Strategic Workshop on Enterprise Integration and Enterprise Computing
(Systems and Software Engineering Standards in the Context of Enterprise Integration)

19-20 November 1998, Sanctuary Cove
(Brisbane's Gold Coast, Queensland Australia)

Position Statement (David Shorter, 6th November 1998)

This statement is in three parts, the first being a general concern about how to get enterprise modelling recognised and used, the second a statement from a European workshop, and the third a personal response from me as CEN TC310 WG1 convenor to Jim Nell’s ‘dialogue’ questions.

1. General

Problems for standards-makers concerned with Enterprise Modelling (EM) in support of Enterprise Integration are that:

• the industrial take-up of EMs as an accepted means of integration and of capturing, applying and re-using enterprise knowledge (‘knowing how to do things’) is slow and often associated with large research consortia,

• what we produce is not immediately usable by industry in its current form, and so

• getting industrial participation in (or even recognition of) our standards-making activities is very difficult (our conferences and workshops often preach to the already converted), and

• even where EMs are developed, they do not (to my knowledge) become a normal part of the on-going operation of the business.

Proposal:

Focus on the business drivers that senior management in industry will respond to, and explain and demonstrate how EM technology will help. We tend to be mired in discussions of technical issues to be resolved and the impact of ‘new’ technology. Let’s put that on one side, document the achievable business benefits, agree what EM capabilities vendors need to be able to deliver to achieve those, who the stake-holders will be and only then move to define what technical problems still need to be solved and where standards should be part of that solution. Then promote these messages to management.

2 CEN STAR workshop recommendations

The following is the final report of a CEN STAR working group (A) produced during a two day meeting in Senlis in June 1998. Members of the working group were: K. Kosanke, CIMOSA Association, Germany, J-J Michel, CETIM, France, M. Petit, University of Namur, Belgium, D. Shorter, IT Focus, UK, R. Weston, Loughborough University, UK. (A previously duplicated paragraph in 2.1 has been moved to 2.3.)

2. Technology Trends

The move in industry is towards autonomous enterprise units (extended and virtual enterprises with agile behaviour) with a significant need for decision support based on real time information. Such information can be identified by and represented in enterprise models and with sufficient ICT support the models can be used to evaluate strategic, tactical and operational alternatives through simulation of business process/enterprise models. Because these models represent both business and technology components, they can be used to maintain the coupling between the business requirements and the technological solutions.

Enterprise models have to coexist and interwork with product models and new models of electronic interworking (EDI, CALS). At present these different kinds of models have different semantic underpinnings (the same word means different things) and it is difficult to relate concepts precisely (say in a form allowing computer support for visualisation of relationships and even assistance with translation). This also makes it difficult to present and assess which standard is addressing what, how they are intended to interwork and what standards should be specified to meet a particular business need at a particular time.

Therefore representation and handling of enterprise models have to be adapted to the needs and capabilities of business people to allow for modifications and their evaluation and verification to be carried by those people themselves.

2.2 Serious Drawbacks

However, we are in a real vicious circle: proactive standardisation has per definition a long pay back period and is a high risk investments, but with a high ROI - if successful. Therefore industry will try to avoid this and will invest in reactive standardisation type activities with low risk and relatively low ROI. The result is a loss in use of leading edge technology, due to not being aware of capabilities and therefore not being able to use more efficient and effective methods, tools and technologies in the enterprise operation.

On the other hand we are supporting technology development with rather large amounts of tax payers money, but we forget about the needs for technology transfer to the user. With the complexity of current technologies the demonstration of applicability and benefits of new solutions is a prerequisite for awareness and acceptance in the user community. However, it is not sufficient to convince a few people in an organisation, but to create a critical mass in the enterprise which is willing and prepared to invest in the new technology.

2.3 Approach for Improvements

We think that new ways and methods have to be explored to create the critical mass of knowledgeable people in the enterprises and thereby achieve the technology transfer on a much broader scale and with an significantly increased speed. At the same time the barriers to take-up need to be lowered by developing and releasing model fragments which are easily accessible at low cost, so allowing industry to make incremental investments within a holistic framework. This will benefit both our industry, which will be again on the leading edge of technology, and the tax payer, who will benefit from his investment in a much shorter time frame with an improved standard of living.

Such an awareness campaign should focus on a selected set of regional enterprises doing some pilot work for demonstrating applicability and potential benefits. These enterprises would then act as show cases for the community.

2.4 Present Requirements

Enhance current standards or pre-standards and ongoing work:

• Revisit life-cycle dimension and contents of genericity dimension in ENV 40 003 (constructs/building blocks, building block types, partial models/components, particular models). (Standards activity).

• Extend modelling construct defined in ENV 12 204 for representation of human related issues, business requirements and virtual systems. (Research activity)

• Further unification and extensions of integrating infrastructure technologies. (Research activity)

New work items:

• Collecting existing semantic models and abstracting from these on a common basis (UML?). The intentions are to increase the base for the unified basic semantics for enterprise engineering (next item). (Research activity).

• Unified basic semantics for enterprise engineering – enterprise modelling and integration. This basic semantics will allow the derivation of user oriented modelling dialects (administration, product development, manufacturing engineering, ...) which map onto the unified basic set. (Research activity).

• Representation of modelling constructs for human use and for computer processing. (Research activity).

2.5 Existing Deficiencies

Overlapping and conflicting standards

• Clarify the use of information model in the different areas of standardisation. (Standards activity).

• Mapping enterprise engineering related standards onto other standards (see STEP/EDIFACT work). (Development activity).

Limited industry involvement in standardisation and use of standards. Need to improve transfer to industry and increase involvement:

• Guide for industry for using standards. (Development and standards activity).

• Establish showcases for partial models via supported industrial pilot applications. (Development activity).

• Establish modelling community clubs. (Development activity).

2.6 Future Requirements

• Registration mechanism for partial models/components (e.g. for a web-site repository) to prepare for industrial use. (Research activity).

• Collecting partial models/components and applying registration in a web-site repository. This is intended to demonstrate the capabilities of enterprise modelling and the improved effectiveness through the re-use of models. (Research activity).

 

C CEN TC310 WG1 convenor’s response to Jim Nell’s ‘dialogue questions’

David Shorter, david@itfocus.demon.co.uk, 6/11/98