Topic F: An Information Age Acquisition Organization


Management gurus of the information age proclaim the need for "networked" and decentralized structures in order for organizations to survive. Perhaps the movement of the military into information age warfare implies similar organizational changes in the development and procurement agencies.


----------- [Moderator] Should the procurement of information technologies be decentralized? Should procurement funds be given to the unified commands (e.g., Pacific Command) and/or individual units instead of the individual services (e.g., Army, Navy, Air Force)? Or is a "joint" acquisition organization (perhaps similar to the Strategic Defense Initiative Office) more appropriate?

[Campen] The framing of this question still implies a focus on THINGS. Information services should be established by functional basic ordering agreements that apply to similar functions in all services, on a global basis. Individual units then access these information services essentially the way they now do food and water, paying a users fee, if necessary, to support a common industrial fund. Using Intelligence as an example, DISA [Defense Information Systems Agency] provides the communications protocols and controls the network at the operational level; functional activities, such as DIA, NSA and CIA provide the functionally tailored information services, and the user taps into the system as required.

[Cebrowski] The spirit of Goldwater-Nichols and other mechanisms address these issues in that: a.) CINCs have more input in defining warfighting requirements_which is inherently decentralized b.) Operational units have significant latitude to apply discretionary funds in a manner that best meets their needs, and c.) The Joint Requirements Oversight Council ensures maximum warfighting efficiency by validating and articulating requirements to the Defense Acquisition Board_in essence, they embody the necessary ingredients of a joint acquisition organization. I prefer joint and decentralized to reinventing communism.

[Cochrane] Speaking from experience of a commercial organisation, decentralisation causes problems. A decentralised organisation results in a plethora of systems performing the same job in different location, on different platforms written in different languages. A centralised approach should provide a co-ordinated approach but needs to be fast and give the flexibility in the field. However, care must be taken to ensure centralisation does not bring the disadvantages of complete standardisation. If an organisation were to standardise on a particular type of hardware, it would be completely compromised if the hardware were to suffer a serious security breach. Are there any reasons for each service to procure information systems separately, let alone each unit? I would think that the needs of one type of military organisation are pretty much the same as the other. One absolute necessity is that the systems should interwork seamlessly. It would also be nice to think that a marine defending an army base should be able to use their equipment in the same way as his own kit. If each service can not work with each other then the game could be lost before it starts!!!!!

[Cohen] Centralized planning_decentralized execution. The price and terms should be negotiated on a DoD-wide contract to minimize costs. The specific equipment should be ordered and processed by the most local person needing it. The acquisition should be based on an established need and method for fulfilling that need identified to and approved by the chain of command.

[Dunnigan] It already is decentralized. Lotsa luck separating the individual eggs from that omelet now.

[Garigue] The question of centralized vs. decentralized organizational structure is a big debate and an essential one. The money should go where the knowledge is. Who knows best should decide. Certainly the architecture and the standards could be centralized and the acquisition and the procurement can be decentralized. But in the case of very large organizations such as the Army, Navy, etc., the question cannot be resolved because all organizations have enough capabilities and resource to suffice to their own needs. None will agree on either solution. So the debate will continue.

[Giessler] Decentralized? Yes. Procurement funds given to unified commands and/or individual units? Yes for most I.T., only unique non-COTS stuff should be centralized. Is a "joint" acquisition organization more appropriate? Only for unique stuff and that should be less than 5 percent.

[Gust] We have had CINC-initiatives with discretionary funds for many years. It has resulted in some duplicative procurement and some non-interoperability horror stories. One cannot even specify UNIX without some disconnect across that product line. I think some centralization at the Service level is appropriate, mainly because that is where the best handling of funding and most flexibility in any reallocation or reprogramming of funds occurs. A joint activity carries too much baggage. My latest and most perplexing problem is to try to start a new common intell broadcast receiver program without a finalized joint ORD [Operational Requirements Document]. I have FY 96 dollars that must be spent, but the joint ORD approval process, which requires CINCS' coordination, will take so much time as to be in the next fiscal year. Jointness in buying JTIDS [Joint Tactical Information Distribution System], GPS, MILSTAR are three examples where time was traded for joint coordination, but Services retained the dollars on these programs. That is probably the best way to go.

[Hazlett] Should be more coordinated, not less. Particularly important from an affordability viewpoint. There is little overlap today between the services, even where there are similar, or even identical requirements. Virtual organization concept has merit in the acquisition world. Services, CINCs, etc., should form temporary virtual alliances when like needs exist, or joint development makes sense. One large acquisition organization may be too unwieldy.

[King] The goal should be to make the procurement process as efficient as possible. Something like a central group that prepares specifications and bid guidelines and then each command doing its own actual acquisition.

[Loescher] I would go further. ALL organizations, including tactical organizations, will be streamlined. Modern command is about moving information to take advantage of openings that happen orders of magnitude faster_and further from the tactical operations area_than war in the past. The officer hierarchy structure is a reflection of 19th Century land warfare and 20th Century ricebowls. With respect to development, a centralized, much smaller, Service-specific R&D outfit with ties to industry labs and Universities is the way to go, at least for the next 5-8 years. Government employees at the mid-level cannot stay in touch with commercial technology to buy it properly; but they can be focused on Service unique problems, which are not insignificant. Procurement of information systems should be decentralized; procurement of large information pools may be better bundled.

A joint office, just means, joint bureaucrats. What we need are new ideas to create new markets for new industry to sell to us_not more bureaucrats to oversee the existing industry, which is a reflection of yesterday's acquisition.

[Probst] Definitely, but interoperability can be guaranteed. But management gurus are not to be blindly trusted.

[Steele] Yes and no. Local authority over the realignment of funds is critical. For instance, PACOM [Pacific Command] briefed me several years ago that they had hundreds of thousands of dollars earmarked for TEMPEST hardware they did not need, and zero dollars for open source collection and production (including their library), so they had to cancel their LEXIS-NEXIS account. Generally the local command knows its needs best. HOWEVER, the current system of penalizing commands for savings is out of touch with human nature. They should be allowed to roll saving overs and reinvest in other priorities. Joint mandates of specific systems and software will produce results as dumb as the Marine Corps suite, where one word processor is mandated and purchased, and ignored Corps-wide, while WordPerfect is bought by hook or crook and is the actual Corps standard.

[Todd] Realizing the short term focus of the Unified Commands, the truly "out of the box" thinking has come more from technology push from the Services (stealth as an example) than from requirement pull from the Unified Staffs. Of course, I realize this is an iterative process and the integrated priority lists are of significant value to identifying what to pursue technology wise. Likewise, a joint office has its place in a very restrictive role. The Services should retain their function in the R&D arena.

----------- [Moderator] Should "acquisition technologists" be put in operational units in order to quickly acquire and transition new capabilities?

[Cebrowski] This is more a matter of process and education than one of organization.

[Cochrane] Three options depending upon how you define information warfare.

- Moving operational people to the technologists would be far less dangerous for everyone involved. Technologists could get military requirements for systems in the same way that they acquire requirements for commercial systems. When getting new kit into the field I would think operational personnel would be happier learning from their peers who know the problems of operating in the field rather than suits from computer companies. - As new information warfare technology can change the face of a battlefield within hours it may be necessary for IW units to contain not only acquisition technologists but also their own front-line research and development teams. - Since information warfare creates a whole new battle field, cyberspace, it creates a need for people that understand warfare to work closely with technology gurus.

[Cohen] No. The person who actually needs a particular piece of technology should be able to go and get it if they want to, but it should be a pre-approved item or technology (except on an emergency basis) from a pre-approved source at a pre-approved price.

[Dunnigan] Put them somewhere out of the way so that the end user organizations can go get what they feel they need.

[Garigue] The logistics officer should be able to buy anything_putting it together is another issue. There needs to be a Information System Officer for that. With new technologies there are new classes of specialists that are required. In the navy we went from sail to steam, with this a new officer cane into the wardroom, then with electricity came another type of officer, now with software becoming the dominant war technology there will be the requirement for a new specialty in the military.

[Giessler] Yes_but that will happen when we select people who think and have an I.W. paradigm and educate them at NPS and AFIT [Air Force Institute of Technology] and all PME [Professional Military Education] at all levels. The Thrashers and Elams and Garcias are already the info science and technology officers and officials. All we need to do is get out of their way. And Comdr Loescher is leading that effort at USN headquarters.

[Gust] The Army still uses its TRADOC organization as the "user's rep" to consolidate requirements. The last thing we need is for each Army division to buy its own technology and then have to join up in a large force side-by-side like in Desert Storm. If you want it bad, you'll get it bad.

[Hazlett] Should stand up several joint and single service organizational and technological testbed units whose manning includes acquisition technologists. Not sure that they really belong or can be kept gainfully employed in regular operational units.

[King] I would centralize the technology part of this and distribute only the actual acquisition part.

[Loescher] In the current model_buying systems vs. commodities_we are wasting our money. Decentralizing acquisition to buy systems decentralizes wasting of money. We need to get our heads out of hardware and software and into information. We don't need acquisition technologists then, we need at new kind of operator, who understands infotech as the technology of modern warfighter. He/she needs to take their places beside the aviation, armored, strategic bombing, etc, innovators of the 1920s to lead us into a different kind of warfare.

[Probst] I think so. Maybe not an operational unit strictly speaking.

[Steele] Acquisition is NOT an arcane specialty that requires magic incantations and special knowledge. That is the OLD acquisition paradigm where convoluted formulas had to be learned over years of study and practice simply to stay out of jail. The new paradigm should rely on open market viability and common sense.

----------- [Moderator] With the focus on use of commercial products, where should the military acquisition organization concentrate its focus? Research and development? Management of contracts and contractors? Integration of systems? Development of interfaces and architectures?

[Campen] Assuming we need but a very small one, its focus should be on architectures, integration and interfaces, and continuity of service.

[Cebrowski] It should focus on simplification, outsourcing, teaming with suppliers, and making itself as small as possible.

[Cochrane] There is no simple answer to this: well defined architecture and interfaces are essential for rapid systems integration and reliability. However, continuous changes in the battlefield can only be met with new technology provided from a well funded and broad based research programme.

[Cohen] No. Yes. No. No.

[Dunnigan] If you buy in sufficient quantity, you can have a partial production run modified to your particular needs. This is done all the time. But you have to act like commercial purchasing operations (i.e., efficiently.)

[Garigue] Acquisition organizations should focus on shorter acquisition cycles and rapid distribution.

[Giessler] Most of it should go away but that which is left should either be a demonstration center which shows off commercially available products and what will soon be available or it should be acquiring unique stuff. The acquisition field must change. Research and development? Not much here and most of this should be done across service if not at the JCS/OSD [Joint Chiefs of Staff/Office of the Secretary of Defense] levels. Management of contracts and contractors? Local user goes out and buys COTS and support necessary_that will be different for each user and may change when someone arrives, then leaves a job. Integration of systems? The local group will decide if they are competent to create interoperability and integration software, hardware, management ware, etc. If not they will hire it done for the 12 to 18 months before they buy new stuff. Some parts of their info system will always be changing by their COTS buys. Development of interfaces and architectures? The C.O. [Commanding Officer] has the vision and his people either develop or hire or buy the interfaces that will allow them to do the job with their constantly adapting architectures.

[Gust] I think the focus has to be on all frontiers, none at the exclusion of the others. We just have to be cognizant of what the other players are doing.

[Hazlett] Military acquisition organizations should concentrate their efforts on identifying those areas where the military can best benefit from economies of scale and on identifying and coordinating the management of cross-organization and organization unique needs.

[King] R&D to keep abreast of changes and development of interfaces and architectures in order to provide the acquisition guidelines.

[Loescher] The first two. There are many Service-unique or at least military-unique R&D problems that will never be solved in industry. Secondly, we need to create a process at the contracting level to capture innovation and to learn how to pay for information, which is different than paying for torpedoes.

[Probst] DoD does almost no research itself, and not much more development. DoD labs are quite different from DoE labs. Only options two and three are feasible.

[Schwartau] R&D for basic technology yes, in cooperation with the private sector, more along the European model. Standards are the key. I bet my bottom dollar that in a contest, a commercial outfit could get a system up and running much faster than a burdened military structure given the same tasking orders. The commercial folks would pick and choose available parts and glue them together for fast functionality. For military applications, a hardening step would be required, especially for mission critical life/death systems.

[Steele] There are always going to be some areas where the private sector simply will not support the kind of R&D that is necessary for unique military requirements. In the C3I area:

_Tactical document acquisition and digitization (such as rapid and rugged scanning of rough documents that are crumpled, wet, and hard to read; take on non-trivial pattern recognition problems

_Automated time and space tags on all multi-media information

_Build the bridges from commercial remote sensing platforms to the NRO/GPS precision points and then return the production of 1:50,000 combat charts with contour lines to the private sector_also build the bridges from commercial imagery through GPS/NRO to precision munitions already mounted on aircraft

_Digitize interactive speech for military police, coalition command & control, prisoner interrogation

_Communications & computing security

[Todd] The military should concentrate on integrating those systems deemed necessary to having a positive impact on the information battlespace. Knowing and having reliable information on our own forces; then being able to acquire, mix, synthesize and distribute information about our adversaries; then being about to orchestrate truly coherent operations against the enemy will have the greatest impact in future conflicts.

----------- [Moderator] What are the institutional impediments to the creation of an information warfare infrastructure?

[Campen] Trying to adapt a structure built to buy, field, install, maintain THINGS into one needed to obtain information services. As the Congress noted in passing the new rules for acquisition of information technology, the hardest thing to change will be the culture.

[Cebrowski] The infrastructure already exists_what's needed is more time and effort for intellectual maturity.

[Cochrane] Impediments to the creation of an information warfare infrastruct ure. - Military and governmental establishments and those that derive power from their positions in such organisations. - The acquisition agencies with their preferences and cozy relationships with the "safe" traditional suppliers. - Selling the concept to the public. If you say that there is a need for information security improvements the implication is that systems are insecure. No one wants to admit to problems, they just want to sort them out on the quiet, the result is that knowledge is not shared. The public are also quick to see a conspiracy and think that faceless governments are going to use the systems to further their own aims.

[Cohen] Were there any?

[Dunnigan] Stepping on someone else's toes.

[Garigue] The mind set. Cyberspace is not seen as a possible battlespace.

[Giessler] The most difficult is to get national security operatives at all levels to drop their old paradigms. The infrastructure is an implicit one that has to be inculcated not defined. It has to be part of the control system of individuals and organizations who understand the goals, objectives and CO's that capitalize on the COTS technologies_and then educate, train and exercise the force to create a revolution in military af fairs.

[Gust] Everyone wants to be in charge. How many times have we have a new initiative and one Service has jumped up to lead so that their service-unique views or needs are served first? Examples are UAV [Unmanned Aerial Vehicles], strategic defense, space-related technology. We cannot all be the leaders, and the leaders have to be more accommodating of the needs of the other services. The followers also cannot be too parochial in their demands for "all or nothing" type solutions.

[Hazlett] Rice bowls and service rivalry, primarily. The services have been way too quick in carving up IW into narrowly defined areas. Some IW areas can best be handled in concert, rather than singularly.

[King] Information warfare is much more abstract than conventional warfare and it will be harder to get people to think about (and fund) it. In order to successfully control a networked nation it will probably be necessary to increase the security and therefore the restrictions. This goes against the current "open" atmosphere.

[Loescher] Institutions. You cannot create a new kind of warfare with old stovepipe warriors. But the stovepipe warriors control promotions_until we find a means to truncate the dynasties or until the dynasties tragically prove their unsuitability to the new world. Historical examples are everywhere, from the Trenches of WW I to Battleship Admirals to today's Information Warfare_which is basically cryptology reinventing itself in the guise of new term. There is a new Information Warfare, absolutely, but you won't get there by renaming the past, which is Navy's current mistake.

[Probst] Established privilege. Empire building. Lack of agreed-on concepts.

[Schwartau] I do not know what an IW infrastructure means here. I could interpret this a dozen different ways. Sorry.

[Steele] The institutional leaders below the Secretary of Defense. ACTUAL_the classification of the threat. The games Navy plays to compartment IW simply to keep its own toys and avoid joint endeavors. The general lack of understanding at the flag level of why intelligence is broken and open sources are a major part of the IW fix. The unwillingness of ASD C3I [Assistant Secretary of Defense for C3I] to stand tall and tell it like it is.