Eric Mack On-Line

"NASA has not demonstrated characteristics of a learning organization after investigators observed mistakes being repeated."  - A statement read from the last Space Shuttle disaster report.  (I wonder how many companies would fall into this same group?)



Presenting: Paul Caraccioli, Program manager of NASA's Marshall Space Flight Center Propulsion System Department (PSD) KM System. (Long title; they deal with engines - fast ones.)

Paul shared a case study of their KM deployment and the lessons learned.

MSFC PSD KM Strategic Plan

(see slide - sorry; too fast.)

"IT is great, but for our KM strategy to work, what we need to bring about a culture change within NASA."

DATA Gathering
Formed a KM team for strategy formulation
Structured interviews to review workflows: purpose: to determine where the collaborative, resource and tool interfaces were.

Cultural Change Survey
Working together
Knowledge sharing
Senior management advocacy
operational processes
to name a few areas

Determined that PSD was ready for a culture change. (Note to self: find this survey, how formulated)

After identifying needs and objectives, developed a solutions map (see slide)  to map to the top 10 day to day needs.
Next, developed functional categories and evaluated software products across the matrix (Lotus Notes missing, but SharePoint/groove present)

Solution Mapping/verification
- What did we do well
- What did we do poorly?
- What did we learn?

LESSONS LEARNED:
"Obtain available industry trade studies on alternative portal technologies too understand key issues and potential solutions." (e.g. CMS Watch, Gartner, Forrister)

Focus analysis upon any integration requirements between software applications that must work as a system. (Systems thinking.)

"Obtain appropriate training on the short list of candidate applications."  ***This was a big lesson learned. Buy a seat of training at the vendors training source, get to know the software.

PLUS
Conduct rigorous, real-world evaluation of candidate software applications. (Unless risk is well defined and is low.) Whenever possible, use a hands-on evaluation of the actual product in the actual or similar environment.

OR, If not practical
"Conduct Power User Interviews'

BOTTOM LINE
"Rigorously evaluate integration aspects of the solution in addition to component functionality."

System Requirements and Design Summary:
Outcome; Solution traceable to user defined objectives, user needs and system requirements.

Problem: The IT products were not integrated and, in hindsight, they should have invested in training before acquisition. A big expensive lesson (not doing the training up front.)

Decided to use the MOSS (Microsoft Office Shared Solution)

INTERFACE IS BIG - If the user's don't like the interface, they won;t use it and the project will fail.

Personal observation: seems like a lot of independent development of collaborative systems going on. Politics, funding, and scale make it harder to implement collaborative systems across and between large organizations. Although the obvious solution would be a centralized solution, I'm not convinced that's doable either. So, the current approach, as an interim solution, makes sense to me. Business could learn much from NASA in this area and I think NASA could learn from business.

The challenges that NASA face from a KM perspective are huge. Consider just the transition from the soon-to-be-retired Space Shuttle program to the next  generation transporter. The knowledge from the Space Shuttle program is largely unstructured in PDFs, design documents, PowerPoint files, personal journals, databases, and applications. How does a KM manager take inventory of this knowledge and organize it in a system where it can be successfully reused? A big challenge.

Slide presented to show the MSFC Knowledge Capture system. One thing that stood out to me is a system that allows a SME's response to a search to be incorporated into the knowledgebase. My take: Imagine if when I did a Google search Google tracked the site (or person or email) that ultimately led to my answer and added that to its own search repository? In theory, the same question would not be asked twice.