are related and/or integrated. In general two
approaches exist here:
– Direct model integration: Direct relations
between concepts in heterogeneous languages
are established. An ontology may be used to
define what concepts semantically are the same,
and which relations may exist between concepts.
A drawback of direct model integration is that
each language that must be added requires O(n)
relations to the n other languages.
– Integrated metamodels: Based on the
metamodels of heterogeneous languages an
integrated metamodel is created synthesised with
mappings to and from the domain-specific
languages. Creasy and Ellis (1993) proposed to
use conceptual graphs of Sowa as language for
the specification of the integrated metamodels.
This is similar to the ArchiMate approach
(Jonkers et al., 2003) in which a rich metamodel
is developed for enterprise architecture models
with mappings to/from domain-specific
modelling languages like UML (OMG, 2003b)
and BPML (BPMI, 2003).
Grundy and Venable (1995) present an
integration environment in which different
modelling languages have their own repository and
editor while changes in one model are propagated
through a central repository based on an integrated
data model to the other models.
Karsai (2000) describes an integration
framework based on model integration. The
architecture proposed is based on tool wrappers that
translate tool-specific models to a syntactic
modelling language resembling ER, after which a
semantic interpreter interprets these models and
stores them in a central repository. The framework
assumes that the presentation layer of tools is left
untouched.
The WOTIF (WOTIF) project aims at
developing an open framework for integrating
design tools for embedded system development.
Design flows today are realized using different,
proprietary design tools, whose integration is a
complex problem. WOTIF provides a meta-model
driven infrastructure for design tool integration,
which facilitates the semantic interoperability across
the elements of a tool chain. WOTIF is implemented
on the basis of Eclipse, an open extensible Integrated
Development Environment (IDE).
5 CONCLUSION
In this paper we have presented the design of a tool
integration workbench that is able to integrate
existing modelling tools. It shows that it is possible
to practice enterprise architecture while at the same
time keeping existing modelling artefacts.
Leaving existing modelling environments intact,
the workbench allows the concurrent design of
enterprise architecture domains: each domain may
still be designed using its own languages, tools and
techniques. More importantly, with the ability to
reason across domain boundaries the workbench
introduces an instrument for collaborative design.
By adopting the ArchiMate modelling language,
the workbench not only allows the integration of
existing modelling languages, but provides a
language to communicate across domain boundaries
as well. Moreover, the workbench serves as a
starting point for the analysis of enterprise
architectures using generic analysis techniques that
rely on the ArchiMate modelling language.
An important success factor left unaddressed in
this work is the mechanism responsible for the
synchronisation of models that share objects. This
subject requires further investigation.
Another key factor in the success of the
workbench architecture is the feasibility of
transformations between tool-specific content and
ArchiMate content. The semantic soundness of such
transformations is particularly nontrivial and thus
requires further exploration.
One of the goals of the ArchiMate project is to
stimulate innovation in the market of tools for the
enterprise architecture design process. The ideas
presented should challenge vendors of such tools to
(1) provide interoperability services, e.g. tool-
specific adapters, such that their tool can be
integrated in environments like the one presented
here, (2) create commercial versions of tool
integration environments and (3) create graphical
modelling tools that allow relating and integrating
existing models.
In the near future, ArchiMate will create
prototype versions of the environment in order to
show that the workbench approach is feasible and
can be turned into commercial products. Attention
will be paid to transformations, synchronisation of
models that share objects, and the analysis of
enterprise architectures. Furthermore, the workbench
will be validated in pilot projects running at the
companies involved in ArchiMate.
ACKNOWLEDGEMENT
This paper results from the ArchiMate project
(
http://archimate.telin.nl/), a research initiative that
aims to provide concepts and techniques to support
enterprise architects in the visualisation,
communication and analysis of integrated
A TOOL INTEGRATION WORKBENCH FOR ENTERPRISE ARCHITECTURE - Integrating heterogeneous models and
tools
477