Globeman 21

TG4 WP1 - Methodology for Business Process Analysis and Design (Enterprise Integration Methodology)


Who did What and When


Participants of TG4WP1

Griffith University, Enterprise Integration Group, School of Computing and Information Technology, Faculty of Information and Communication Technology, Griffith University, Nathan (Brisbane) Queensland 4111 Australia, tel+61-7-3875 5039 fax+61-7-3875 5051

Peter Bernus (TG4Wp1 leader, 3 years)

Greg Uppington (for 2 years)

Raymond Woo (for 1.5 years)

Brane Kalpic (for 3 months)

CSIRO Division of Manufacturing Science and Technology Locked Bag No9. Preston, VIC., 3072 Australia, tel +61-3-9662 7700 fax +61-3-9662-7851

Laszlo Nemes (3 years)

Mingwei Zhou (2.5 years)

John Mo (1 year)

Acknowledgements: this work was carried out in collaboration with many people to whom the TG4WP1 members are indebted for their continuing support and collaboration. In particular we must mention from the Globeman 21 Consortium the VRIDGE Team members from Toyo Engineering Corporation, from Takenaka and from Yokogawa, and from NCR-METIS, the Concept team, the Toronto EI Laboratory, members of the TU Denmark and Odensee teams as well as others with whom we did not work closely together but whose work inspired us through seing their results and problem statements at the Globeman 21 plenary meetings. From outside Globeman 21 we would like to mention the members of the IFIP-IFAC Task Force, the ISO Working Group TC184/SC5/WG1, and individuals who for shorter or longer time visited our groups to exchange ideas or participate in our work, in particular Chris Menzel, Ted Williams, Guy Doumeingts and Gunter Schmidt. We apologise if someone if left out by pure oversight.

When

The Globeman 21 project ran from mid 1996 till mid 1999. In Australia it was funded by the Department of Industry, Technology and Tourism.

What

Aim

1. Develop a comprehensive but easy-to-use methodology for enterprise integration. Several integration architectures and enterprise engineering methodologies are described in the literature, however, these developments are not yet at a stage where they can provide an easy to use methodology which is understandable by a cross-functional design team. TG4WP1 was to provide increasingly refined versions of an enterprise integration architecture and methodology and build on the most promising existing integration architectures known.
2. Provide guidance and consultancy to GM21 Demonstrator projects on the basis of the best practice methodology and previous experience. In return the demonstrators were to serve as a test-bed for new ideas and methods to demonstrate their feasibility as usefulness in industrial context.

Results

I. GERAM, Change Identification Methodology and Application to a Demonstrator Project

The development, based on the work of the Griffith and CSIRO teams and on the work of the IFIP-IFAC Task Force, of a generalised architecture and methodology which allowed enterprise management to select and combine elements of best practice and create a complete capability of enterprise engineering for small or large scale change.

A thesis specifically adressed the problem: is it possible to provide enterprise management with a methodology for the initial phases of change, that allows the selection of those tasks which corroborate the belief of management in the feasibility of one or another type of change. Through this methodology several contradictions which surfaced in the business process re-engineering community disappear; e.g. the "to do as-is analysis frrst or not" question becomes a simple decision process based on management insight and available knowledge of the company and its current state. The VRIDGE demonstrator project has supported the belief that this methodology coul dbe utilised in a wide variety of cases. For example, the Purdue Guide for Master Planning (the methodology of the Purdue reference architecture) and the GRAI-GIM methodology (previously believed to be competitors) become special cases of the same change methodology, their differences being reduced to the fact that they live with different assumptions about the state of an enterprise that will attempt change.

GERAM, the Generalised Enterprise Reference Architecture and Methodology and the above change identificatiopn metholology have been developed and were applied to the VRIDGE Demonstrator, to identify and develop models of various enterprise entities in which substantial change was required.

Please read about the official release of GERAM or follow the GERAM tutorial.

Since GERAM is a generalisation of several integration architectures, this allowed the end user in the VRIDGE demonstrator to combine the elements and draw on the strengths of the PERA, CIMOSA and GRAI GIM architectures:

Understandably, various stages of the design process need the description of the same enterprise entity from multiple points of view, and on various levels of generality. For example a functional model (expressed e.g. in IDEF0) is suitable for the identification of the interfaces between (among) business functions. At the same time executable process models (expressed e.g. in CIMOSA/FirstStep) are useful for the investigation and optimisation of the business process to achieve optimal cost or time characteristics of the process.

II. Model Maintenence in Enterprise Engineering Tools

It has been recognised that Enterprise Engineering Tools with meta-modelling capability (such as METIS, ProcessWise, PTech) are needed as the basis of an Enterprise Engineering Workbench. The workbench could contain an integrated meta-model of those enterprise modelling languiages which are needed during the enterprise engineering / modelling process.

It has been obvious that there is a need for a tool that is capable of maintaining the consistency, and if possible translate among, various enterprise models. Although some form of translation is possible among models on the same life-ciycle level (e.g. two forms of process model), automatic translation between requirements and design models would amount to automatic desig, which in general is not possible. A study of semantic translation possibilities and of model consistency maintenance and fact propagation has also been undertaken.

We believe that the next generation of enterprise engineering tools will have the following characteristics:

III. Publications

A number of journal and conference papers have been published during the course of the project. Also the Handbook of Architectures of Information Systems was published co-edited by the TG4 WP1 leader collecting relevant enterprise modelling languages and reference models and adopting the GERAM architecture as an ordering principle. The second volume, the Handbook of Enterprise Integration is in preparation as of January 1999, and is going to collect a great deal of publishable GM21 results in one volume.


Please send feedback to eigroup@cit.gu.edu.au
Last updated 27 January 1999