An agile manufacturing firm forms partnerships with other manufacturers as necessary to design and manufacture a product quickly in response to a market opportunity. Ideally, this partnership is a virtual enterprise in which the participating manufacturing plants realize a portion of the product design and act not only as subcontractors but also as partners who cooperate to lower the product cost, improve its quality, and reduce the amount of time necessary to bring the product to market. In the virtual enterprise, the partners electronically exchange the necessary information for design, process planning, production planning, inventory management, testing, delivery, and billing. In addition, the partners must have business processes that allow them to exchange such data and to arrange the necessary financial transactions.
This project addresses a special case of the virtual enterprise: one firm designs the product and partners with other firms for the product's manufacture. In order to form such a partnership, the firm needs to create a superior design and select the optimal partners. This study supports both needs. First it provides methods to evaluate, early in the product life cycle, a proposed design with respect to the capabilities of candidate partners. Second, it provides tools for comparing candidate partners. The design evaluation has two significant benefits: the manufacturing firm critiques the design considering the partner-specific strengths and weaknesses that are related to the manufacturing requirements (not the manufacturing plant's general performance), and the firm receives vital design feedback at an early point in the design cycle, when modifications are less expensive.
The result of this project is an integrated approach for design evaluation and partner selection. The approach employs novel information models for representing product designs, the capabilities of manufacturing processes, and the capabilities and performance of manufacturing firms, and it provides two types of design evaluation methods (variant and generative) as well as a multi-criteria partner selection procedure. The variant design critiquing method uses data about the products fabricated by a candidate partner to describe that manufacturing plant's capabilities and performance. The procedure searches for product designs that are similar to the new design, and the plants that manufacture these similar products are candidates for partnering, since they have the desired capabilities. The generative procedures use information about the manufacturing processes available at the manufacturing plants. The procedures generate and evaluate manufacturing alternatives for the new product. These alternatives and their performance depend upon the plants' process capabilities. The partner selection procedure allows the firm to stress different performance measures (cost, quality, and lead time) and find superior manufacturing alternatives. A selected manufacturing alternative will identify manufacturing plants with the desired capabilities.
The primary application of this approach is a class of mechanical and electromechanical products. Specifically, we consider flat mechanical products and microwave modules (MWMs), which comprise flat mechanical substrates of complex shape, an artwork layer, and electrical components. We have implemented this approach in a software system that a designer can use to define a feature-based product model, generate concise product descriptors, search for and sort similar products, generate alternative plant-specific process plans, evaluate those plans, and compare them to find the most suitable combination of processes and manufacturing partners.
The most significant contribution of this work is the use of design evaluation as a method for partner selection, facilitated by the partner-specific design evaluation techniques. Important results include the development of novel information models (an integrated STEP-based product model, an object-oriented Group Technology information model, and a manufacturing resource model), and the decision support methods for the formation of virtual enterprises (variant design critiquing, generative design critiquing, and partner selection).
This web page describes the functionality of the system that has been developed and provides an overview of the approaches that were developed to support this functionality. The web page provides references to papers that describe the details of the approach and the system implementation. The remaindering pages are structured as follows: