|
Friday, November 27, 2009
Thoughts on an Enterprise Architecture Research Program
The questions below are taken from discussions at the Penn State IST Enterprise Architecture Forum. The answers are mine and mine alone.
What questions should be asked as part of an Enterprise Architecture
research program? - What is the deep structure of EA (i.e. what do all EA frameworks/approaches have in common)? What is the ROV (Value = Investment
or Mission or Quality) of EA (an ROV model should be evident on the basis of the deep structure)?
- What
is the Venn Diagram/Euler Diagram of all EA frameworks/approaches where the area within Venn/Euler closed-curve for each framework/approach are the tools and
techniques?
- Where does the EA Space Venn/Euler Diagram overlap with Strategy Theory, Management Theory, Process Improvement,
Operational Optimization, Project Management, Portfolio Management, and/or Senior Management?
- Based
on the deep structure of EA, what areas of the EA space have yet to be developed?
- Based
on the Venn/Euler diagram, what are the boundaries of the EA space?
- What are the strengths and weakness of
each framework/approach i.e. what is the value of the tools and techniques in each?
- Strengths and weaknesses
may well be relevant to the specific, point-in-time context of the enterprise being architected, in which cases what are the
characteristics of an organization which maximize the strengths and minimize the weaknesses of any framework/approach?
- This
is especially important in the context of "EA lite" or "EA a la carte" where an organization
is constrained by limited budget, time and/or resources and cannot engage in full-scale EA. What is the
ROV of "EA lite" or "EA a la carte"?
- Can individual enterprise architects be mapped
onto areas of the EA space if a third dimension which measures enterprise architect competency is added to EA space?
- What
is competency in the context of the enterprise architect?
- Presuming that competency is more than certification knowledge,
how can enterprise architect competency be measured?
Who is asking these
research questions today and what do they do in practice? - Many perceive these gaps,
but I think the general feeling is that these questions are too large for an individual to answer. In practice
they merely do what they know, relying on the mold of previous experience and the inspiration of intuition. Those
who are promoting frameworks/approaches tend to do so at the exclusion of other frameworks/approaches and have no incentive
for an unbiased investigation of the entire EA space.
- I think the big questions of EA aren't around specific practices
(as in tools & techniques of specific frameworks/approaches), the big questions are what is EA? what is not EA? how can
EA be rationally applied to a specific enterprise to maximize ROV?
What long-term participation models make
the most sense to pursue these research questions? - I would say that the model of
project managed open-source software is the best. A small team takes the lead by developing processes, methodologies and software
to support research into a specific area of inquiry. A project plan is then put together and led by this
small team. Anyone in the larger community can then signup to work on specific, well-defined tasks with
clear end-points and demonstrable deliverables. This type of of model would work very well for the Venn/Euler
diagram population. Alternately, a less controlled, more Wikipedia like model could be adopted.
Personally I think that the wikipedia model would be too loose and too prone to flame wars, sock puppeting and editorial inanity.
What other areas of fundamental research expertise will need to be brought to bear? This includes expertise
in experiments, surveys and bringing several theoretical perspectives. - There
will be a need for computer science, business school and social science expertise (with a moderate dose of statistics).
There will be a need for graduate research assistants and thesis students to fill-in the unknown areas in a systematic
way. There will in fact be a simple academic research need, by which I mean literature searches, industry
surveys and statistical analyses. There will be a need for an established infrastructure that supports
applying for industry, government and private grants and more importantly the infrastructure to administer and disburse these
grants.
How will thought leaders in industry be engaged to work with academic researchers on these questions?
- Thought-leaders with a point-of-view will have a stake in making sure that their favorite framework/approach is accurately
included in the Venn/Euler diagram. Thought-leaders without a point-of-view will have a stake in defining
ROV and mapping individual EA competency to ROV within specific enterprise contexts.
How the needs that academics have for
research publications be balanced with the immediate answers needed in industry so that we can motivate researchers to
work on problems important to industry? - The basic collaborative model is publishable,
as are the tools. Any results on ROV or mapping framework/approach to an enterprise or on EA competency
are highly publishable. Furthermore, any enterprise engaged in EA or considering EA will be very interested
in ROV and Competency in order to improve/augment ongoing EA. Any organization considering EA, especially
those considering "EA lite" or "EA a la carte" will be very interested in the results there.
2:47 pm est
|
|
|
2009.11.01
|
What's New? (minor updates are happening all
of the time!)
12/31/2009 - Content updates 11/27/2009 - Minor updates to content & formatting;
First blog post 10/06/2009 - Updated My Resume and Favorite Links Pages 10/06/2009 -
Added Definitions Page
 |
|
