This time I would like to receive some feedback from my readers as I believe the topic I am discussing here might be similar to a PLM / ERP discussion – a discussion between religions. I have preached the past two years a more data-centric approach for PLM, instead of file management and related tot this data-centric approach, the concept of a PLM platform / Business Platform – CIMdata/ Innovation Platform – Gartner becomes clear.
What´s the issue?
As I wrote in my earlier post (random PLM future thoughts), I realized that talking about platforms is not that straight-forward when meeting companies with their history and terminology. Some claim they are already using a business platform, others have no clue what makes a platform different from a their current PLM implementation ? Therefore I will summarize the different approaches I have seen in my network and give a non-academic opinion as a base for discussion. Looking forward to your opinion.
The platform approach
My definition of a PLM platform:
- A central repository of data based on a core data model. Information is stored as data in a unique way
- On top of this repository, applications can run, using a subset of the overall data elements, proving dedicated functionality and user interface to a particular user / role
- Access to the platform is provided through web-technology. Storage could be on the cloud.
- External applications and data can be connected through an open (standardized?) API embedded or federated
- The PLM platform can be a collection of services and functionality coming from various vendors / suppliers – the app store concept
- The platform approach is THE DREAM for business, being flexible to combine and edit data in any desired context in dedicated apps / environments
In the PLM world, Dassault Systems with their 3DExperience approach is following this trend although here you might argue about the ease of use to add external apps to this platform – is it open ? Aras and Autodesk might also claim they have a PLM platform, where you might question the same and if the depth of the data model and the provided solutions on top of the data model are mature enough. Finally also SAP can be considered as a platform, but I would not name it a PLM platform at this moment in time. An important question for me would be: How can achieve openness of a PLM platform?
The PLM backbone approach
My definition of a PLM backbone:
- The core PLM functionality is provided by a single, proprietary PLM system
- Additional functionality that is not part of the core development (acquisitions) is connected to the backbone through proprietary interfaces
- External authoring tools are linked to the backbone through integrations or interfaces which could be developed by third parties
- External system can interface to the PLM backbone through open interfaces
- The PLM backbone is THE DREAM for engineering, as historically this was the domain where PLM started to be implemented
I would consider Siemens and PTC (see picture) the best examples of a PLM backbone approach with their PLM portfolio. Teamcenter and Windchill are both rich PLM systems further connected to several systems, covering the product lifecycle. I am not expert enough to state that the same conclusion is valid for Oracle´s Agile, where I believe the backbone is bigger than the PLM system. What do you think ? Will these PLM vendors also move to a platform approach? And what will be the platform?
The Service Bus approach
My understanding of the Service Bus (I am not an IT-expert):
- Service Bus has a standardized interface to request for data or to post data that needs to be stored in other systems
- The Service Bus approach reduces the amount of (custom) interfaces between systems by requiring standardized inputs and outputs per system
- Providing a user with information that is not entirely available in a single system, the service bus needs to acquire the data from other systems, which might not give a high-performance as expected by business people
- The Service Bus is the IT DREAM as it simplifies the complexity for IT to manage point-to-point solutions between systems and makes an upgrade strategy easier to support.
From a very high-level view, the service bus approach has some similarities to a platform. The service bus concept allows business to select the systems they like the most (provided they connect to the service bus) – Image property of IBM.com
The main difference would be the persistence of information, where is the real data stored? I came across the service bus approach more often in the past, where the target was most of the time to integrate the PDM functionality (PLM as an enterprise solution was never in scope here).
For the Service Bus approach, I am curious to learn its relevance for future PLM implementations as the challenge would be to provide any user in the company with the relevant information in context. Is the service bus going to be replaced by the platform? Who would be the major players here?
The Business Intelligence approach
This method I discovered in project-centric companies (Oil & Gas companies, EPCs, Construction companies) but strangely enough also at some manufacturing companies, where I would assume integration of systems would bring large benefits.
- Each type of information is managed only in one single system avoiding interfaces or duplication of data.
- Only where needed, data will be pushed from one system to other systems
- Business Intelligence applications extract information from the relevant system and present this in context to the user, giving him/her a better of understanding
- Business users will work have to work in multiple systems to complete their tasks
- The BI approach is the ULTIMATE IT DREAM as it simplifies their works dramatically and shuts down business demands.
I have seen an example where IT dictated that for document management we use product ABC (well-known Content Management system). Next for internal documents we use SharePoint. For CAD, we use product PQR as much as possible (heavily adapted) or AutoCAD 2D (to support the minimum). For ERP, the standard system is XYZ (a famous ERP system – you do not lose your job by selecting them) and of course everyone uses Excel as a common interface of information between people.
It was impossible in this company to have a business view on the solution landscape. As you can imagine, this company’s margins are not (yet) under pressure as their industry is very conservative.
What do you think?
Is the future for PLM in platforms? If Yes, what about openness? Who are the candidates to offer such a platform? Or will lack of industry standards and openness block wider adoption? If No, will there be a massive PLM system in the future, connected to other enterprise systems (ERP/CRM)? Or will PLM be implemented as a collection of smaller systems communicating through an enterprise service bus?
I am looking forward discussing the topic here and soon during the upcoming Product Innovation conference in Düsseldorf