I picked up a copy of the Align Journal at the BPMI Think Tank this week, and found inside an interesting article on Con-way Freight’s [*1] use of SOA and EDA, together with CEP [*2], to improve real-time business performance.
One interesting aspect was that one of the business drivers for adopting EDA and CEP was “automating business experts” [*3]:
Many business activities, when observed, seem too complex to be automated. A business expert fueled by experience and skill makes “instinctual” business decisions and may find it difficult to explain the web of data points, rules and constraints that went into a choice of action. That knowledge must be teased out into its component parts:
1. Events
2. Triggers
3. Relationship of events to other current and historical
events
4. State transition model with timing expectations
5. Business rules or optimization algorithms to apply to
specific patterns of events and business object states.
The application architecture seemed to have a pretty good ROI: the part of the application dealing with Canadian and US Customs alone saves 500 hours of paperwork a day. Overall this is an excellent article for business and IT people that helps explain the value of CEP.
Notes
[*1] Con-way Freight is a subsidiary of Con-way Inc, a $4.2 Bn pa freight and logistics company.
[*2] Con-way Freight uses TIBCO SOA software for this application, including TIBCO BusinessEvents for Complex Event Processing.
[*3] Veterans of the rules business may recognize the similarities with the “expert systems” and “knowledge based systems” (KBS) that proved trendy 15-20 years ago. Of course, these systems have continued to be developed (as “rule-based systems”) even if the terminology fell into disuse. Rule-driven CEP technology provides many of the capabilities required for real-time versions of such systems.
