I participated at an OMG meeting two weeks ago in Berlin that during a side conversation I was talking with some peers about the new CMMN designed to model and execute non prescriptive, standard, “bpm”, whatever you want to call process types.
I skimmed across the beta release and I did not found nothing extraordinary that BPMN could not do to model Case Management approach, by the looks and feel, I would say that CMMN is a subset of BPMN.
People from the CMMN committee told that the difference is how the language is executed, once is based on stage transition.
For me it was a surprise, because stage transition is what “BPM” processes (structured) are all about., the process moves, or change stage when activities reach and end state. ACM, or even Case Management is not about stages, is about availability of data, is user and data driven processes, as such is much more object oriented that defines the path the process takes.
Thus is CMMN missing the target and actually is a BPMN subset with a different name, or is something different?
My argument is X Management (Case management, Adaptive Case Management, Purpose Case Management, Production Case Management … ) is object oriented, not task oriented. Actually we can handle a case with no tasks at all, it’s possible to combine multiple approaches to do it: activity streams, documents, etc. The difference is that is data and data availability that is transformed into information that drives the case. Most part of the data is coming from sources outside the form of the case (if it exists).
For example, if you are analyzing a complain and you get the contract to understand the type of conditions that were setup, the penal clauses or if an opinion coming from Legal Dpt about how the complain should be handled regarding the context and contract that was signed, this is what helps people involved in the case to steer direction.
The big difference as I see is that X Management is object aware approach. The overall process (case) is structured around object types involved and outcomes of its manipulation (goals, tasks, documents, attributes, etc) and may refer to other object types or be referenced by them.
It’s not my intention to start a “holy” war against CMMN, I just say that looks like it’s stuck in the middle of the bridge. It does not means that in future it can evolve and can cross the bridge.
I’m arguing that as a design principle, the language is based on stage transition, like prescribed / structured process does not look right to me, because X Management is all about processing of data that helps to decide the path.