GOALI: Quality Analysis in Flexible Manufacturing Systems: A Systems Approach

  • Li, Jingshan (PI)

Grants and Contracts Details

Description

The objective of this research is to establish an analytical framework to investigate the impact of flexibility on product quality in flexible manufacturing systems and apply the results to production systems at GM. Specifically, methods addressing the following problems will be developed: Analysis: Analytical modeling for quality performance in flexible manufacturing systems. Improvability: Identifying the machine and the product type that impede the quality performance in the strongest manner. Design: Determining the appropriate batch size or production sequence so that both the quality requirements and customer orders are satisfied. Applications: Implementing the techniques to the painting and machining operations at GM Detroit Hamtramck Assembly Center and Romulus Engine Plant, respectively. Manufacturing systems are becoming more and more flexible, e.g., GM is making multiple models of vehicles with many options on one production line, and flexibility is the key to success of its global platform strategy. Similar trends also occur in other companies. Although flexible manufacturing systems have been studied extensively, most of the research is devoted to the issues of flexibility measurements, investment cost, tradeoffs between flexibility and productivity, etc., while neglecting the question of quality. However, flexibility and product quality are tightly coupled. Therefore, an analytical method to investigate the coupling between quality and flexibility is necessary and important. Unfortunately, the current literature does not provide such a method. The approach of the research is based on analytical investigation of Markov processes that describe the flexible systems at hand. Quality performance is modeled as a function of transitions among multiple states characterized by product types and sequences, and propagation of quality variations among the machines. The challenge of this research lies in discovering an appropriate mathematical description of such function and its arguments. The intellectual merit of this research is the establishment of a novel analytical method to study the interaction between flexibility and quality, and to provide insights for production system design from the point of view of quality. Such a method will enable us to evaluate quality performance, identify bottlenecks, and design scheduling policies (sequencing and batching) to achieve the best quality and reliable demand satisfaction. The successful completion of this research will open up a new direction in manufacturing systems research and build a solid foundation for integrated study of cost, productivity, quality and flexibility. The broader impact of the research is in providing the production engineer with quantitative tools for analysis, improvement, and design of flexible manufacturing systems with respect to quality. The results obtained will be implemented in GM, disseminated through lectures and presentations, and included in the course "Production Systems Engineering" offered by the PI in University of Kentucky. The graduate students involved will be trained to be experts in the area, receive significant exposure to important industrial problems, and gain practical experience. Women and students from under-represented groups will be actively recruited to participate in the project and outreach activities will be carried out through exhibitions and visits for K-12 students. Both the PI and Co-PI will collaborate to contribute to all research addressed in this project and supervise students jointly. Up to 30% of the PI's project time will spend in GM and 25% of the Co-PI's time will be dedicated to this project. The PI's effort will be centered on developing analytical methods and that of Co-PIon data analysis, model validation and implementation.
StatusFinished
Effective start/end date10/1/076/30/10

Fingerprint

Explore the research topics touched on by this project. These labels are generated based on the underlying awards/grants. Together they form a unique fingerprint.