days of operation with component production of 10 000 units/day, and treat daily production as a batch. The specification limits of both the needle and body are scaled to be    5, with nominal values equal to zero. Four types of processes are studied. The process capability of body manufacturing

pk

 C^b has deteriorated to 1.68 due to the active adjustment, but itis still within the acceptable range (>1.0). More importantly, the quality loss of the matched assembly is reduced to 0.12, which is fairly close to the theoretical bound.

V. CONCLUSION

In this research, we quantify and compare the impact of different information-sharing strategies on process and product quality. We indicate that real-time information sharing may lead to dramatic quality improvement for an assembly process, an example being the two-stage supply chain. In reality, some practical issues need to be addressed. First, to motivate the information-sharing strategy, it is critical to suggest a mechanism that spreads the benefits of information sharing between the supplier and manufacturer.