【正文】 f Virtual factory. The activity Execute simulation for virtual factory is a separate simulation model of VM system. With this virtual factory, parameters (, utilization, operation time, etc.,) associated with operating a flexible manufacturing cell are simulated. And these results can provide the possibility of controlling manufacturing processes and predicting potential problems in the real manufacturing. 3. Object modeling for virtual flexible manufacturing cells Objectoriented technology may provide a powerful representation and classification tools for a virtual flexible manufacturing cell. It may also provide a mon platform for the information sharing between submodules, and provide a richer way to store/retrieve/modify information, knowledge and models and reuse them. In the context of an object oriented approach, a model is simply an abstraction, or a representation of an objects or process. VFMC requires a robust information infrastructure that prises rich information models for products, processes and production systems. As shown in Fig. 2, three models, that is product model, facility model, and process model, are developed for virtual flexible manufacturing cells. A product model is a generic model used for representing all types of artifacts, which appear in the process of manufacturing. It represents target products, which include conceptual shape information as well as analysis module for a specification, productivity, and strength. A facility model contains information about machines consisted of a virtual flexible manufacturing cell. By using the model, innovative tooling and methods can be evaluated without the cost of physical machine prototypes and fixture mockups. A process model is used for representing all the physical processes that are required for representing product behavior and manufacturing processes. Product model A product model holds the process and product knowledge to ensure the correct fabrication of the product with sufficient quality. It acts as an information server to the other models in the VFMC. It also provides consistent and uptodate information on the product lifecycle, user requirements, design, and process plan and bill of material. An instance of Class Part provides detailed information about a part to be fabricated in VFMC. Subclasses like ProcessPlan, BOM, and NcCode, are aggregated into the class Part. Classes Process Plan and BOM manipulate information and data associated with process plans and bill of materials, respectively. Class NcCode deals with NC programs, which interacts with CAD/CAM systems. With incorporation with the facility model, this developed NC programs can be verified and checked for collisions and interference with any workpiece or tooling in the fixture. This can avoid costly machine crashes and reduce risk during initial equipment installation and produce launch. Furthermore, productivity can be improved by avoiding nonproductive time for program prove out on the machine tool and by using the 4 simulation environment to train operators of new machines. Facility model Real manufacturing cell may consist of NC machines, robots, conveyors, and sensory devices. The architecture of class corresponding to the real manufacturing cell is shown in , and represents the factory model. In VFMC, characteristics of the factory model include a detailed representation of machine behavior over time, a structure to the model that can configure and reconfigure easily, and a realistic and threedimensional animation of machine behavior over time. Virtual machines defined within this model may be used to estimate accurately the merit of a process plan, and, based on this evaluation, determine appropriate process conditions to improve (and even optimize) the plan. Virtual robot contributes to unload and load parts into/from machines, and is used to find optimal paths without any collisions. With virtual operation, the fidelity of the machining and robot utilizing time and cost estimates is expected to improve. In addition, accurate modeling will predict the quality of the machined part, which cannot be determined easily and reliably without producing several physical prototypes. This information is invaluable to both the designer and the process planner. Physical entities such as machines and workpieces have the explicit representation as 3D models for their shapes, positions, and orientations. 3D models are conveniently used for calculating, geometrical attributes, checking spatial relations, and displaying puter graphics. Process model By assigning a finite set of states to each device in a cell (idle, busy, failed, etc.), the process of cell control can be modeled as a process of matching specific state change events to specific cell control actions, decision algorithms, or scripts. With this model, cell processes are represented a Task Initiation Diagram (TID) using an objectoriented approach. The methodology behind developing TID regards the tasks to be performed by the cell or any of its constituent machines for being primal, and employs the multilayered approach. Sensory signals indicating the change of state of machines are used to trigger or initiate tasks. A task may be simple and require a relatively short time to execute, or may be plex and lengthy. Formally, a Task Initiation Diagram (TID) is defined as the fourtuple TID=(T, SR, C, O). Task Initiation Diagrams are posed of two basic ponents: a set of Rest states SR and a set of tasks T. Tasks, in turn, are classified into three groups: the c