The Crash of the ACS

JAMES HASIKwww.jameshasik.com • 512-299-1269 • [email protected]

The Crash of the Aerial Common SensorIntelligence, Surveillance, and Reconnaissance systems in Counterinsurgency, and the Economics of Network-Centric Warfare

23 January 2006

——————

In January 2006, the US Army cancelled its Airborne Common Sensor (ACS) project with Lockheed Mar-tin. The twin-engined turbojet ACS was to have replaced three US signals intelligence aircraft: the Army's RC-7 Crazy Hawk, a four-engined turboprop based on the DeHavilland Dash-7; the Army's RC-12 Guardrail, a twin-engined turboprop based on the Beech King Air; and the US Navy's EP-3E Aires II, a four-engined turboprop based on the old Lockheed Electra. The program may have seemed small given the number of airframes procured—38 for the Army, and just 19 for the Navy—but the cost of the elec-tronics and software meant that Lockheed Martin's development contract would have been worth $879 million. Subsequent production would add another $5 billion to that.

The difficulty that lead to the cancellation was that the Army and the Navy's requirements for the air-craft did not match, and the airframe that Lockheed Martin selected, Embraer's ERJ-145 regional jet, of-fered little growth space. After months of working with Lockheed to resolve the problem, the Army and the Navy decided to terminate the contract for convenience (which signaled significant governmental culpability in the denouement), and to reschedule the competition for 2009. At that point, it was hoped, the project's requirements difficulties could be resolved.

The ACS debacle is a lesson not just about problems with multi-service procurement programs, but also about strategic priorities. The Army and Navy were very happy to pool their resources to purchase a single set of aircraft, and thus keep the mission out of the hands of the US Air Force, which operates roughly 60 intelligence, surveillance, and reconnaissance (ISR) aircraft that are mostly based on the Boe-ing 707. The problem is that the respective services' purposes for these aircraft were very different. The Navy's airplane was meant to monitor the Chinese and other transoceanic threats; the Army's was to snoop on insurgents and other overland opponents.

The two missions require equipment with sufficiently divergent parameters that designing a single air-craft for both purposes was challenging—at least into the form-factor of the ERJ-145.1 From the start, the Army wanted an aircraft with room for just four workstations in the back; the Navy desired six, since it had a wider range of signals analysis tasks in mind.2 Lockheed’s design team (with rather lacked senior aeronautical engineers) assumed that since the smallish Embraer had room for the seats and consoles, it could take the workstations. The weight of the consoles and their wiring, however, was another matter.3

It is notable that the next leading solution in the competition—Northrop Grumman with General Dy-namics' G450 jet—suffered from the same problem.4 Both bidders had proposed aircraft quite close to the limit of feasibility so as not to propose an aircraft with any greater procurement or operating costs than was necessary. Small wars, that is, generally call for solutions with restrained costs, lest the insur-gents drain the treasury through their asymmetric violence. On the other hand, a larger airplane would make more sense economically if meant for the Chinese threat: the longer inherent range would be use-ful over the Pacific, and the smaller number of individual units would have been acceptable given the geo-graphical specificity of the collection targets. Handling that sort of mission from a single platform calls out for a significantly larger airframe.5

Counterinsurgency, however, could be seen to call for tailor-made ISR solutions, as the targets are not the flotillas of submarines, squadrons of jet fighters, and battalion of tanks for which Cold War ISR sys-tems were intended. Targets as these can be difficult enough to find and track, but big ISR systems like Northrop Grumman's E-8 Joint Surveillance Target Attack Radar System (J-STARS) aircraft showed their value through the results in the 1991 and 2003 campaigns in Iraq.6 The trouble is that they have been markedly less useful for tracking fleeting individuals who dress and act as unremarkably as any passerby, save for the plastic explosives hidden in their belts. Expensive aircraft like the J-STARS and the AWACS are essential assets for the potential-though-grave threats like China, but not the optimal asset for to-day's ongoing fight against Taliban and Ba'athist insurgents.

Is there a general solution for these problems that provides systems for both areas of need? Perhaps. The networked technologies that comprise the elements of today's ISR systems are modular and recombina-tive—modular because the building blocks are easily specified and reproduced, and recombinative be-cause a great deal of the technical progress in military materiel consists of finding new plug-and-play uses

The Crash of the ACS: ISR in Counterinsurgency and the Economics of NCW page 2 of 8

JAMES HASIK • [email protected] • www.jameshasik.com

1John Young, the Navy Department’s Assistant Secretary for Acquisition, noted just two months before the contract cancellation that “it could be very hard to meet all the Navy’s requirements” in a joint program. See ‘Navy May Scrap Ties with Lockheed Spy Plane Project,’ Bloomberg News, 8 November 2005

2 Megan Scully, ‘ACS to Provide Early Intelligence in Future Battles,’ ISR Journal, special conference coverage, 17–18 November 2003

3 Jonathan Karp, ‘As It Adapts to Information Age, Lockheed Fumbles Key Project,’ The Wall Street Journal, 26 January 2006

4 'ACS Reconnaissance Plane: The Kerfuffle Around the Shuffle,' Defense Industry Daily, 12 October 2005

5 Bill Sweetman, 'Swooping to Conquer,' Jane's Navy International, October 2005

6 See Karl Schwartz, 'Alles im Blick,' Flug Revue, July 2003

for existing modules.7 If ISR architectures can take better advantage of modularity, then systems of sys-tems can be more appropriately scaled for the specific level and distribution of threat.

Consider, as an example, how the ACS program could diverge into separate but related programs. Some time ago, the Army could have chosen an even smaller airframe for the ACS, had it not needed to incor-porate the Navy's requirements. Army signals intelligence aircraft fly either immediately behind the troops, or directly over the heads of insurgents (and just above the range ceiling of their shoulder-fired missiles). As the Army's current fleet consists of comparatively shorter-ranged aircraft, a substitute for the ACS might involve signals intelligence payloads on the Army's RQ-5 Hunter drones (built by North-rop Grumman under license from Israeli Aircraft Industries). Indeed, the Army is already considering just such an option. It could even feature signals intelligence roles its new MQ-1C Warrior UAV, a version of the General Atomics' very successful Predator drone. Larger numbers of UAVs could comb signals from the field from many directions, and if one were lost to ground fire, another could be dispatched in short order. Indeed, the ACS was intended from the inception of the program to work with UAVs8, but a more ambitious strategy would entail replacing many of the manned aircraft with unmanned ones.

The Navy seems certain to rely, at least in part, on a larger aircraft. In June 2004, the service announced that it would replace its P-3 Orion maritime patrol aircraft (the basis of its EP-3s) with Boeing P-8s, armed versions of its 737 jetliner. Just 108 of the faster, higher-flying P-8s are now scheduled to replace the nearly twice as many Orions, and maintenance costs are expected to be lower as well because a global support network exists for commercial 737 aircraft. So, only a few weeks after the ACS cancella-tion, Boeing revealed a design for a signals intelligence version of the 737. The proposed EP-8 would fea-ture several advantages: a production line already producing military 737s (for programs like that of the Royal Australian Air Force’s Wedgetail radar surveillance aircraft), a proven electronics architecture, and longer antenna baselines and separations for longer-ranged and more accurate location of enemy trans-mitters. The EP-8 would carry a crew of fourteen, and could accommodate up to twenty-five comforta-bly. The larger airframe, with greater power, could even be used for stand-off communications jamming.9

The question, however, is whether an aircraft like the Predator, which has been so successfully used over Kosovo, Afghanistan, and Iraq, be used by over the open ocean? It could very possibly, as the Predator's maritime version, the Mariner—has been the favored contender in the Navy's Broad Area Maritime Sur-veillance (BAMS) program. The Navy is hoping that more can be achieved with a squadron of drones than could be achieved by a handful of manned aircraft, or by at least several unmanned aircraft controlled by the crew of a single manned one. No single drone could carry all the sensors of an ACS, but the right sensor could be carried at the right time. Further, multiple UAVs flying in dispersed formations could construct very large synthetic apertures. The same aircraft that carry signals intelligence payloads one day could carry magnetic anomaly detectors for hunting submarines the next. The day after, they could swing to the overland role, packing Hellfire missiles and electro-optic sensors for hunting down insurgents, or just signals intelligence pods for tracking their wireless telephone calls.



7 For an early exposition of this concept, see Andrew Latham, ‘Military-Technical Revolution: Implications for the Defence Indus-try,’ Canadian Defence Quarterly, June 1995

8 See Tom Smart, 'Aerial Common Sensor: The Eyes and Ears of the 21st Century Warfighter,' Military Intelligence Professional Bulletin, October–December 1998

9 David Fulghum, ‘737 Spacious, Modern Replacement for EP-3,’ Aviation Week & Space Technology, 25 January 2006

Rather smaller drones could be even more useful in this context: the wing pylons that the P-8 would use for carrying anti-ship cruise missiles could instead, on the EP-8, mount semi-disposable, air-launched jamming and surveillance drones. There is, of course, the added advantage that the Chinese could not nab another aircrew by running a fighter into an unmanned surveillance plane. The downing of an EP-3E Aires II by the Chinese Air Force in April 2001 was offensive, but the capture of the plane's aircrew was par-ticularly galling. Destroyed drones just drop into the ocean and sink.

Better yet—from the Army and Navy's perspective—replacing part of the Air Force's wide-bodied ISR fleet with smaller aircraft could be cost-effective, if done in an intelligently modular fashion.10 The tech-nology may not completely ready today, but the duality of geopolitical interests—the need, that is, to stop both the Chinese Communists and the political Islamists—suggests a direction for applied research funding for the Pentagon. Training drones to fly in formation, avoid basic threats, and cooperate without constant external coaching are already objectives of unmanned aircraft programs. There is yet more to be done, however, if smaller, modular, distributive platforms are to supplement the fewer and less flexible platforms that programs like the ACS represent. Sensors must be taught to recognize and process signals, imagery, radar telemetry more independently, and to coordinate their responses semi-autonomously. Connectivity between ground-based and aerial battle management personnel and their ISR aircraft and distributed ground sensors must, of course, be guaranteed.11 Networks must be designed to be self-healing (rather like that ancient beneficiary of federal applied research, the Internet) so that a swarm of distributed sensors proves more robust in battle than a handful of expensive platforms.

Fortunately, there are things that the Defense Department's procurement authorities can do to system-atically encourage the development of flexible and modular ISR architectures that will be useful both in counterinsurgency work and broader campaigns:

First, invite users and industrialists to collaborate in setting requirements. The arms industry, to be sure, has been the source of many innovations hatched outside the traditional requirements process. The very idea behind the Predator reconnaissance drone originated with Leading Systems, the small company that was bought by General Atomics in 1993.12 The armed forces in the US and many other countries have tradi-tionally not asked industrialists and engineers to participate in defining its concepts of operations. This exacerbates the problem posed by the frequent unfamiliarity with technical state of the art amongst the operational staffs writing the operational concept documents.13 Worse, the design of many new products today is itself a recombinative process: the systems integrator mixes and matches existing modules with some new developments in a sophisticated fashion.14 End users as well should be involved in the devel-



10 Michael Kennedy, Thomas Hamilton, Gregory Hildebrandt, Joel Kvitky, Manbing Sze, and David Vaughan, Common Replacement Asset (CRA) Study Results Briefing (DB-419-AF), RAND, 2003

11 Kennedy, et al., op. cit.

12 Stacey Evers, ‘’Gnat-750 may raise profile of military UAVs,’ Aviation Week & Space Technology, 7 February 1994, p. 54

13 Dave Shingledecker, John Weber, and Mark Klicker, CONOPS Development Process Recommendations, working paper of the 2003 USAF C4ISR Summit: Transforming C4ISR Into Decision Superiority

14 A. Arora, A. Gambardella, and E. Rullani, 'Division of Labor and the Locus of Inventive Activity,' Journal of Management and Gov-ernance, November 1997, pp. 123–140

opment early on, and this does not mean just asking the staff people writing the requirements.15 Fortu-nately, this is changing—modern practice generally expects that software engineers be involved in devel-oping operating concepts for the commercial organizations on whose behalf they work, and much of the functionality of modern weapons is found in their software.16 This does not mean that military services should surrender responsibility for trade analyses to contractors. It merely means that early involvement in requirements development by small, innovative firms with clever operational concepts can enhance operators understanding of the possible rather than merely the traditional. The good news is that emer-gent counterinsurgency requirements (e.g., slat armor for countering rocket-propelled grenades, jamming systems for countering roadside bombs) have relied heavily in the past few years on both views from the field and insights from a wide range of contractors. The bad news is that the consolidation of programs of the past fifteen years has in other domains restricted industrialists' input to those of a handful of firms.

Second, start more programs for niche uses. The late Admiral Art Cebrowski pointed out that long, consoli-dated programs with sole sourcing tend to compel industrial consolidation amongst platform builders. One can either deal with the monopoly, or pay over the top to maintain two sources. Paying for two sources of production, however, is not as useful in the long run as having more competitions for new products.17 Today, this is a potential problem posed by the proposed consolidation of the ACS into the USAF's E-10 program: lumping all the functions of the Guardrail, the Crazy Hawk, the Aries II, the E-3 AWACS, the E-8 Joint STARS, and all the various versions of the RC-135 into a single airplane would provide but a single opportunity to get everything right. Fortunately, increasing modularity in these sys-tems suggests that more programs could be cost-effectively maintained if the aircraft and their payload were smaller—the minimum efficient scale of production could be considerably reduced. More program starts means a greater likelihood of covering niche requirements (like jamming radio bomb triggers) cost-effectively.18 Increasing modularity also suggests a decrease in the ideal size of the firm19, which leads to the next point.

At a minimum, do no harm to small firms. Doing business with the Pentagon is not easy; at times it has been epically difficult. Small firms have a relative disadvantage with respect to these barriers because they lack the scale necessary to employ enough contract administrators and lawyers to navigate the US military's labyrinthine acquisition process. Simplifying the process would encourage more small and commercially oriented firms to do business with the military, but since this advice has been in the literature for dec-



15 Brian Ippolito & Earll Murman, Improving the Software Upgrade Value Stream, Lean Aerospace Initiative working paper #RP01-01, The Massachusetts Institute of Technology, September 2001

16 See International Standards Organization software engineering standard 15288

17 Vago Muradian, 'Do Big U.S. Programs Stifle Innovation?' Defense News, 10 May 2004

18 The Navy is currently undertaking this task with the BAE Systems USQ-113 communications jammer carried by its rather aged fleet of E-6B Prowler aircraft. The Prowler's operating costs are rather greater than those of an aircraft like the Predator. See comments by Captain Chris Field, Commander of the US Navy's Electronic Attack Wing (Naval Air Station Whidbey Island), at the annual convention of the Association of Old Crows, 24 October 2005, Virginia Beach.

19 R.M. Henderson & K.B. Clark, 'Architectural Innovation: the Reconfiguring of Existing Product Technologies and the Failure of Established Firms,' Administrative Science Quarterly, March 1990, pp. 35–50; and R. Sanchez & J. Mahoney, 'Modularity, Flexibility, and Knowledge Management in Product and Organization Design,' Strategic Management Journal, special winter issue 1996, pp. 63–74

ades, there is little reason to say more at this point. It is remarkable that many of the good ideas of late in counterinsurgency material solutions have come from small firms. Finding Serbian tanks tucked into Bosnian barns in the mid-1990s became much safer with the unmanned electro-optics of the Predator, but the team at Leading Systems that originated the idea consisted of just a dozen engineers. The Cougar armored cars that the Marine Corps uses to sweep roads for mines are built by Force Protection, a 300-person firm with a single plant outside Charleston, South Carolina. Emergent threats have tended to loosen the sclerosis in the arteries of funding, but money does not always flow after threats cycle down. However, when properly encouraged, small firms (particularly commercially-oriented ones) may more quickly port commercial technologies into military products, and on development cycles more akin to wireless telephones than that of the Joint Tactical Radio System (JTRS—see below).

Experiment with operationally acceptable gear before making final decisions. Modular product architec-tures—which tend be best supplied by systems integrators and independent suppliers—are well suited to fast-moving industries in which the flexibility with which products and networks can be upgraded is critical.20 Progress with those architectures, however, depends on recombination and experimentation. Outside responses to emergent requirements, the Pentagon's procurement authorities have been mark-edly and institutionally disinterested in experimentation with not-ready-for-prime-time technologies. Changing this, however, would require a fundamental commitment to ongoing experimentation that has not been seen outside the Advanced Concept Technology Demonstration (ACTD) program. In actuality, the ACTD program is not about high science or groundbreaking new component technologies. The Predator drone was developed through the ACTD program, but all its major components were com-posed of relatively off-the-shelf technologies. It was the recombination of these systems in a novel form that produced such low-cost and useful weapon.

Stiff-arm requirements creep within programs. The success of the Predator as a low-cost reconnaissance aircraft depended strongly on the narrow scope of its mission—a host of potential customers wanted a do-everything aircraft, but in the words of one Pentagon official, requirements creep was “successfully stiff-armed.”21 This enabled a wider set of users to field the equipment much faster than would other-wise have been expected. Faster fielding creates more opportunities, sooner, for that critical experimen-tation in the field, and this has an option value all its own. Two retired admirals have agreed on this mat-ter with respect to JTRS: the aforementioned Art Cebrowski and Bill Owens, formerly Vice Chairman of the Joint Chiefs of Staff, and more recently (and pointedly) CEO of Nortel Networks, CEO of Teledesic, and COO of SAIC (of a few companies with some knowledge of telecommunications). Each has argued that the US Army and the USMC could have deployed very capable wireless systems based on rugge-dized, commercial Internet-protocol communicators to troops in Iraq in time for the campaign in 2003. Instead, the troops had to make do with older radios of lesser performance while the JTRS program continued its all-encompassing development process. Three years after the program launched, no equip-ment had reached the troops, and users who urgently needed upgrades were needing to secure waivers



20 M.T. Ulrich, 'The Role of Product Architecture in the Manufacturing Firm,' Research Policy, May 1995, pp. 419–440; and Char-les H. Fine, Clockspeed: Winning Industry Control in the Age of Temporary Advantage (Perseus, 1998)

21 David A. Fulghum & John D. Morrocco, ‘CIA to deploy UAVs to Albania,’ Aviation Week & Space Technology, 31 January 1994, pp. 20-22

through Washington from orders to avoid procurements that did not flow through the JTRS program office.22

Maintain open architectures for affordable interoperability. Firms doing business in network industries face choices as to how interoperable their products should be with those of other firms. In the long run, they tend on their own towards compatibility, because this equilibrium maximizes both industry profit and social welfare. Perhaps surprisingly, it also tends to actually decrease consumer welfare.23 The good news is not just that the customers demand this compatibility within established standards. They are also mo-nopsonists, so they can generally extract the rents that would otherwise accrue to the winning produc-ers. Competitive advantages (vis-à-vis other contractors, that is) can still be secured by influencing the establishment of standards in the firm’s interest, since defense contractors are quite heterogeneous with respect to individual technical capabilities. Boeing, in particular, has been promoting the Network Centric Operations Industry Consortium as a semi-official entity that would help the Pentagon develop common communications and information processing standards for networked C4ISR systems.24 Lockheed Martin initially opposed the idea, openly stating that it could provide Boeing with an undue lead in this field25, but eventually climbed on board for fear of being left out of the process. Whoever sets the standard, however, accessibility for smaller firms must be ensured so that the government can take advantage of the widest range of ideas from industry.

The Global Positioning System has been an excellent example of these dynamics. The requirements of the program were ambitious but simple: as Stanford's Brad Parkinson, the first program manager, put it, to drop five bombs in one whole with a $10,000 guidance system. The earliest GPS receivers were not very impressive, but the emergent equipping of Coalition ground forces in 1991 opened a huge combat laboratory that led to a wide range of improvements and developments. Industry now drives much of the improvement, since its commercial customers arguably have as much interest in accuracy and reliability improvements as military ones do. The best military uses of the technology, that is, were only developed after its civil usage caught on after the 1991 campaign (by which point the electronics had been suffi-ciently miniaturized for commercial applications). Hundreds of companies compete today to to develop the new, new thing in satellite navigation. This is possible because the signal structure of GPS's C/A co-



22 Bob Brewin, ‘JTRS Should Promote Performance: Lack of replacement radios leaves warfighters with old equipment,’ Federal Computer Week, 27 September 2004

23 This is because the network externalities or the value of compatibility is sufficiently important that consumers will pay for it even if they have to pay much higher prices, and because they are have significant lock-in costs that are brand- or equipment-specific. See Oz Shy, The Economics of Network Industries (Cambridge: Cambridge University Press, 2001), chapter 2. The firms' gain is greater than the consumers' loss, and with diverse shareholding prevalent amongst large industrial firms in North Amer-ica and the UK, the consumers could be said to own the firms anyway. The result is thus not that alarming. This argument follows from two different studies that considered the problem even in the absence of externalities: Nicholas Economides, ‘Desirability of compatibility in the absence of network externalities,’ American Economic Review, December 1989, pp. 1165–1181; and Carmen Matutes & Pierre Regibeau, ‘Mix and match: product compatibility without network externalities,’ RAND Journal of Economics, Summer 1988 pp. 221–234

24 ‘Communication consortium by Boeing,’ Aviation Week & Space Technology, 22 June 2003

25 Brad Peniston, ‘Boeing, IBM Join to Bid on U.S. Net-Centric Work,’ Defense News, 20 September 2004

de26 has been publicly available since 1983, when Ronald Reagan decided to provide the service to the world free-of-charge.

Thus, while the challenge to the procurement system is significant, recent history indicates that it can be do well when attention is focused. The Western powers—and particularly the United States—have at their disposal an information systems industry that can handle the challenge if their defense ministries can manage to buy their intelligence systems intelligently. The alternative is a more expensive global cam-paign against terrorists, or a force less ready for deeper threats.



26 This was originally termed 'Coarse Acquisition' when only military applications were considered; today it is at least as often dubbed 'Civil Access'. See Michael Russell Rip and James M. Hasik, The Precision Revolution: GPS and the Future of Aerial Warfighting (Naval Institute Press, 2002)

News & Politics

The Crash of the ACS