Does BPEL – or even BPM – matter? (part 2)
BPM as a principal concept fails to deliver realistic achievements. There are no vendor independent studies that prove that rigidly managed process flows are beneficial to a business. Executives chase the illusion that they can implement a business model with IT that will allow them to run the business by remote control.
For decision making and monitoring the business achievements against goals, it is necessary to have access to real-time business data. BPM people speak of using ‘a common semantic set’ when they mean process descriptions and metadata definitions that correlate to the service interface data. Therefore not only BPEL has to be compatible but also business architecture metadata. BPEL offers nothing for both needs at all. Presenting real-time business data to the business user in the process requires Java code to pull the data from the service interface and present them.
The other element that is needed for pragmatic use of process management is business rules. One could convert business rules into BPEL, but then they cannot be executed as complex rule sets that are triggered by data changes or business events. Business rules also act on business data so you need to code more Java to pull them from the service interface and pass them to the rule engine. Business rules have however multiple purposes with the most important one being so called boundary rules that are essential for auditing. The better your boundary rule set is, the better you will catch exceptions and violations automatically. Thus those rules should not really be an integral part of your process but rather an independent definition but tightly integrated into the process execution. There are some non-BPEL product that handle real-time data and rules quite well.
Some vendors propose BPM 2.0 and other visionaries already propose BPM 3.0. They say that it will not happen without BPEL. I think they are right. Therefore there will be no BPM 2.0. Or even 3.0. BPM in its current rigid form with or without BPEL and BPMN will not make businesses more dynamic in the long run. New technology concepts are required.
What would this new technology need to look like? If any technology concept can foster successful process management, it has to be business facing and enable real-time metadata-driven model execution, while engaging business directly in continuous process change cycles. To enable business participation, a secure change management environment is required that controls the lifecycle and manages deployment.
My proprosal: The only way that business users can be involved in creating processes is by using graphical means of real-world process representation. The best way to plan and represent a process is by defining the user roles involved, the data entities required and how they are serviced by backend applications, business rules, user presentation and content definitions. Content state drives the process forward. The problem of complex summary states of content will need innovative functionality. Consider a software agent who will monitor user activity in a business process (case) and automatically discover the data and state patterns that repeatedly cause user activity.