Part flow orchestration in distributed manufacturing processing
Abstract
All of the standard approaches to the management of the flow of workpieces through manufacturing systems rely either explicitly or implicitly upon the notion of centralized control to achieve "optimal" performance according to some criterion. There is, at least in principle, an alternative. A manufacturing system could be viewed as a collection of intelligent machines serving a population of intelligent workpieces, with each member of each type of entity making its own decisions. One advantage of this perspective, if it can be made to work, is that the software could be completely modular. That is, changes in any of the individual system components would not require modifications anywhere else in the system. This thesis provides an investigation of this new concept by proposing a method for part flow coordination that capitalizes upon the concept of both intelligent workpieces and intelligent machines, demonstrating its feasibility, and establishing that the approach does yield promising performance. The coordination of both physical and information flow is provided by a unique distributed bidding procedure. In this procedure, machines bid for operations on workpieces according to their own capabilities and current status, while each workpiece autonomously evaluates the bids offered according to its own requirements and objectives.
Degree
Ph.D.
Advisors
Solberg, Purdue University.
Subject Area
Industrial engineering|Artificial intelligence
Off-Campus Purdue Users:
To access this dissertation, please log in to our
proxy server.