IFIP-IFAC Task Force on Enterprise Integration
Workshop on Application Experience
Detroit 3 November, 1997
Contents
Letter to Members
Minutes
Ford Case Study
Chrysler Case Study
IMS Globeman 21 Virtual Enterprise Engineering
European Automotive Industry Consortium
International Conference on Enterprise Integration and Modelling
Task Force on Enterprise Integration - Future Direction
See also programme of workshop
To Members of the IFIP-IFAC Task Force on Enterprise Integration
28 January 1998
Dear Members
I apologise for the lateness of this report about the Detroit meeting of the Task Force, now that the Summer holidays are over (the summer is not, we still had 36C in Brisbane in the last few days) I hasten to send out the minutes summarising the meeting of the IFIP-IFAC Task Force on Enterprise Integration which was held in conjunction with Autofact in Detroit on the 3 November 1997.
I would like to thank to our host Dr Gustav Olling, who is now CAE/CAD/CAM R&D Executive of Chrysler Corporation for the arrangements of the meeting at Cobo Center in Detroit and the assistance of the SME Secretariat, especially Ms Shelley Stover and Ms Deirdre Corbett for the arrangements.
On the 4 November IFIP TC5 held its strategic meeting and the TC5 chair proposed a name change to simplify our name to IFIP IFAC Task Force on Enterprise Integration. The corresponding IFIP WG5.12 and IFAC TC MIA names would also change as a result; from IFIP's perspective the name change has been approved by TC5.
The Task Force mailing list has seen in January a discussion on the directions that we should take, and I agree with those members who expressed their opinion about the need to address strongly the human issues in enterprise integration. Since these questions are treated by other relevant groups in both IFIP and IFAC I propose that I as chair approach them (please provide names of groups and contact persons who You think would be the most appropriate) and propose to them that joint activities should be planned for 1999 and the year 2000.
As you will see the meeting also identified other important issues that the Task Force should not abandon in the area of education, research, case studies and standardisation.
best regards and thank you very much
Peter Bernus, Chair
Peter Bernus, Griffith U., Australia (task force chair)
Guy Doumeingts, Bordeux U., France
Jim Nevins, consultant, Burlington. MA, US
Laszlo Nemes, CSIRO, Australia
Francois Vernadat, Metz U, France (vice-chair IFAC TC MIA)
Ted Williams, Purdue U, US (task force vice-chair)
Gyorgy Kovacs, Computer and Automation Institute, Hungarian Avad. of Sciences, Humgary
Guests
Gus Olling, Chrysler Corporation
Mike Morton, Chrysler Corporation
Gene Nelson, Ford Motor Company
Gale Copple, Ford Motor Company
Note on the materials distributed together with these minutes:
Please note that for confidentiality reasons the slides presented by the industry speakers are not available for distribution (at the time of the workshop); I hope that the summary presented below is a suitable way of conveying information that was relevant for our purpose.
Also Professor Ted Williams asked that the attached paper by Shelby V.Laurents of Flour Daniels not to be circulated before January 98. By the time you receive these minutes the request will have been honoured.
Materials that were not available electronically have been sent to Task Force memebers by air mail. Some video recordings were made during the meeting but due to technical reasons the quality was not good enough for them to be circulated.
Chair's Report
The aim of the meeting was to discuss,
For this reason, with the generous help of our founding members Jim Nevins and Gus Olling we have been able to secure two talks about significant projects in the automobile industry which fall in the category of Enterprise Integration.
The Ford speaker gave a 1.5 hour presentation on how Ford transformed its production, dramatically dropping production costs through the application of product and production process sharing and reuse between various Ford products.
Unfortunately the talks were given to the Task Force members present and the detailed slides and video could not be made available for dissemination.
Brief summary of the talk given by Gene Nelson and Gale Copple, Ford Motor Company
"Ford had a major goal: to dramatically drop its production costs without any compromise (or with improvement) of quality an thereby increase its investment efficiency. Since normally in the car industry complete factories are built to support the manufacture of a line of product, or a model, automobile manufacturers end up with slight variations of the same (type) of process, without the benefit of reuse of what has been developed in one model for the production of another.
Even though sharing of parts between models has been a common practice in the global car industry, the sharing of technology (processes and tooling for example) has not been. This is a major cost factor and increases the fixed cost of production immensly."
The brief analysis of the approach is given below.
Ford’s objective seems unattainable without a global view and approach to change, because no local change can yield the desired effect. In order to change the practice that causes the waste a complete overhaul of all products and productions facilities is necessary from this respect.
Ford needed a project that was able to
a) identify the opportunities in the present and future planned product mix and the respective processes
b) develop a metrics which allowed the project to measure the impact of the process reuse on the total life-cycle cost of the vehicles produced
Obvioulsly the project did not only need technical competency to identify the process reuse opportunities, but a great deal of management skill to convince designers to make the necessary changes that followed from the need for process reuse. We believe that the development of a metrics is an important component in being able to quantify the benefits that flow from the desired modifications.
How does the Ford project (Ford 2000) relate to enterprise integration, what do we believe Ford could benefit from EI methods and what results could be useful for similar future projects?
Since the change at Ford was of the nature which took the company from one level of integration to a higher level, it could not be treated as an isolated technology improvement exercise. The result of integration is an approach that assesses the total life-cycle of the products and feeds the findings of this assessment back to the product and production facility development process.
In the particular case presented the assessment was concentrating on reuse and therefore reduction of fixed cost. However, such optimising behaviour is an important property in the company for any future change, including the technical and management channels that were built out to make this change possible. E.g. should the result of metrics calculation change dramatically we expect that the company will have learned in the present project how, based on the results of evaluations at any given time the necessary product or process changes can be instituted. In other words we believe that apart from the impressive economic savings on production costs quoted at this time, another less visible but important result is a new capability of the company to optimise itself based on a metric (or performance criterion). [Note Bernus July 2000: the Ford Case study has now been released (April 1999)]
Brief summary of the presentation made by Mike Morton and Gus Olling of Chrysler:
The Chrysler speaker presented another approach that uses extensive simulation of both product and production to optimise production costs and improve product quality
"Chrysler recognised that for the integrated management of Chrysler products a number of information management requirements must be satisfied, such as maintaining a product database, manage manufacturing data, maintain an engineering bill of material, implement comprehensive change notification, vehicle specification support, enterprise resource planning, and all this integrated in a virtual product and process modelling system.
This is an overall approach to the information flow in the product design and production design and will improve the quality of the product itself and, or reduce the cost of the process thereby effecting the profitability of the company. The aim is a complete product data flow integration including design processes and process equipment and to close the loop of the product and plant design by simulation.
While product design is a highly automated process (i.e. is supported by sophisticated CAD tools) plant design is not as developed.
Virtual product AND process modelling will close the gap between the two life-cycles."
The recognition that products have a life-cycle which needs to be understood and managed is not a new one, neither is process simulation (which is for example often applied to optimise factory layout). However, these approaches traditionally propose local optima and solutions for individual details of products and processes. We believe that these methods will have to be utilised in the future but in conjunction with global life-cycle modelling tools and methods such that product and production are simultaneously engineered.
In the past 10 years it has become more and more clear in many industries (electronic, software, automotive, agricultural production, services) that almost any improvement of the product is intimately connected with the quality of the process that produces it.
Industry sector specific process assessment methods have been developed, with standards (like ISO9000 series) and more recently a great variety of STEP standards is developed to allow the integrated information flow of information not only in product design but also in process and plant design.
Chrysler is moving from the Part centric to the Product and Process centric organization.
The simultaneous design of product and process is expected to result big savings and quality benefits through global optimisation.
In addition to the Industry talks presentations were given by Task Force members on the activity of research consortia, namely the Globeman 21 IMS consortium (Global manufacturing), by Laszlo Nemes, the European AIT consortium, by Francois Vernadat, and the use of EI, especially PERA in Flour Daniels, by Ted Williams.
See slides and relevant articles attached on the AIT Consortium, Globeman consortium, and the Use of PERA by Flour Daniel (in "Building Electronic Data Management...")
Guy Doumeingts gave an overview of the new use of the GRAI GIM methodology, which has been extended with several new components to include benchmarking, development of performance criteria and successfully used in the service industry (finance / credit company).
Given the fact that the week preceding the Task Force meeting the International Conference on Enterprise Integration and Modelling was held in Torino, Italy Francois Vernadat also gave an overview of the outcomes of ICEIMT97.
The view of the future (conclusion)
The discussion continued with the participation of all members present (except the colleagues from Ford, who had to leave earlier).
The questions put to the members of the discussion were:
In other words we wished to identify those needs of industry where significant benefit could be expected from the work through the IFIP-IFAC Task Force on Enterprise Integration.
In other words our aim was to develop a strategy to follow to achieve the above objectives.
Disseminate GERAM associated results in form of applying these to particular typical cases to show the connection for end users. Through the publication of such cases, case studies and other forms of results the generic findings will be easier to understand by the end user community
Publications, such as the planned Handbook of EI (Springer) and the participation in Education programmes (e.g. through SME) both for management and engineering professionals
Case Studies (demonstration of benefits)
Case studies could for example demonstrate a particular industrial case (real or imagined but realistic) in terms of the GERA modelling framework and methodologies, and identify various enterprise entities and their life cycles; demonstrate what is covered and what is not; what tools are available for the uncovered areas
.
As GERAM is a very broad scope architecture, its components can not be considered in isolation. It should be demonstrated what is the relationship between GERAM and related architectures such as PERA, GRAI-Gim, CIMOSA as well as other standards (STEP, ODP,WPDL,…) or proposals in all the areas in the scope of GERAM (such as enterprise modelling languages, reference models, methodologies, enterprise modules such as integration infrastructures) in various industries.
To organise workshops, conferences in conjunction with various interest groups about the basic questions identified in the ICEIMT process. A few of these issues are listed here (please refer to the attached report on the ICEIMT conference and workshops):
It is important that the Task Force does not do an isolated attempt to solve the above tasks, because of the multidisciplinary nature of the problems
As an immediate task the Task Force is participating in the ISO TC184/SC5/WG1 work and we shall attempt to harmonise our activities with the strategic directions that SC5 is going to propose at its meeting in Paris.
After the Boulder Colorado meeting we shall then decide a future work programme, including meetings and other activities as described above under Education, Case Studies, Research and Standardisation.