Saturday, March 26, 2011

More on decision model and event processing

I have written before about the "Decision Model" book, when Larry Goldberg visited us in HRL. 
I met Larry again this week in the OMG meeting, and also had dinner with him and Michael Grohs from KPI  in Cary, NC 
(I am now staying nearby in Durham -- almost packed to go start the long journey back).

From their testimony it seems that the decision model is catching quite fast, and hundreds of organizations are using it already in one way or another.  In the OMG meeting they brought their flagship customer Freddie Mac representative, who reported a big success,  there were already several vendors implementing it to be executable.   The benefit is the simplicity and riding on the "spreadsheet" table-like thinking with their methodology.    They are now working on extending the model which started with partial coverage but is evolving,  e.g. they are adding concepts of views and contexts.

Our challenge in event processing is larger, since the complexity of operations is such that I am not sure can be easily expressed in tables, but the seek for abstractions that will enable business analysts and semi-technical users to construct systems is still there.     This is one of our activity areas, which I hope to report progress at some time.   More - later.

Friday, March 25, 2011

On the Virtual event processing symposium: the EPTS use case report

Yesterday we held the EPTS virtual symposium - a kind of live webinar that was broadcast from the OMG Technical meeting in Arlington, Virginia.   This was a complex logistic operation - some of the talks were live on the site, some were live on the phone, and some recorded.  Some obstacles, but it generally worked. There were 77 remote participants and few in the room.    This is behind me and I can breathe some air  - now on a one day visit in the IBM site in RTP, North Carolina.    The EPTS meeting consisted of presentation of the event processing manifesto, that I've described before,  some discussion on standards that I'll Blog about soon, and some report of EPTS working groups.    
The first of them that has been presented was the report of the use case working group, with a first public summary of the survey results.  Pedro Bizarro posted the presentation on slideshare.
While the results reflect those who answered the survey and is not necessarily an accurate sampling of the market situation there are several interesting observations:

  1. Event processing applications can be found now in multiple industries (not just financial), some of the applications were in energy and utilities, defense and aerospace, Transportation and logistics, Healthcare, manufacturing and more.
  2. The most cited reason for using EP is enhance/improve user services - and this probably relates to adding functionality that was difficult to add other way, others are - cost reduction, agility (faster development) and compliance
  3. Many of the applications use database/file as sources, which is more of the traditional way, other are using streaming sources like messages or subscriptions.
  4. Most applications are doing various types of notifications -- only a quarter are triggering automatic actions (this is somehow related to the decision oriented observations of James Taylor that I've blogged about recently).
  5. Most applications do not require high throughput ( less than 1000 events per second), around 5% of the application need more than 100K events per second -- this is an observation made before by Roy Schulte.
  6. Most applications have 100 event producers or less (actually almost half have 10 event producers or less)
  7. Among the non-functional properties: high availability was the biggest requirement, while security is still not a big concern 

As written - this relates just to a sample of applications and is not necessarily represent the entire market, but I think that these observations show some of the current trends.  I'll write about other sessions in the webinar later.

Thursday, March 24, 2011

Decisions in smarter systems

Arlington., Virginia,  Hayatt Hotel

I am here for the OMG technical meeting.      I have participated in (part of) the decision modeling day organized by Paul Vincent and Christian De Sainte Marie.  Their ultimate goal is to get to a standard on decision modeling, and they have issued proposal for RFP on that issue.  A good survey of that day can be found in James Taylor's Blog.  I sat near James, and he is blogging in real-time.    James himself gave an interesting keynote on the 
importance of decisions  

James concentrated on operational decisions and said in many of the organizations the role of computerized systems is to provide data (in various ways) to manual decision makers when they ask for it.    The smarter systems have larger portion of automated decisions, they are active rather than passive - determine when a decision is needed, and the decisions are measurable with quantitative metrics, so they can be evaluated.  

While James did not talk explicitly about event processing,  it is obvious that it has a significant role in his vision, it has several roles:

  1.  determine when a decision is needed
  2.  the automated decision itself is often context dependent and  the context can be determined by event processing context mechanism (temporal, spatial, event history related...)
  3. the decision itself may depend on event pattern
  4. Last but not least -- the extension of event processing to proactive computing coupling with the metric that measures the decision's result can trigger decision to mitigate undesired predicted deviation from the result,(I discussed this one with James during the reception in the evening).  

The EPTS virtual symposium - tomorrow.   A lot of logistics to get it running!