Dynamic Processes

I got a BA with a major in biology, and every so often, to try to at least keep aware of how that field is developing, I get a book that let's me stay up-to-date with one or another aspect of biology.  With this in mind, I recently acquired A First Course in Systems Biology by Eberhard Voit.  It's an interesting survey, and you can get some feel for its contents just by reading the titles of the first few chapters:

1. Biological Systems, 2. Introduction to Mathematical Modeling,  3. Static Network Models, 4. The Mathematics of Biological Systems, 5. Parameter Estimation….

For those with no particular knowledge of biology, you'd be surprised at how familiar much of this seems to someone who is active in process work.  It's about modeling biological processes, with a mathematical twist.

Those interested in business processes, who are mathematically inclined, have probably already found their way to John Sterman's great book:  Business Dynamics:  Systems Thinking and Modeling for a Complex World.  Sterman's book describes his MIT course where he teaches business people to apply Systems Dynamics Theory to business processes.  Without saying too much about Systems Dynamics Theory, let's just say that it begins with a diagram that shows the relationships between a number key variables.  Thus, a diagram of a firm has a box for the firm, and within it, sub-boxes for Sales, Order Fulfillment, and Capacity Acquisition.  Outside the firm is an external box for Customer Demand.   Arrows connect the various boxes, indicating things like Revenue, Customer Contacts, Orders, etc.   In other words, we reverse a traditional BPMN diagram and represent stocks of things as boxes and processes as flow arrows.    More important, however, one doesn't ultimately focus on the model and boxes but on a set of graph lines represented on a matrix.  One graph line will show orders growing or declining.  Another will show the fluctuation in customer contact or in revenue.  The real value of the initial model is simply to establish relationships, while the real insight to be gained is the dynamic interaction of the variables over the course of time.  (If you want to read more on systems dynamics, go to www.bptrends.com and search on “systems dynamics.”

Ultimately, this is where dynamic process analysis will go.  We will move away from models that simply show activities and relationships.  Increasingly flows won't be as important as the networks of relationships that make up a business system.  And, ultimately, management decisions will be based on the interaction of the network elements over the course of time.

At the moment, an OMG group is working on a standard for Case Management.  Case Management is meant to describe complex processes that change frequently.  Unfortunately, as far as we can tell, at the moment, the standards group is mostly focused on using business rules to incorporate decisions within what are, ultimately static process models.  Incorporating rules is certainly a worth undertaking, and perhaps a necessary step toward a more sophisticated model.  Most process people really aren't ready for dynamic process modeling.

I can remember all the books I have read on biology that pictured relationships — food webs that showed who ate who else, or diagrams on energy circulation in an ecological system.  Those diagrams were a first step — they developed our basic understanding of the relationships involved.  Systems Biology is a second step, where diagrams are added to real time data and used to generate dynamic pictures of changing relationships.  If one wanted to exercise control over such a dynamic systems, one can imagine establishing rules that defined what was likely to follow from certain patterns.  If x is rising, and y has just dropped while z remains steady, then m will likely happen, unless one intervenes to slow the rise of x.

I'm not sure business managers or process practitioners are ready to think of their modeling efforts in these terms yet, but in the near future I am confident that many companies will have people who can think in these terms.  Between now and then, lots of process people are going to have to relearn the basics of their craft and add a facility from some mathematics. New BPMS tools will undoubtedly help with this transition.  In the long run, however, we will shift from drawing modeling of processes that are often ignored to creating tools that will let business managers manage their companies on a day by day basis.