The objective of the workshop was to provide a clear understanding of the benefits of computer supported tools for enterprise modelling and simulation to better manage the enterprise and react to changes in the business environment. To reach this goal, the agenda was composed of both presentations and small group sessions with tools. The particular approach of the workshop was , that the participants could get involved in developing or customising models of typical business processes by working in small groups. Thereby people could gain practical experience about the business value of enterprise modelling. In addition, the participants could understand and compare the properties of the presented tools and assess their particular capabilities in process modelling and simulation.
The workshop was organised by the CIMOSA Association, jointly with ADITEC and with the 'Fraunhofer Demonstration Centre for Simulation in Production and Logistics'. ADITEC, the Aachen Demonstration Laboratory for Integrated Production Technology, is a pilot factory located at the Technical University Aachen . The workshop was supported by the European Union through CIM Europe and the Accompanying Measures Program. The workshop was attended by 45 participants , half of which were users of enterprise modelling and simulation tools from industry (12) and from academia (10).
In the first presentation, Professor Weston from the MSI Research Institute at the Loughborough University of Technology elaborated on the requirements for enterprise modelling as well as for model enactment, and on the necessary properties of tools. He gave on overview of the candidates for modelling methods and languages and reported together with Paul Gilders on their experience with methods and model driven CIM tools in the enterprise engineering workbench SEW-OSA at the institute. He concluded that a consolidation of the 'best' tools is needed and that tools must become more usable for model enactment and for supporting the exchange of reference models or partial models, in order to make modelling more effective through re-use of model parts.
In the second presentation Valter Vico firstly reported on enterprise engineering with two CIMOSA applications at FIAT AUTO. The first application was concerned with model engineering to manage gearbox production in a component plant. Modelling the real environment of the gearbox production system with CIMOSA provided a significantly reduced model engineering time, highly increased modelling flexibility and much more meaningful analytical capability
The second CIMOSA application was concerned with modelling the product design process, as a task of Workpackage 5.2 of the AIT Initiative . FIAT has selected the CIMOSA methodology because it enables
Another development of business engineering with CIMOSA is concerned with the 'Virtual Factory' , where entire production process areas can be modelled to evaluate the behaviour of the system . To date, 17 production areas have been modelled, emulated and simulated in order to investigate the production control flow and work-in-process behaviour. Furthermore, an ongoing project to develop a reference model of the logistic chain with a library of standard building blocks for Small and Medium Sized Enterprises (SME's) was reported.
In the third presentation, Dr. B-D. Becker from the Fraunhofer Institute for Manufacturing Engineering and Automation (IPA) in Stuttgart/Germany was addressing business engineering in production and logistics with simulation methods and tools. Simulation has a particularly high business value in production logistics with fast changing constraint driven demands. Cost benefit ratios of greater than one to six have been achieved in more than two hundred simulation projects where IPA Stuttgart was involved. Two examples were outlined in some detail, one on the complex resource planning in the maintenance facility for the ICE high speed trains of the German Federal Railroad, the other on manufacturing resource planning in a copper plant. Both simulation examples were carried out with the SIPMLE++ tool and demonstrated a very effective role of the human in the simulation task. Future developments will provide solutions to distributed modelling and simulation tasks for instance solving of complex, constraint limited scheduling tasks in large organisations and following the paradigm of the Fractal Enterprise . This concept was explained with the example of decentralised planning in a very large steel producing enterprise.
Eleven tools for process modelling, for simulation or for both have been shown, in alphabetical order :
ARIS (IWi Saarbrücken), CASTSim (University Dresden), CREATE (IML Dortmund), FlowMark (IBM), GTVOICE/McCIM (EP VOICE), MO2GO (IPK Berlin), PACE (GPP Munich), PMT(IISST Dortmund), (PROPLAN (WZL Aachen), SEW-OSA (University Loughborough), SIMPLE (IPA Stuttgart). Most of these tools have been presented in small work sessions to develop or modify process scenarios as shop floor control, order processing and manufacturing resource planning. The objective was to provide practical hands-on experience with different tools to business users and to encourage users to become more involved in enterprise modelling and simulation technology.
Starting with statements on the present situation and future developments of tools for business engineering made by the panel of tool providers, several issues came up in the plenary discussion .
Today the existence of tools for modelling and simulation is not well known though there is a high potential for benefits. For instance, the German market for potential savings with simulation is estimated to 18 Billions DM of which only 5 % have been exploited , (compared to 66 % in the CAD market sector). Only 8% of potential users know about process modelling tools based on Petri-nets. Therefore information dissemination through appropriate channels as workshops, television and multimedia is highly recommended. Local events held in the native language could be more effective to reach SME's than events on European level.
It was agreed that all promotion activities for enterprise modelling and simulation have to start with identifying and solving important problems of the end user demonstrating a true added value to his business.
With the present landscape of tools for modelling and simulation, depending on their purpose, functions and capabilities, an interchange of data or models is still very rudimentary. It was agreed that connectivity between tools is definitely needed, some results have been achieved
( interfaces ARIS - FlowMark , ARIS - SIMPLE++) .
With regard to the methods, the meaning of the constructs and its semantics (What is a process, an activity ?) must be unambiguous. Furthermore, a basic set of graphical symbols (which can be customised for tools or for applications) should be common.
As well known, the standardisation process is usually slow and tedious. In areas which are under development as enterprise modelling and simulation, discussions are ongoing what should be modelled and in which level of detail. (Example ENV 40003 ). It was agreed that standardisation of a framework can speed up the main avenues of development in enterprise integration and can be of common benefit to both users and vendors. On the other hand, standardisation must not hamper technology development.
Martin Zelm
www-cimosa@cnt.pl, 04.08.1995 (last update: 14.01.1996)