Search This Blog

Moving the Goal Posts

OMB has also revised the definition of "data center" to include anything from "wiring closets" to facilities 500-square feet large and more, said Federal Chief Information Officer Steven VanRoekel said. (1)

OMB now says government agencies are on track to shut down 962 data storage abodes by the end of fiscal 2015. OMB's initial goal was to close 800 by that time, with its target based on a program that only included centers of 500 square feet or more.

 This move is a welcome revision of the original goals that were based on a count of data centers with more than 500 square feet. (2) When the services finally count all of the wiring closets dispersed through offices they may discover that the actual number of “data centers” is likely much larger than originally estimated. DoD components will be now able to report meeting OMB consolidation targets by relocating isolated servers into locations where shared physical space makes the target counts easier to report. Whether this will actually deliver savings is arguable, because the cost of rewiring and relocation without corresponding software and operator savings will not contribute to the large expected savings.

We have already some numbers on 2010-2011 data consolidation savings in response to the 2010 Kundra’s mandate. OMB published savings numbers on October. So far the reported savings are $14.6 million. (3) That’s 0.18 percent of the Fed IT budget. The reported savings are from Agriculture $4.2M; EPA $1.5M; Justice $1.9M; NASA $0.3M; State $1.7M and Transportation $5M.

Summary
How the ambitious data center cost reductions will be accomplished and the results will be accounted for is anybody’s guess. OMB should require that in the future all of the reported savings should be shown in long term reductions in capital and operating dollars, not only on data center counts.


  (1) http://www.fiercegovernmentit.com/story/audio-federal-cio-steven-vanroekel-announces-data-center-consolidation-upda/2011-10-06
  (2) http://pstrassmann.blogspot.com/2011/01/data-center-consolidation-is-feasible.html
 (3) http://meritalk.com/blog.php?user=SteveOKeeffe&blogentry_id=3070&utm_source=AdaptiveMailer&utm_medium=email&utm_campaign=MeriTalk%20Newsletter&org=2021&lvl=100&ite=1702&lea=2252793&ctr=0&par=1

A View of Fractured CIO Roles

There has been a steady increase in Congressional attention to information technology. This is shown by requests to GAO to report about the ways in which legislative mandates have been carried out. The most recent report, “Federal Chief Information Officers: Opportunities Exist to Improve Role in Information Technology Management”  is critical how the government handles the accountability of CIO leadership.1

GAO listed the statutory responsibilities that require CIOs to be responsible for IT. As a precedent there is a long list since 1994 of reports that has documented a consistent lack of accomplishments.2  The GAO offered the following findings:

1. The majority of CIOs were not fully accountable for the 13 areas that were assigned to them by statute. Most of the CIOs did not have responsibility for information collection, records management, privacy and statistical policies. As result, statutory requirements were not met.
2. About half of the CIOs were holding jobs in addition to serving as the CIO. Holding multiple positions is contrary to the federal law.
3. Federal law calls for agency CIOs to report to the head of an agency. Only 17 of 30 CIOs had such a relationship.
4. The median tenure time for CIOs was just under 2 years. That is inadequate since it takes at least 3-7 years to implement major changes.
5. CIOs reported that they were hindered in exercising their authority over agency IT budgets, component IT spending, and staff.
6. Close to a half of the CIOs did not have the power to cancel information technology projects.
7. CIOs faced limitations in their ability to influence agency decisions on integrating systems within their agencies.
8. In a number of agencies the CIO did not have visibility into a majority of the agency’s discretionary investments and could not ensure that the investments were delivering expected returns.
9. In many cases the CIOs did not have hiring capability for IT staff, influencing component agency CIOs’ performance ratings.
10. CIOs do not always participate in selections for candidate personnel and did not participate in the component CIOs’ performance reviews.

SUMMARY
The top CIO leadership position in DoD is that of the OSD CIO.  There were five executives in that job in seven years. Of these two OSD CIOs were political appointments, which require an extended period for orientation. The term for serving as the top DoD CIO requires a much longer tenure, of at least five to seven years.3

There were five Air Force, five Army and four Navy CIOs during this period. In almost every case the terms for the top DoD CIOs did not overlap with services CIOs, which made the launching of shared programs difficult.

The Clinger-Cohen legislation established the position of the CIO. It was inadequate for the creation of a solid foundation for managerial accountability for DoD IT spending. Its primary focus was to clarify acquisition responsibilities. All other critical questions, which addressed matters such as control over budgets or personnel, were left to prevail along lines that had roots in processes dating back to “cold-war” administrative methods.

This is why the latest GAO report should be read as an unfavorable view of the inadequacies how leadership and accountability for DoD IT is managed.




 1 http://www.gao.gov/products/GAO-11-634)
 2 GAO/AIMD-94-115; GAO-01-376G; GAO-04-49; GAO-04-823; GAO-04-394G; GAO-09-566; GAO-10-872T 
 3 Strassmann served as the first de-facto appointee in this position as Director of Defense Information, before the title CIO was established.

Flaws in the DoD Architecture According to Apple

Apple is the most valuable technology firm on the planet. DoD is the most valuable and most costly user of information technologies (well over one percent of global spending for IT).
Is there anything that DoD could learn from the success of Apple? Here are four fatal flaws, which explain what is now generally agreed that DoD IT is not performing well:

Flaw #1: Apple designs its software, products and operations together. 
Apple’s approach to software and hardware is tightly linked. Apple is organized for central management and control, with long-term leadership. Apple systems have short-term delivery schedules. They have tightly defined tactical increments that fit a long-term design.

The DoD software design is subdivided into a planning, software acquisition, software development and software testing phases. Operations are subdivided into acquisition and operations phases. All of these are disjointed and assigned for execution to separate organizations. The linking between these efforts is obtained only by means of elaborate procedures.

DoD systems are fractured to get anything accomplished. Coupling across applications is never achieved because of elongated schedules. DoD systems are not built to an overarching architecture. Systems projects do not contribute to a coherent design. DoD is organized for local management and changing leadership.
SUMMARY: Apple is coherent. DoD is disjointed.

Flaw #2: Apple designs reflect unified leadership. 
Apple controls the contents, the format, the applications and the hardware that deliver its products. It is a closed system that promotes innovation by executives who drive the entire organization to deliver superior products and services but only from a limited set of proprietary options. The management is totally centralized and organic to the firm. It subcontracts only when that is economically advantageous though always under Apple control.

DoD does not control the contents, the format, the applications or any hardware. DoD only defines generally its requirements, leaving to contractors to dictate how implementation is accomplished.
SUMMARY: Apple leadership runs the enterprise. DoD contractors run how the enterprise operates.

Flaw #3:  The Apple enterprise is vertically integrated.
Apple has a coherent end-to-end visibility of its products and services. Changes in the architecture, design, operations and products are managed for inter-dependency. Though a part of its value-chain is subcontracted, that is never done with a loss of unified control. Apple manages everything what it delivers.

DoD has no end-to-end visibility of its systems. The management as well as the budgets for delivering a systems function to its customers is widely distributed. Accountability is diffused.
SUMMARY: Apple management controls its systems, including security. DoD has distributed control over its diverse systems and is unable to impose a coherent approach to assuring its security.

Flaw #4: Apple manages its products and services from source to use.
Apple dictates everything from design to manufacture to distribution. This is inclusive of prescribing proprietary solutions as to what microprocessor should be used, how and where services will be deployed and what marketing method will be applied. Fundamentally, Apple operates in a closed environment and therefore is in a position to offer security safeguards.

DoD does not dictate software or hardware choices but only the acquisition that picks whatever meets procurement requirements. DoD deals with the problem of information security by safeguarding a wide range of vendor offerings.
SUMMARY: Apple manages everything and therefore can offer greater security. DoD tries to safeguard its systems by attempting to protect against every possible vendor risk. 

Significance of the Apple 4S SIRI Cognitive Assistant

Among the features included in the new Apple iPhone4S there is a personal assistant called SIRI. It uses speech recognition and artificial intelligence. It offers a tight integration to the iPhone's built-in apps by creating a "personal assistant". 

SIRI was not developed by Apple but was a recent acquisition. It was a spin-off project from a DARPA-funded project called CALO. This was a five-year initiative involving 300 researchers to adapt AI technologies to create a "cognitive assistant" – which is a good way to describe what SIRI delivers.
CALO was an Artificial Intelligence project funded by DARPA under its Personalized Assistant that Learns (PAL) program. It was a well-funded five-year contract, which brought together researchers from US top university and commercial research institutions. The goal was to build a new generation of cognitive assistants that can reason, learn from experience, be told what to do, explain what they are doing, reflect on their experience and respond robustly to surprises. SRI International was the lead integrator responsible for coordinating these efforts.

SIRI sends messages, sets up reminders and searches for information. It understands not only conversational speech but also the meaning of sentences. In this way unstructured talk can be used for natural communications. SIRI can use information that is already in the iPhone—such as a person’s GPS location, contacts and relationships—in order to provide personal assistance.

For instance, SIRI can: 
1. Uses machine learning algorithms to build a model of who works on which projects, what role they play, how important they are and how specified documents are related.
2. Can help a user put together new documents by using what has been learned about the structure and content from previous documents accessed in the past.
3. Assist a user how to interact with people, both in electronic forums (e.g. email) and in meetings. It can automatically generate a meeting transcript, tracks action item assignments and detect what are the roles of participants.
4. Automate routine tasks such as travel authorizations. It can be taught new procedures and tasks by observing and interacting with the user. It can a person’s preferences for getting things done while managing a busy schedule.

SUMMARY
From the standpoint of DoD tactical warfare operations the SIRI technology offers capabilities that vastly exceed what is currently available to our war fighters. By offering unstructured interactive voice conversations the military can now proceed to replace the current radio chatter that fills bandwidth especially if a unit is under attack. The ability to “train” hand-held devices that fit into a shirt pocket to answer specific queries makes it possible to simplify how command and control of forces can be applied.

Voice recognition makes a SIRI device adaptable to the characteristics of a soldier’s speech pattern.
The Apple is the first one with an offering of an intelligent communicator based on DARPA innovations. However, there is no question that other smart-phone firms will be also coming out with new features such devices that deliver added security safeguards.

The new Apple 4S device costing $299, weighing 4.9 ounces, sized 2.3 by 4.5 inches, with an 8 megapixel camera, a built-in GPS and 200 hrs. of standby power is ideal for Special and Expeditionary forces engaged in asymmetric warfare. 
 
When viewed from the standpoint of the history of information technology this first commercial availability of an intelligent “personal assistant” should be recognized as a revolutionary event.  Such devices will offer solutions that will have major effects on the ways the military can now shift to new class of battlefield computing devices. 

Google vs. Apple

Two giants are emerging with rapid growth. Google and Apple are defining the structure of consumer computing. Google (revenue $33 billion, market capitalization of $166 billion) and Apple (revenue $100 billion, market capitalization of $353 billion) differ from firms engaged primarily in enterprise computing such as IBM (revenue $104 billion, market capitalization of $209 billion) and Oracle (revenue $36 billion, market capitalization of $145 billion). Microsoft (revenue $70 billion, market capitalization $208 billions) straddles between enterprise and consumer sectors.

The distinctions between consumer computing and enterprise computing are vanishing as office personnel are becoming increasingly mobile. Over half of office work does not take place any more by employees sitting at their desks. Business is now done by employees moving and changing locations. This is why people that move must rely on devices that are available wherever a person is located. Personal computing now means the access to the Internet and to application from any anywhere. Apple with iOS and integrated portable devices are confronting Google devices that are managed by the Android software that support a variety of vendors.

In the next few years it will be Google or Apple that will shape how organizations start thinking about ways how to manage their computing. IBM, Oracle and Microsoft will continue operating in their respective market segments. These firms have not emerged as pacemakers of IT architectures. HP, the one time leader, is now sitting on the sidelines.

Google and Apple are fundamentally different. For Google, the Web is the center of the universe though this firm does not own the Web. It only manages access to it primarily for marketing purposes. Its ownership of applications is marginal. They do not own wireless phones or appliances. Google can be categorized as an un-integrated firm with the exception of operating a well-organized network of special purpose data centers. Other than the hardware and software of its data centers, Google outsources everything.1

For Apple your device is the center of everything. It owns a formidable collection of phones and appliances and owns or controls its logistics pipeline as well as its applications. Apple controls its integration from microprocessors all the way to dedicated cloud data centers.
Google makes money by making access to the Web easy using open source browsers plus its own proprietary Chrome and then derives profits from advertising revenues. Apple sells proprietary hardware with proprietary applications and its own proprietary browser (Safari) and gets profits from selling premium priced devices plus selling Apple-approved applications.

Google’s entire strategy is based on the future, and not the Internet as it is today. Google is betting that the world will have low-cost, ubiquitous Internet soon, including fiber connections in offices and homes and super-fast mobile wireless broadband in every nook and cranny of the planet. Google’s apps are connection-dependent in the Internet. All of the data is stored on Google’s servers in the cloud or on servers on the edge. Google controls only the browser and a limited set of applications.

Apple’s approach is not to use the cloud as the computer-in-the-sky that runs everything. Apple doesn’t want or need everything to happen in the cloud. Instead, it views the cloud as the conductor that makes sure all of applications are synchronized. Apple controls the security at every point of the network as well as its network. The Apple Cloud is the hub that is controlled and synchronized only by Apple. Users manage their data and prefer to store personal data on Apple devices that are easily upgraded.

Summary


DoD implication: A GIG that is based on access via secure servers on the edge will be the lowest cost and most flexible solution, but more risky. However, Google cannot be adapted to the GIG with its limited applications and a high diversity in smart phones.

Apple’s view is more relevant because it is far more integrated. It offers a more secure set of technologies. Unfortunately, neither Google nor Apple is as yet structured to support enterprise level systems. This leaves it to the established enterprise providers – Microsoft and IBM to compete for DoD’s enterprise clouds until such time when either Google or Apple will offer products that extend from personal to enterprise uses.

From the standpoint of the individual, Apple offers by far the lowest cost of ownership on account of its overall ease of use. Apple provides an overall integration of every component that is required to run the user end of IT operations.

All other vendors manage only a part of the total set it takes to run IT. Whether the monolithic and proprietary approach pursued by Apple may prevail is not clear, though rising security requirements seem to favor such an approach. In the next ten years the dollars and the preferences of satisfying the personal computing needs will most likely give to Apple the revenue and the profit edge.

  1  http://www.zdnet.com/blog/btl/philosophical-differences-the-google-cloud-vs-the-apple-cloud/50245