Modeling Agile Manufacturing Cell using Object-Oriented Timed Petri net

An Object-Oriented Timed Petri-net (OOTPN) is proposed for modeling agile manufacturing cell (AMC). It consists of four phases: (1) grouping manufacturing resources into classes; (2) constructing general OOTPN block for each manufacturing resource class; (3) defining joint transitions and suitable control decision rules; (4) generating complete OOTPN model for agile manufacturing cell. This OOTPN approach possesses the characteristics of both object-oriented method and Petri-nets. It can describe the attributes and behaviors of AMC. Furthermore, the model based on this approach has the high degree of modularity, flexibility and reusability. In comparison with other modeling methodologies, the main advantage of OOTPN is that the cell model can be regenerated quickly by selecting or modifying appropriate pre-defined modeling objects when the agile manufacturing cell is reconfigured. INTRODUCTION Manufacturing enterprises have recognized that the rapid introduction of new and customized products to meet market demands is a key factor that ensures their survival from fierce competition, which in turn requires manufacturing systems be reconfigured quickly according to different products. The manufacturing cell is the basic unit for the manufacturing system, which carries out the final accomplishment of the production plan. The main task of a manufacturing cell is to accomplish assigned operations on assigned parts with reasonable cost in a time horizon and guarantee processing quality. In the future, it should be able to reconfigure rapidly to meet changing and unpredictable market demands. The concept of AMC (Agile Manufacturing Cell) arises from the idea of agile manufacturing. AMC could be a virtual manufacturing cell, that is to say, the physical layout of the manufacturing resources of a workshop is fixed, but the logic composition can be redesigned or reorganized to meet different production requirements. The essential characteristic of AMC is that it can be reconfigured rapidly according to production tasks or state variation of a manufacturing environment. Its agile configuration brings additional values such as rapidly new product introduction, accommodation to unpredictable demand, higher equipment efficiency, etc. In order to accommodate the changes of production tasks and manufacturing environments, it is necessary to have suitable methodology and tools for rapid and agile configuration of manufacturing cells. To achieve this goal, rapid and reusable modeling techniques are demanded. There have been some important modeling methodologies such as queuing networks, mathematical programming, activity cycle diagrams, IDEF0 diagrams, Petri-nets, O-O (object-oriented), and so on. However, all of these methods have certain limitations. The assumption of some random processes and their parameters may not well reflect the real manufacturing processes. Activity cycle diagram cannot represent the relationships of conditions and events in discrete event dynamic systems. Although Petri-nets have been widely applied to modeling, controlling, and analyzing the system’s dynamic behaviors due to the characteristics of the graphical representation and mathematical analysis of control logic, the efforts of modeling and analysis become quite cumbersome when the system becomes more complex. Furthermore, these methods lack of reusability. The O-O modeling technique provides a one-to-one correspondence between the entities in a real system and the objects in its model that represent them. It possesses high modularity, maintainability and reusability. These advantages have led many researchers to apply O-O methods to manufacturing system modeling. Nevertheless, a powerful tool supporting quantitative analysis and validation is still unavailable in the O-O method and its depiction to real system lacks abstraction and accuracy. This paper aims at developing an OOTPN (Object-Oriented Timed Pert-Net) modeling method for AMC modeling. This method is an instance of composite modeling, which possesses the characteristics of O-O technique and Petri-net. The rest of this paper is organized as follows. In Section 2, the characteristics of AMC and its requirements on system modeling are discussed. In Section 3, the definition of OOTPN is presented. Section 4 describes the modeling process of AMC using OOTPN method. An example is provided in Section 5. Finally, conclusions are given.