An Integrated Object Model as a World of Model Components for an Activity Network Based Simulation Approach

One of the main obstacle for an integrated use of simulation over different planning areas and stages are the various views on a production system. Therefore, a world of model components is developed, which enables the co-existence of different views and levels of detail in the same simulation model while maintaining its consistency. This is achieved by combining a network based simulation approach with the object-orientated technology. The fundamental idea is to offer the opportunity to re-use existing models for the investigation of different aspects of a production system. The approach is abstractly described as a generic object model and is thus, independent from a concrete simulation language, tool or environment. This paper describes the generic object model as a world of modeling components of an object-orientated simulation approach. Furthermore it discusses the modeling requirements for such a simulation approach. ACCEPTANCE OF SIMULATION TOOLS IN PRACTICE For the design and control of production systems, simulation has proven to be a powerful tool. However, investigations of the market situation have shown that many companies are not willing to use simulation as a permanent planning tool (Schmittbetz 1998, p. 18; Zülch et al. 2002, p. 39). One of the most named reasons for this is the expenditure associated with modeling a production system (Rabe 1999, p. 3). The problem is that most simulation tools which are available on the market today, do not support the re-use or exchange of models e.g. exploiting them for different planning aspects. In order to realize the exchange and the re-use of models there can be used two fundamental possibilities. Either one defines a standard interface between different tools or one combines the different views of a production system in one world of model components, e.g. the view of material flow, the view of manufacturing or the view of personnel planning in one simulation application. During the development of such a simulation concept the problem of semantic compatibility is likely to occur. Each view of a production system shows different parts of it. If one wants to combine these views, the world of model components, on which the simulation application is based, has to contain the main elements of all these views. If an additional task of a simulation tool is to reach a flexibility similar to that of other planning tools, e.g. operation plans or plain text descriptions, the modeling methodology has to support a step by step development of the simulation model. In this case, it should be possible to simulate the model in every development phase. Thus, the simulation concept has to support various levels of detail and also the addition and removal of model parts. THE IDEA OF AN INTEGRATED OBJECT MODEL AS A WORLD OF MODEL COMPONETS OF A PRODUCTION SYSTEM In order to take a step towards a solution of the described problems, an object-orientated simulation concept is under development at the ifab-Institute of the University of Karlsruhe. The fundamental idea is the development of a generic world of model components (cf. VDI 3633, p. 11) of a production system including the simulation functionality of the model’s elements. In this world of model components, various application views are combined. Thus, the derived models can be used in different planning steps. For the development of the world of model components object-orientated modeling principles were chosen. With the help of the object-orientated technique of specialization the functionality of an object model can be structured hierarchically. In this manner, abstract object classes may be defined with more universal functionality compared to traditional approaches. This universal functionality can be adjusted to the requirements of a specific application field, with the help of specialization. In this way, common features and variations in different application views can be brought into a defined context. Separate objects can be combined using the principle of composition. Thus, more complex functionalities can be realized. Experience has shown that the hard part of representing a production system is to depict the processes and their respective control because these aspects are quite abstract and it is rather difficult to get an intuitive correspondence between model and reality. One possibility is to depict the processes by modeling them based on activity networks. This technique is found mainly in project management and business process modeling (Schmid 1998, pp. 20). It makes intuitive handling of the activities together with their releases and dynamics to a great extend possible.