Page 73 -

34 Enterprise Data Governance

product repositories, generating integration difficulties with
the rest of the Information System.

PLM is different in some ways to PIM. In particular, PLM
has important functions for collaborative work. These
functions enable concerned parties in R&D to work
collectively on successive versions of a product as it is being
designed. Furthermore, PLM integrates itself with
CAD/CAM systems through data interfaces on technical
nomenclatures and digitally-controlled machine tools. Once a
product is put forward by R&D, it is then copied into the
PIM repository to make it accessible to marketing systems,
and to manufacturing and sales. With PIM, product
configuration operations occur, but on the basis of pre-
established parameters at the time of product design by
R&D, i.e. on exiting PLM.

The close relations between PIM and other data
repositories (supplier, customer, factory, marketing, etc.)
make the use of MDM inevitable. Thus, MDM must replace
PIM. It is important to have a semantic data model for
product configuration in order to build an MDM system that
can be integrated with other data repositories, without any
10
unnecessary breakdowns .
With PLM, the choice of its replacement by an MDM
system is a more delicate affair. Indeed, an MDM system
needs interfaces enabling it to interact with subsystems used
by an R&D department, such as CAD/CAM, etc. This effort is
justified if a company notices that the integration of PLM
with the data repositories in production, be it a PIM or an
MDM system, is problematic. In an economic environment
where the design and manufacturing lifecycles are ever
increasing, a breakdown between a R&D and the other
marketing, manufacturing and sales teams could have a

10. This approach is also possible when using software packages (see
Chapter 3 later in this book).