End-to-end OOTB vs. open, standards-based PLM

I liked Oleg's recent post PLM Platform Wars: Who is Right or Who is Left? .

As commented several times, I agree that product data is the main asset of a company as PLM applications come and go over time. But I would like dig into another trade-off that PLM customers are facing when trying to act on this: the choice between cost-efficient, end-to-end solutions deployed “out of the box” (OOTB) from one vendor versus a multi-vendor PLM architecture based on open standards.

Product data is the main asset

Protecting this asset e.g. by knowing your logical and physical data model and by constantly monitoring the data quality is essential. An open, standards-based solution in this field basically means STEP for most of the meta-data model, potentially complemented by JT for geometry and newer incarnations of STEP such as PLM Services. Most PLM platform vendors have not adopted STEP as a major component of their product architecture – for reasons such as maturity of the standard and the need to differentiate the offering from the competition. You might find STEP interfaces for import and export, but as long as the internal core data (and services) models of the PLM applications are not STEP-based, the interfaces are rather weak links and create a lot of overhead for data mapping and data exchange. Integrating complex PLM architectures with hundreds of applications on this basis is still a bold venture that not many companies go into.

End-to-end PLM solutions OOTB

Sounds too good to be true, right? And I think it is. Even if one vendor had a good offering, the risk of committing your product data into a black box governed by one vendor is not very promising. In the current case of Dassault V6, I would distinguish two cases:

• MCAD-PDM: CATIA V6 with integrated Enovia team data manager
• Full PLM: the complete Dassault V6 vision of mechatronics PLM (RFLP)

I tend to think that the benefits of an end-to-end CAD-PDM solution could outweigh the risk (if you can still control your product data...), but I would think twice for my full PLM architecture.

What do you think?

Note: this view is driven by working with rather large, OEM-type companies in automotive and aerospace. Small and medium businesses might have different priorities.