A Huge Price Tag for a De-Centralized Implementation. Twenty Four Billion dollars.

 The GAO report is here:  http://www.gao.gov/assets/590/589796.pdf

 Are you at least curious what I’m talking about?  Of course, you are, you are still reading.  Or watching a Google advertisement – I insist you do one or the other before you leave.  Check out this video of a 3-minute walk through of what the Army says Global Combat Support System (GCSS) is:  http://www.youtube.com/watch?v=yMgDJNo2_PM Did you gloss over 1:06?  Watch it again and listen to LTG Charles Mahan (Ret) emphatically say that SAP (also known as Systems, Applications, and Products in data processing) was the only company that could accomplish this.  (Soldier for 35 years and a VP for SAP circa ’05/’06 via LinkedIn) I’m only mentioning this because in the world of contracting the pool either has one ring, two rings, or three rings.  (I don’t want you to think that hard; most kiddie pools stop at three)

Could they (SAP) not have also accomplished this for the USMC?  In 2020 allow the ability to perform data mining on the same systems too.  Together.  One team.  One mission and purpose.  There is my problem – being logical.  Let’s not get carried away–that’s too big of a task and purpose! (like healthcare.gov)  If we build one system structure, then it will be undermined by security nuances.  Wait, why is it that there is so much information out there about how GCSS is being built (read: if you know how I created it; you can undermine it)?  I heard someone once say that SAP, Oracle, whatever, they would all be interfaced to one another.  Let’s consider:  Does a contracting company who has invested over 10+ years in building this solution – will they make in easy integration?  I’m looking at you sideways; only you can’t see me.

What confounds the concept is that the ideas of centralizing the efforts, nay, stopping to take a moment of pause, and recognize that implementation of an SAP (for GCSS) system for the Army and an Oracle-based solution for the USMC didn’t make good business/data (ha) sense but decentralizing fundamental concepts did.  Consider hacker attacks-they build to target one type of system, for the most part, so the complexity of attacking both SAP and Oracle a little more difficult.  At least that’s what we can argue.  It is (dramatic pause), complex.
A thesis, public information, was written in September 2010 by Mark Jones.  This paper identified the struggles that GCSS was plagued with, the Government Accountability Office (GAO)’s identification of some issues (in 2010) and clear recommendations to continue. Thesis paper here  I’ve extracted the meat (grain fed organic!):
This two release strategy reinforced the philosophy that designing and implementing smaller increments of capability over short periods of time allows capacity, even if it is reduced from the original plan, to be fielded earlier to the user (it’s like the waterfall to software developers).  Know that the “user” is the warfighter.  The men and women we laud on days committed to them and invest our reverence to their very existence.  (Without them; we can’t write informational blogs.)  This is a critical concept in building systems to support DoD – at the end of the day if your requirements don’t fulfill that purpose means; then fail.  Bean counters sweat bullets.  If you catch my drift.  But, is the end product supporting real users.

The question I pose (oblivious to the one above), why should the DoD spend as much money as they have in six systems when maybe one system could be developed to be used by all four Services with crucial security components explicitly tailored to thwart common hacks.  Again, I’m visited by ghost past: Healthcare.gov

Effort: RICE (customized Oracle) Objects ACTUAL– Severe customization. SHOULD HAVE – Conform to core software functionality and change business processes of each Service (but we’ll say USMC because, well yes).

Post Deployment System Support – ACTUAL – SCATTERED to the winds with multiple contracts, configurations, and release mgmt.  SHOULD HAVE – Centralized, single contract, setup, and release mgmt.

The GCSS-MC program recognized the complexity and technical challenges in customizing the COTS product and moved the RICE objects for the cross-domain solution and data synchronization capability into Release 1.2 allowing the GCSS-MC program to field the essential skill of supply and maintenance management sooner to the warfighter and speed up the process of eliminating the legacy supply and maintenance systems.

I couldn’t have stated it any better than Mark, “Therefore, the perception of buying a COTS product to minimize development effort is exactly that, a perception. In addition to not supporting the functional and technical requirements, the COTS product does not provide the unique Service Component business processes nor does it provide the unique ...financial requirements. The implementation of the unique DoD requirements translates to a large development effort in terms of customization and configuration.”  I’m sorry; I know you are bored…
The Strategic Plan intends to share information across DoD and with mission partners.  In summary this includes: the data itself; the Department’s management over the information life cycle; the processes associated with managing information to accomplish the DoD mission and functions; activities related to designing, building, populating, acquiring, maintaining, operating, protecting and defending the information enterprise; and sharing of related information resources such as personnel, funds, and equipment.

To develop such a system, a Lead Service or external entity could act as the Program Office and manage the entire lifecycle support of the system.  —-that right there?  Is just insane.

But I don’t think it is.  I have experienced the lack of subject matter expertise in software and system development, and I’m almost certain that years of drowning in legacy systems only inebriated value-added growth.  One vision, multiple missions, and end product that is usable, repeatable, and has outputs that drive decisions.  Magic.  I heard somewhere that magic was expensive.  It indeed is when division comes into play.