An integrated methodology for the analysis and design of cellular flexible assembly systems
Abstract
This research proposes a methodology for the planning, analysis, and design of cellular flexible assembly systems (FASs) with a higher degree of sophistication and automation. A prototype system, which is called CASCADE (Cellular flexible Assembly Systems using a Computerized Analysis and Design Environment), integrating a design framework with a decision supporting tool, has been developed. The decision supporting tool starts with the creation of an object-oriented data and knowledge schema as the common information model for cellular FAS design. A prototype SQL language has been built to facilitate the query process. A planning system has been devised to explicitly describe products, processes, plans, and the assembly system with levels of abstraction so that the analysis and design can follow. It eases data entry and maintenance and reduces data inconsistency by interfacing with the database. The design framework first selects the best assembly technology for each product and its operations through a multi-product Econ-Tech model. Based on the Econ-Tech result, a four-phase model for simultaneously selecting workstations, assigning work elements to workstations, generating assembly plans, balancing lines, and minimizing assembly costs has been developed. This model satisfies constraints of assembly precedences, production rates, cycle time requirements, and facility availabilities. To group products into families and workstations into cells, a general clustering method (GCM) has been formulated. The GCM not only strengthens weak points of existing group technology models but also explores new concepts for solving practical manufacturing clustering problems. Finally, the flow and layout analysis is performed by the use of a heuristic pattern matching algorithm. The proposed methodology can be used effectively in the planning, analysis, and design of a cellular FAS. By applying the same concept, this methodology can be easily adapted to solve a wide range of cellular manufacturing system design problems. Equally important, the methodology behind this research, where many distinct application modules share a common model, offers significant benefits in complex manufacturing problem solving.
Degree
Ph.D.
Advisors
Moodie, Purdue University.
Subject Area
Industrial engineering
Off-Campus Purdue Users:
To access this dissertation, please log in to our
proxy server.