2
Naval Vision: Operations and Autonomous Vehicle Applications

This chapter describes the naval vision of operations and the potential for autonomous vehicle (AV) applications within the context of that future vision. The section below summarizes the naval operational environment and vision for the U.S. Navy and Marine Corps. Subsequent sections then address specific naval mission needs and potential applications and limitations of autonomous vehicles.

NAVAL OPERATIONAL ENVIRONMENT AND VISION

Broadly speaking, the role of the Navy and Marine Corps in the U.S. military is to provide credible, sustained combat power from the sea when and where it is needed.1 Many future naval combat operations are likely to be in the littorals, that is, close to shore, in order to project power ashore and to provide an umbrella of defense for forces ashore. In the littorals, naval operations are expected to be contested with mines, diesel submarines, swarms of small boats, and antiship cruise missiles. Marine expeditionary operations can be contested by shore batteries, ground forces, and mines in the surf, on the beach, and inland. In addition, there may be Marine Corps objectives in urban terrain, a complex environment that compounds the difficulties of combat operations and increases risk. If a serious military competitor arises in the future, there is the possibility of war at sea (i.e., opposing naval forces engaging on the high seas).

1  

ADM Vern Clark, USN, Chief of Naval Operations; and Gen Michael W. Hagee, USMC, Commandant of the Marine Corps. 2003. Naval Operating Concept for Joint Operation, U.S. Government Printing Office, Washington, D.C., September 22.



The National Academies | 500 Fifth St. N.W. | Washington, D.C. 20001
Copyright © National Academy of Sciences. All rights reserved.
Terms of Use and Privacy Statement



Below are the first 10 and last 10 pages of uncorrected machine-read text (when available) of this chapter, followed by the top 30 algorithmically extracted key phrases from the chapter as a whole.
Intended to provide our own search engines and external engines with highly rich, chapter-representative searchable text on the opening pages of each chapter. Because it is UNCORRECTED material, please consider the following text as a useful but insufficient proxy for the authoritative book pages.

Do not use for reproduction, copying, pasting, or reading; exclusively for search engines.

OCR for page 20
Autonomous Vehicles in Support of Naval Operations 2 Naval Vision: Operations and Autonomous Vehicle Applications This chapter describes the naval vision of operations and the potential for autonomous vehicle (AV) applications within the context of that future vision. The section below summarizes the naval operational environment and vision for the U.S. Navy and Marine Corps. Subsequent sections then address specific naval mission needs and potential applications and limitations of autonomous vehicles. NAVAL OPERATIONAL ENVIRONMENT AND VISION Broadly speaking, the role of the Navy and Marine Corps in the U.S. military is to provide credible, sustained combat power from the sea when and where it is needed.1 Many future naval combat operations are likely to be in the littorals, that is, close to shore, in order to project power ashore and to provide an umbrella of defense for forces ashore. In the littorals, naval operations are expected to be contested with mines, diesel submarines, swarms of small boats, and antiship cruise missiles. Marine expeditionary operations can be contested by shore batteries, ground forces, and mines in the surf, on the beach, and inland. In addition, there may be Marine Corps objectives in urban terrain, a complex environment that compounds the difficulties of combat operations and increases risk. If a serious military competitor arises in the future, there is the possibility of war at sea (i.e., opposing naval forces engaging on the high seas). 1   ADM Vern Clark, USN, Chief of Naval Operations; and Gen Michael W. Hagee, USMC, Commandant of the Marine Corps. 2003. Naval Operating Concept for Joint Operation, U.S. Government Printing Office, Washington, D.C., September 22.

OCR for page 20
Autonomous Vehicles in Support of Naval Operations Navy Vision and Environment The vision of the Navy’s capstone concept Sea Power 212 is summarized in the concepts of Sea Strike, Sea Shield, and Sea Basing, enabled by FORCEnet, as described briefly below. Sea Strike Sea Strike is a broad concept for projecting precise and persistent offensive power from the sea. According to this concept, networked, autonomous, organic, long-dwell naval sensors, integrated with national and joint systems, will provide persistent intelligence, surveillance, and reconnaissance (ISR), enabling the development of a comprehensive understanding of an adversary’s capabilities and vulnerabilities. Closely integrated with these ISR assets will be the capability to strike time-sensitive and moving targets so as to defeat any plausible enemy force. Sea Shield Sea Shield is the concept focused on the protection of national interests by sea-based defense resources. Traditionally, the Navy has maintained vital sea lines of communication, protected its own offensive forces, and provided strategic deterrence through nuclear-armed submarine patrols. Under Sea Shield, in the future the Navy will also project an umbrella of theater air defense ashore, assist in providing ballistic missile defense for the U.S. homeland and for forces in theater, and extend the security of the United States seaward by detecting and intercepting vessels of hostile intent. Sea Basing As stated in “Sea Power 21,” “As enemy access to weapons of mass destruction grows, and the availability of overseas bases and ports declines, it is compelling both militarily and politically to reduce the vulnerability of U.S. forces through expanded use of secure, mobile, networked sea bases.”3 Sea Basing will support versatile and flexible power projection, enabling forces up to the size of a Marine Expeditionary Brigade (MEB) to move to objectives deep inland. More than a family of platforms afloat, Sea Basing will network platforms among the 2   ADM Vern Clark, USN. 2002. “Sea Power 21: Projecting Decisive Joint Capabilities,” U.S. Naval Institute Proceedings, Vol. 128, No. 10, pp. 32-41. 3   ADM Vern Clark, USN. 2002. “Sea Power 21: Projecting Decisive Joint Capabilities,” U.S. Naval Institute Proceedings, Vol. 128, No. 10, pp. 32-41.

OCR for page 20
Autonomous Vehicles in Support of Naval Operations FIGURE 2.1 Schematic portrayal of the Navy’s vision of FORCEnet. SOURCE: CAPT Robert Tylicki, USN, Deputy Branch Head for FORCEnet Requirements, OPNAV N61FB, “FORCEnet—Making Network Centric Warfare a Reality,” presentation to the committee, January 30, 2003. Expeditionary Strike Group (ESG), the Carrier Strike Group (CSG), the Maritime Prepositioning Force (MPF), the Combat Logistics Force (CLF), and emerging high-speed sealift and literage technologies. It will enable Marine forces to commence sustainable operations and enable the flow of follow-on forces into theater and through the sea base, as well as expediting the reconstitution and redeployment of Marine forces for other missions. FORCEnet FORCEnet is the Chief of Naval Operations’ (CNO’s) vision for enabling network-centric operations for the Navy (see Figure 2.1). According to the CNO, FORCEnet is the “operational construct and architectural framework for naval warfare in the information age, integrating warriors, sensors, command and control, platforms, and weapons into a networked, distributed combat force.”4 While broader in concept than just communications networks, it includes “dynamic, multi-path and survivable networks” among the capabilities to be provided. 4   ADM Vern Clark, USN. 2002. “Sea Power 21: Projecting Decisive Joint Capabilities,” U.S. Naval Institute Proceedings, Vol. 128, No. 10, p. 37.

OCR for page 20
Autonomous Vehicles in Support of Naval Operations Marine Corps Vision and Environment The Marine Corps vision is encapsulated in the concepts of Expeditionary Maneuver Warfare (EMW), which serves as the Marine Corps’s capstone concept for the 21st century.5 EMW includes the following: Operational Maneuver From the Sea (OMFTS),6 the concept for the projection of maritime power ashore, focusing on the operational objective using the sea as a maneuver space and pitting strength against weakness with overwhelming tempo and momentum; and Ship-to-Objective Maneuver (STOM),7 the concept that applies the principles and tactics of maneuver warfare to the littoral battlespace.8 Consistent with the philosophy espoused in Joint Vision 20109 and Joint Vision 2020,10 Marine Expeditionary Forces (MEFs) are envisioned as being rapidly deployable, distributed and networked, taking advantage of information superiority and speed of execution to cut off an adversary’s options. Central to the vision is the ability to deploy STOM directly from ships in the sea base to inland objectives, without the necessity of seizing, defending, and fortifying staging areas on the beach. Expeditionary Maneuver Warfare depends significantly on enhanced joint ISR capability, improved command and control (C2), and new platforms that will enable assault forces to maneuver rapidly to strike an enemy at its weakest points in the battlespace. The concept is likely to require moving a combat-credible ground force several hundred miles inland with great speed and sustaining it there for considerable periods of time. On the ground and in urban terrain, distributed Marine forces will maneuver in a coordinated advance to exploit enemy weaknesses. This will require command and control, assured communications, persistent ISR with supporting firepower, and effective air defense. 5   Gen James L. Jones, USMC, Commandant of the Marine Corps. 2001. Expeditionary Maneuver Warfare: Marine Corps Capstone Concept, Warfighting Development Integration Division, Marine Corps Combat Development Command, Quantico, Va., November 10. 6   Headquarters, U.S. Marine Corps. 1996. Operational Maneuver From the Sea, U.S. Government Printing Office, Washington, D.C., January 4. 7   LtGen Paul K. Van Riper, Commanding General, Marine Corps Combat Development Command. 1997. Ship to Objective Maneuver, Quantico, Va., July 25. 8   ADM Vern Clark, USN, Chief of Naval Operations; and Gen Michael W. Hagee, USMC, Commandant of the Marine Corps. 2003. Naval Operating Concept for Joint Operation, U.S. Government Printing Office, Washington, D.C., September 22. 9   GEN John M. Shalikashvili, USA, Chairman of the Joint Chiefs of Staff. 1996. Joint Vision 2010, U.S. Government Printing Office, Washington, D.C. Available online at <http://www.dtic.mil/jv2010/jvpub.htm>. Accessed on May 13, 2005. 10   GEN Henry H. Shelton, USA, Chairman of the Joint Chiefs of Staff. 2000. Joint Vision 2020, U.S. Government Printing Office, Washington, D.C., June. Available online at <http://www.dtic.mil/jointvision/jv2020.doc>. Last accessed on April 5, 2004.

OCR for page 20
Autonomous Vehicles in Support of Naval Operations NAVAL MISSION NEEDS AND POTENTIAL APPLICATIONS OF AUTONOMOUS VEHICLES Among the world’s naval forces, the U.S. Navy and Marine Corps are second to none, and at present they have no close competitor. However, the current national security environment places increased demands on the Navy and Marine Corps—for presence in a larger number of strategically important geographic locations than in the past, for more rapid and flexible response to emerging crises, and for new capabilities to enable decisive victory over determined adversaries employing asymmetric means. As the post-Cold War era has evolved, it has become increasingly clear that many military legacy systems have minimal or no utility in meeting many emerging challenges. As part of the ongoing transformation of the U.S. military Services, the envisioned Navy and Marine Corps concepts of operation require a transformation of naval forces. And, as discussed below, autonomous vehicles will play a major role in the transformed force. These discussions note that although an AV may have the capability to perform a task, it is not necessarily better than manned systems in performing the task. All else being equal, AVs with a high degree of autonomy can potentially reduce training, support rapid change in tactics (i.e., capitalize more rapidly on the digital battlefield), enable reductions in force personnel, and help reduce the logistics footprint, to name a few advantages. The rest of this section discusses the potential AV applications to meet the needs of Sea Strike, Sea Shield, ground warfare, and other missions. Sea Strike: Needs and Potential Autonomous Vehicle Applications Today the Navy and Marine Corps project power ashore against ground targets by three primary means: (1) manned strike aircraft that are carrier-based or based on large-deck amphibious assault ships (LHDs/LHAs) and that carry precision weapons, (2) cruise missiles launched from surface ships and submarines, and (3) Marine Expeditionary Forces transported ashore from amphibious support ships by manned amphibious landing craft, amphibious assault vehicles, and helicopters. The sea-launched cruise missiles and frequently the air-launched precision weapons are targeted using coordinates obtained by means of electro-optical (EO) or infrared (IR) imagery, often from national intelligence resources. With adequate cueing and good weather, strike aircraft have capable sensors enabling them to detect and identify targets. According to recent studies,11 naval strike capabilities against known, fixed targets are good. However, considerable 11   For example, Task Force on Options for Acquisition of the Advanced Targeting Pod and Advanced Targeting FLIR Pod. 2001. Report of the Defense Science Board Task Force on Options for Acquisition of the Advanced Targeting Pod and Advanced Targeting FLIR Pod (ATP/ATFLIR), Office of the Under Secretary for Acquisition, Technology, and Logistics, Washington, D.C., February.

OCR for page 20
Autonomous Vehicles in Support of Naval Operations improvement is needed in strike capabilities against time-sensitive or moving targets, in large part because of surveillance and targeting limitations, especially in adverse weather. Furthermore, there is only minimal capability for detecting, intercepting, or even reacting to opposing forces employing asymmetric means. Sea Power 21 envisions future naval forces employing autonomous vehicles in a number of Sea Strike mission roles, ranging from surveillance and targeting to weapons delivery. The Navy has been slow to adopt AVs for Sea Strike, and one significant impediment in this regard is that operational forces have been given little opportunity to experiment with AVs and to experience their usefulness. In order to provide operational experience useful for evolving operational concepts and for defining requirements, many currently operational AVs and prototype AVs could be employed by the Navy. Operational experience thus gained can serve to accelerate the introduction of AVs and realize the Sea Strike capabilities envisioned in Sea Power 21. Modeling and simulation, virtual and constructive, could play an important role in this area, particularly for experimenting with less mature systems and concepts. Such experimentation in simulation could accelerate the development of requirements, and could focus and help integrate research and development (R&D) efforts. Surveillance and Targeting Persistent (or continuous) surveillance and targeting of threats constitute an important goal expressed in the DOD’s 2001 Quadrennial Defense Review12 and in the Navy’s Sea Power 21, and the goal is critical to the STOM concept of the Marine Corps as well. The most likely path to the successful achievement of this goal for all Services is to have sensor platforms of various kinds deployed and to integrate them into a network. An effective, persistent surveillance and targeting network is likely to consist of national imaging and electronic intelligence (ELINT) sensors, a constellation of space-based radars, manned aircraft such as the Joint Surveillance Target Attack Radar System (JSTARS) and the U-2, and various kinds of unmanned aerial vehicles (UAVs) and unattended ground sensors (UGSs), depending on the specific mission. In most circumstances, high-altitude and -endurance (HAE) UAVs (such as the Global Hawk) and medium-altitude and -endurance (MAE) UAVs (such as the Predator) can be key contributors to the network. Regardless of whether or not the Navy owns or controls these UAV assets, the data acquired by them are likely to be crucial in developing an accurate and timely picture of the battlespace. 12   Donald H. Rumsfeld, Secretary of Defense. 2001. Quadrennial Defense Review Report, U.S. Government Printing Office, Washington, D.C., September 30. Available online at <http://www.defenselink.mil/pubs/qdr2001.pdf>. Accessed on May 13, 2005.

OCR for page 20
Autonomous Vehicles in Support of Naval Operations As another key part of a persistent surveillance and targeting network, Navy strike groups need to have organic sensor platforms. The reasons for this are as follows: For the foreseeable future, the majority of the U.S. military theater surveillance and targeting capabilities will be supplied by the U.S. Air Force. Whether or not a network exists, ships on the scene have an obligation to the Joint Force Commander to contribute to surveillance and targeting. The proximity of the Expeditionary Strike Groups or Carrier Strike Groups to the target gives them an advantage over other potential sources of surveillance and targeting information (e.g., an ability to perform a given task rapidly). Surveillance and targeting coverage from national and theater sources may be missing under some circumstances, possibly at a time when the need for targeting data is urgent and acute. Such a situation could occur when cloud cover blinds the EO and IR sensors on satellites and high-altitude aircraft. In such circumstances, ship-based organic assets could fly below most cloud cover to obtain the needed information. The organic sensor platform can have special capabilities that other elements of the network lack (e.g., the ability to approach the target closely and view it from many aspects in a timely fashion). To provide surveillance and targeting for Sea Strike, the organic surveillance and targeting platform needs to be capable of being launched from ships in the naval force so that it can be on station in a timely fashion and capable of providing data to enable the detection and identification of ground targets of interest. This platform will require the following capabilities: Range from the force of no less than weapon delivery range, Position accuracy commensurate with that of weapons (certainly no less than that of the Global Positioning System (GPS)), Endurance and survivability for the length of attack (certainly more than several hours), and Minimal risk to U.S. personnel. The detection and identification of targets could be accomplished either by humans using data linked back to a ship or ground station or by the platform itself. Speed, endurance, and survivability requirements for the platform would depend on the specific task to be carried out. UAVs appear to be primary candidates for the organic platform role. A recent Defense Science Board study13 13   Task Force on Options for Acquisition of the Advanced Targeting Pod and Advanced Targeting FLIR Pod. 2001. Report of the Defense Science Board Task Force on Options for Acquisition of the Advanced Targeting Pod and Advanced Targeting FLIR Pod (ATP/ATFLIR), Office of the Under Secretary for Acquisition, Technology, and Logistics, Washington, D.C., February.

OCR for page 20
Autonomous Vehicles in Support of Naval Operations examined various trade-offs in allocating the burden of performance between the targeting system and the weapon to achieve a precision kill. This trade-off will not be considered further here, but it is an area that could benefit from further investigation, especially for UAVs. The following subsections describe the naval forces’ specific Sea Strike surveillance and targeting needs in executing the tasks of naval fire support and deep strike. Surveillance and Targeting for Fire Support. The Navy is currently designing the DD(X) ship, a destroyer, to have a significant capability for naval fire support. The ship’s Advanced Gun System (AGS) will have a higher rate of fire, much greater range, and possibly a larger magazine than that of 5-in. guns on today’s surface combatants. With high accuracy, the AGS will fire a rocket-propelled guided munition to hit targets designated by GPS coordinates. Unless programmatic directions change, it appears that Navy surface ships will have little organic capability to provide target coordinates for the AGS when it becomes operational. As discussed above, for the surveillance and targeting of ground targets, the DOD and the Navy can deploy sensor platforms of various kinds, including HAE and MAE UAVs, and integrate them into a surveillance and targeting network. As also discussed above, in order to provide surveillance and targeting for Sea Strike, a surface ship needs to be capable of launching its own surveillance and targeting platform. For a surface ship with AGS, the platform’s range needs to be at least equal to AGS range and to be able to provide targeting data at a rate enabling all AGS batteries in the force to fire at maximum rate. A vertical-takeoff-and-landing (VTOL) UAV appears to be a strong candidate for this role. Surveillance and Targeting for Ship-to-Objective Maneuver. As part of the Marine Corps STOM concept, in addition to forces crossing a beach in amphibious vehicles, some elements of a Marine Air Ground Task Force (MAGTF) will be transported several hundred miles by VTOL aircraft (such as the MV-22 tilt rotor) and may have to be supported there for extended periods of time. As indicated earlier, the Marine Corps will rely heavily on Navy and Maritime Prepositioning Force (Future) (MPF(F)) ships offshore for fire support, air defense cover, and logistics support for such maneuvers. The Navy currently has no ship-based capability to provide the level of ISR required. Both shore- and ship-based AVs are a likely prospect to meet this need. Surveillance and Targeting for Deep Strike. As discussed above, persistent surveillance and targeting for deep strike are best met with a network of sensors of different kinds, including both joint and organic UAVs. The organic UAV needs to be capable of carrier launch and recovery and to have range equal to that of the task force’s deep-strike weapons, together with significant endurance.

OCR for page 20
Autonomous Vehicles in Support of Naval Operations Weapons Delivery and Suppression of Enemy Air Defenses An uninhabited combat air vehicle (UCAV) is potentially a very attractive means of weapons delivery. It blends the best characteristics of today’s cruise missiles and tactical aircraft precision weapons systems. In just a dozen years since their first use in Operation Desert Storm, cruise missiles have become vital components in the U.S. arsenal. They have been widely used in a number of limited engagements since that time and have been used routinely in significant numbers early in major conflicts (e.g., Operation Iraqi Freedom) before enemy defenses have been eliminated. Cruise missiles are accurate, reliable, and survivable, and they do not put pilots at risk, but since theirs is a one-way mission, they are also a relatively expensive way to put bombs on target. Alternatively, carrier-based tactical aircraft can close to within a few miles of a target area and drop inexpensive GPS-guided munitions or laser-guided bombs onto multiple targets and return to the carrier to reload, but of course these operations put pilots at risk. UCAVs combine the best features of these capabilities. Another potential mission for a UCAV is suppression of enemy air defense (SEAD) (see Figure 2.2). Today this mission is accomplished with manned air- FIGURE 2.2 As illustrated in this schematic, a Joint Unmanned Combat Air System suppression of enemy air defense (SEAD) mission requires coordinated control of multiple unmanned aerial vehicles. SOURCE: Glenn D. Colby, Navy Chief Engineer for Joint Precision Approach and Landing System (JPALS) and Lead for Naval Unmanned Combat Air Vehicles Carrier Air Operations Development, Naval Air Systems Command, “UCAV-N, Naval Unmanned Combat Air Vehicles, Carrier Integration Challenges, Automatic/Autonomous Flight Operations,” presentation to the committee, April 24, 2003.

OCR for page 20
Autonomous Vehicles in Support of Naval Operations craft, but it is dangerous for them, as explained below. Since the Air Force retired the EF-111 from service many years ago, the EA-6B aircraft has been the DOD’s only platform capable of providing electronic jamming support to tactical strike aircraft. As stealth alone is inadequate protection against networked air defenses, jamming support is seen today as a necessary adjunct to all manned deep-strike missions. Though its electronic warfare capabilities are modern, the EA-6B airframe itself is aging, and the Navy is developing a SEAD variant of its F/A-18 E/ F aircraft to replace the EA-6B. But the SEAD mission is a dangerous one for a manned aircraft: the high-power jamming transmitters are an unavoidable liability—multiple ground sites can triangulate and locate the aircraft, and antiradiation missiles can home on the transmitters as on a beacon. A UCAV, perhaps using a data link to a ground station for control of the jamming suite, appears to be a strong candidate for this role in the future. A special potential application of a UAV in the SEAD mission is as a “stand-in,” as opposed to a “stand-off,” jammer. Modern microwave radars are capable of adaptive nulling that can render ineffective the jammers standing off at a distance. To the degree that our adversaries employ this technique in the future, it may be necessary to jam from a vantage closer to the protected forces, and a UAV appears to be the best platform for this mission. Sea Shield: Needs and Potential Autonomous Vehicle Applications The most serious potential threats that the surface Navy and Marine forces may encounter when entering a littoral region are mines in the sea and surf, on the beach, and on land; diesel submarines; swarms of small craft; and antiship missiles (ASMs). Countering Mines, Submarines, and Surface Craft Threats The Navy has begun concept development for a Littoral Combat Ship (LCS), whose purpose is to secure littoral regions for Navy and Marine activities. Navy plans appear to place considerable responsibility on the LCS to protect naval forces from a number of significant asymmetric threats, such as mines, diesel submarines, and swarms of small surface craft. It is planned that the LCS will carry mission-specific modules that can be changed as necessary for various missions, and that those modules will make extensive use of various kinds of AVs. The LCS appears to be the natural home for certain types or classes of UAVs, unmanned surface vehicles (USVs), unmanned undersea vehicles (UUVs), and unmanned ground vehicles (UGVs). To date, the Navy has concentrated on defining the LCS hull, while much remains to be done to define the concepts of operations and systems for executing missions. In particular, it is yet to be determined how the AVs’ sensor systems will detect stealthy submarines and mines;

OCR for page 20
Autonomous Vehicles in Support of Naval Operations how AVs of different kinds, carrying different sensors and weapons, might be networked and controlled to conduct missions; and how the interchangeable mission modules might be designed to house, operate, support, and maintain these AV systems. Countering Mines. Mines are a major impediment to naval forces in the littorals. They are relatively inexpensive and can be widely deployed in the sea, in the surf, on the beach, and on land—and they can be very difficult to detect in any of these environments. The methods employed today for countering mines are cumbersome, very slow, expensive, and inconsistent with the rapid operations envisioned in Joint Vision 2010,14 Joint Vision 2020,15 and Transformation Planning Guidance.16 After decades of alternating periods of very brief emphasis and very long neglect in this area, the Navy now has under way a number of promising developments that together may form the basis for a more effective countermine capability. Since there appears to be no single approach possible for countering mines, a systems approach, such as described below, is the most likely path to success. Under such an approach, before naval forces enter a littoral area, the combatant commander could use a variety of joint ISR assets to support a joint task force (JTF). National imaging and electronic intelligence systems, manned aircraft with a ground moving target indicator (GMTI) capability, HAE UAVs, or the future space-based radar could provide initial indications and warning of preparations for mining activities in coastal areas of interest to naval forces. Surveillance data received from the aforementioned systems would cue clandestine reconnaissance assets in the fleet, but such cues would have to be provided in a timely fashion. Navy attack submarines (i.e., the nuclear-powered submarine (SSN) and nuclear-powered guided-missile submarine (SSGN)) would employ UUVs in the littoral areas of interest to begin bottom mapping as well as identification of the boundaries of mined areas. Of equal importance will be the identification of mine-free areas all the way to the beach. This entire process would be accomplished in advance of the arrival of the JTF. The entire area will be kept under surveillance to prevent additional minelaying. In addition, small UUVs can conduct detailed reconnaissance of surf zones, waterways, and port 14   GEN John M. Shalikashvili, USA, Chairman of the Joint Chiefs of Staff. 1996. Joint Vision 2010, U.S. Government Printing Office, Washington, D.C. Available online at <http://www.dtic.mil/jv2010/jvpub.htm>. Accessed on May 13, 2005. 15   GEN Henry H. Shelton, USA, Chairman of the Joint Chiefs of Staff. 2000. Joint Vision 2020, U.S. Government Printing Office, Washington, D.C., June. Available online at <http://www.dtic.mil/jointvision/jv2020.doc>. Last accessed on April 5, 2004. 16   Donald H. Rumsfeld, Secretary of Defense. 2003. Transformation Planning Guidance, U.S. Government Printing Office, Washington, D.C., April. Available online at <http://www.defenselink.mil/brac/docs/transformationplanningapr03.pdf>. Accessed on May 13, 2005.

OCR for page 20
Autonomous Vehicles in Support of Naval Operations include these: a network of UGSs employing acoustic and visual sensors may be useful for monitoring road traffic, UAVs may be useful for deploying UGSs, and UGVs are expected to be useful in reconnoitering over the hill. Urban Warfare Urban warfare is an especially dangerous environment for the ground warrior, and AVs can and should play a critically important role in reducing U.S. casualties in such operations. UAVs, especially vertical-takeoff-and-landing tactical unmanned aerial vehicles (VTUAVs), can provide persistent overhead surveillance using EO/IR, GMTI, signal intelligence (SIGINT), and chemical/biological/radiological sensors. Additionally, the UAV can serve as a communications relay and a pseudolite (ground-based reference station) for GPS extension into built-up areas. For example, a small UAV propelled by a ducted fan and equipped with a television camera and communications link can provide surveillance of an area by perching atop a selected building and/or looking into windows. In the future, armed UAVs would be useful in an urban environment to provide another vantage for fire support to ground forces. UGVs in a variety of forms can also be particularly useful in urban warfare. Examples are teleoperated machines for looking around corners, semismart small “tanks” containing sensors and weapons, and small devices that a Marine can throw through a window to gather information immediately, implant sensors, or attack the room’s occupants. UGVs may be especially useful underground, in sewer systems or subways. A low-cost, teleoperated UGV for simply drawing enemy fire may also be useful. The Marine Corps and Army have a joint program office in the UGV area in which it is most important that the Marine Corps maintain a high priority. Other Potential Missions The DOD’s missions are changing, and these changes will likely result in new and different missions for the Naval Services and new needs for AVs. For example, HAE UAVs may find application in maritime surveillance for homeland defense and for ballistic missile defense. A specific mission that cuts across applications discussed above is the collection and dissemination of environmental data. Up-to-date information on the weather in denied areas can be very valuable. Thus, any AV that transmits data to a station can enhance its force’s military effectiveness by including weather information in its communication. Instruments to sense environmental conditions are inexpensive and unobtrusive, and communications requirements to transmit the data are minimal.

OCR for page 20
Autonomous Vehicles in Support of Naval Operations TABLE 2.1 Potential Applicability of Autonomous Vehicles in Naval Missions Mission and Task Autonomous Vehicle Applicability UAV USV UUV UGV Providing ground surveillance and targeting •   For nearshore fire support • For support of Ship-to-Objective Maneuver • • • • For deep strike •   For suppression of enemy air defenses • For looking over the hill •   • For urban warfare • • Countering mines in the sea • • •   Countering mines in surf on beach •     • Countering land mines •     • Countering diesel submarines • • •   Countering surface craft • •     Countering antiship cruise missiles • •     Detecting chemical, biological, radiological attack • •   • Providing maritime surveillance •   Environmental monitoring • • • • NOTE: UAV, unmanned aerial vehicle; USV, unmanned surface vehicle; UUV, unmanned undersea vehicle; UGV, unmanned ground vehicle. Little can be said in this unclassified document about AV applications in intelligence operations, but their ability to undertake missions too dangerous or too stressing for humans has made them especially valuable in such roles. Summary of Potential Applications of Autonomous Vehicles Table 2.1 lists likely future naval missions and potential applications of AVs. POTENTIAL AND LIMITATIONS OF AUTONOMOUS VEHICLES Autonomous vehicles exhibit great potential to enhance naval operations, but they are limited by basic physical principles (see Appendix B). It is useful to examine their future potential while at the same time considering their limitations. Factors That Limit Autonomous Vehicles There are several factors that limit AVs from achieving their full capabilities and potential. These factors are discussed in this section.

OCR for page 20
Autonomous Vehicles in Support of Naval Operations Cost AVs show promise for greatly reducing the cost of accomplishing many missions. Because they do not have to provide for the needs of humans for space, life support, and special threat protection (e.g., armor), AVs can often be made much lighter and smaller than they would otherwise have to be. Historically, the cost of a vehicle is roughly proportional to its mass (about $3,300 per kilogram, or $1,500 per pound for a typical military airframe for manned vehicles). Thus, reductions in mass yield substantial savings in procurement costs, often resulting in a proportionate cost savings in the support required for the vehicle. Cost reduction is such an important consideration that it is worthwhile to project the limits of miniaturization of autonomous vehicles. For example, with sufficient miniaturization and cost reduction, the use of expendable AVs might make some missions feasible that would otherwise be too costly. Such trade-offs need to be carefully considered as the technology for AVs advances. Although current experience with AVs does not appear to exhibit the level of cost savings now enjoyed by computers, there is good reason to believe that as AV technologies mature and production levels increase, their costs will follow the well-established trends of manned vehicles (i.e., $1,500 per pound). Onboard Computing The onboard computing capacity of AVs is likely to continue to follow Moore’s law as the commercial technology advances. Intelligence, Surveillance, Reconnaissance, and Targeting Sensors In addition to onboard computing, other aspects of AVs, such as imaging sensors, will presumably also continue a rapid advance driven by other markets. However, such systems often have theoretical limits that will restrict further advances in certain areas. For example, many imaging sensors are now able to detect nearly all of the available light entering the camera aperture, with sensor noise near the lower limits set by physical laws. Thus, sensitivity to light will not increase, but improvements to these imaging sensor systems will come mostly in terms of increased total size of the imaging array. Thus, panoramic-type images can be expected; such images preserve the fine detail needed for sophisticated image interpretation. Advanced sensors for intelligence, surveillance, reconnaissance, and targeting (ISR&T) often require optics that cannot be miniaturized. For example, to recognize faces (~1 cm, or 3/8 in., resolution) in an image taken from 1 km (3,300 ft) away requires a camera aperture of about 10 cm (~4 in.) in diameter. This size is dictated by the wave nature of light and is not subject to miniaturization through the application of advanced technology. Thus, very small AVs, carrying

OCR for page 20
Autonomous Vehicles in Support of Naval Operations small sensors, will need to approach their targets relatively closely in order to get good ISR&T data, while larger AVs might be able to stand off a considerable distance to accomplish the same purpose. Since the size of the camera aperture is proportional to the range to the target (for the same image resolution), to recognize faces from 10 km would require a 1 m (40 in.) aperture. Thus, if good images of the ground or sea surface are required, a high-altitude reconnaissance UAV would have to be a relatively large vehicle to accommodate a camera of the necessary size. A similar problem applies to the acoustic sensors used on underwater vehicles, leading also to the conclusion that the longer-range sensors cannot be accommodated in smaller vehicles. However, the wave effects governing acoustic sensors can be somewhat overcome by synthetic aperture sonar by integrating the signal from a moving sensor to get the same resolution as that from a stationary sensor. In similar fashion, the sensors on UAVs can employ the same techniques as those employed by manned aircraft with synthetic aperture radar (SAR) imaging. However, the application of such phase-coherent techniques to natural visible and infrared (IR) light is so challenging that it will almost certainly not be feasible within the planning horizon of this study, if ever. Another important physical characteristic of atmospheric sensing is the effect of water vapor on high-frequency electromagnetic radiation. In general, clouds are opaque to electromagnetic radiation at optical and infrared frequencies, while radiation at lower frequencies is not impeded. Hence, radar emissions pass readily through clouds, while optical and infrared images do not. Thus, when clouds or fog obscure the ground, high-altitude UAVs with EO, IR, and radar sensors (e.g., the Global Hawk and the Predator) will lose their EO/IR capabilities, but they will maintain their radar sensing, including SAR and moving target indicator (MTI) capabilities. Endurance Characteristically, small vehicles tend to have relatively short ranges and loiter times, whereas larger vehicles can have much greater ranges and loiter times. Physical laws dictate that both aerial and underwater vehicles have approximately the same endurance versus mass relationship. These limitations are exhibited in data for actual vehicles (see Appendix B for a more detailed discussion of the scaling of AVs and a plot of the endurance versus mass relationship). Since a vehicle of a certain size can carry only so much fuel, it can oppose natural winds or ocean currents for a limited amount of time. Most current operational UAVs or UUVs with endurance greater than 24 hours have a gross weight of at least 1 ton. By contrast, small and much lighter-weight aerial or underwater vehicles that can be hand-launched have a maximum endurance of a few hours.

OCR for page 20
Autonomous Vehicles in Support of Naval Operations This limitation, which is primarily the result of basic physics, is not readily amenable to improvement by advanced technology. Partly for this reason, the Defense Advanced Research Projects Agency (DARPA) has been funding the development of small rotorcraft that can “perch and stare,” and so perform long missions without running out of fuel. Vehicle Types There are three broad categories of AVs, which can be characterized by size, that are both feasible and attractive. One type, being very small, has limited sensor resolution and endurance and must get relatively close to the target to perform its mission, as discussed above. Its small size and mass mean that it might be very inexpensive, hand-launched, and “attritable” (expected to survive only a limited number of missions), or even expendable (not retrieved at all). The vehicles in this category can only maneuver for a few hours, but might be able to extend their useful missions by perching, for example, on the local terrain. Such vehicles might be so small and unobtrusive that they would not be very vulnerable, despite the fact that they approach their targets very closely. Current examples of such systems are the Dragon Eye UAV and the Remote Environmental Monitoring Unit System (REMUS) UUV, both of which have demonstrated their operational utility: the Dragon Eye was used by Marines for close-in reconnaissance at the battalion and company levels in the drive to Baghdad during Operation Iraqi Freedom, and REMUS was used to scout the waters of the port of Umm Qasr for mines at the beginning of the same conflict. Another broad class of AV, somewhat larger, has a dry (unfueled) mass in the range of ~100 kg (220 lb) to a few tons. Depending on its payload, such a vehicle can maneuver as much as a day or two without refueling and carry sensors that can obtain superb reconnaissance data without getting very close to the target. Even though they are moderately large, such vehicles may not be very vulnerable because they can loiter far enough away from a target to be hard to detect. Examples of such systems are the Predator UAV, which has been so successful in the conflicts in Afghanistan and Iraq, and UUVs, which can be deployed from a standard 21-in. diameter torpedo tube to perform missions such as mine hunting using high-resolution, side-scan sonar. These vehicles can be much lighter and much less costly than a piloted vehicle, which can perform the same mission, and they can operate far longer than a lone, onboard, human pilot could endure. The last broad category of AV is larger still, with a mass of many tons. Such vehicles are suitable for carrying large payloads (e.g., munitions or heavy, power-hungry sensors) or desirable for having extreme range or endurance. While these vehicles may not be very much lighter or cheaper than manned vehicles intended for the same mission, they can operate in extremely hazardous environments and

OCR for page 20
Autonomous Vehicles in Support of Naval Operations persist for extreme durations (i.e., tens of hours for UAVs). Examples of such vehicles are the Global Hawk UAV; the Navy’s planned uninhabited combat air vehicle (UCAV-N); or a large UUV capable of tracking and trailing a submarine (i.e., following a submarine for many days or weeks). A disadvantage of such large vehicles is that, once again, cost is roughly proportional to mass, so they will be relatively expensive. The foregoing discussion applies mostly to unmanned aerial, undersea, and surface vehicles, with somewhat different physical limitations applying to unmanned ground vehicles. The latter can always simply stop moving in order to reduce or eliminate most power drain. Although “perch and stare” may be developed for rotorcraft UAVs, most aerial vehicles, sea-surface vehicles, and underwater vehicles do not have that option, with the possible exception of a UUV that settles onto the seafloor. However, when UGVs are maneuvering, their energy consumption is not too different from that of the other vehicles. Therefore, it takes a vehicle mass of 1 ton or a few tons to perform sustained maneuvers for more than a day or two, and small vehicles will have maneuvering endurance measured in hours. One approach to improving the endurance of AVs is through on-station or in-flight refueling. This technology will be very beneficial for certain types of missions. However, because there is generally a severe endurance penalty for moving at high speed, most AVs will be designed to move relatively slowly, so as to have as much endurance as possible. As a result, most UAVs will not be able to fly as fast as the stall speed of the current fleet of refueling tankers maintained by the Air Force. Alternatively, UAVs might autonomously refuel UAVs of similar performance. Refueling of UUVs and USVs probably will require them to dock with or be brought onboard mother ships. For any mission class, there is an important trade-off to consider between on-station refueling and just having another similar vehicle replace the exhausted one. Communications In the area of communications, the differences between UAVs, USVs, UUVs, and UGVs begin to emerge strongly. In the air it is relatively easy to communicate along a line of sight, so UAVs can be part of interconnected networks able to relay huge amounts of data. Fortunately, Earth’s atmosphere readily propagates most radio frequencies up to about 100 GHz (gigahertz), even through heavy rain. For example, it is possible, using small (~8 in.) directional antennas, to exchange about 10 Gbps (gigabits per second) between high-altitude UAVs and surface stations, over typical slant ranges, using only 1 W of radiated power. Two UAVs can be separated by 500 km (~300 mi) and still maintain a line of sight above bad weather. Even with these theoretical data rates reduced by a factor of 100 to give an antijamming margin, a network of UAVs can create a densely

OCR for page 20
Autonomous Vehicles in Support of Naval Operations interconnected communications grid that provides service that is the equivalent of high-definition television (HDTV) quality between surface units in the battlespace, the UAV network, and the fleet offshore. A low-bandwidth system of small, omnidirectional antennas, similar to cellular telephone systems, can service requests for access to the high-data-rate network as well as provide limited service to ground units through foliage and other background clutter. While it is difficult to provide reasonable communications to ground units (including UGVs) without aerial relays, such a UAV network can provide high bandwidth to surface units. Most current UUV missions require that the vehicle periodically raise a small device to the surface to get a fix from GPS. At such times the UUVs could tie in to the UAV network to exchange large amounts of data with human operators and offboard automated systems. Communications underwater are extremely challenging. The most capable systems offer only about 10 kbps (kilobits per second), using acoustic communications that have very high signature for detection and localization by the adversary. While the Office of the Secretary of Defense (OSD) is moving aggressively to implement the Global Information Grid (GIG), a well-crafted Transformational Architecture for communications, significant issues remain to be fully addressed in developing the architecture for a workable network of highly mobile nodes. Fundamentally, the issue is to enable each node to maintain efficient routing and connectivity while different vehicles come in and out of local range and view of one another. This problem is one aspect of a larger challenge that includes airspace deconfliction (in a mix with piloted vehicles), AV resource allocation and tasking, the interoperability of different AV systems, and the management and distribution of information with different levels of security classification. While it is apparent that these issues are soluble, they are closely interrelated and need to be aggressively addressed as an integrated set of problems. Endurance and range increase relatively slowly with the mass and cost of AVs (see Appendix B). For example, doubling the range and endurance of a typical AV having a modest fixed payload might be expected to increase its mass and cost by about a factor of 10. It is therefore very desirable to base AVs in support of naval operations on ships. Both the tremendous premium on range and endurance and the tremendous bandwidth of a theater-to-fleet, point-to-point AV communications network strongly favor ship basing for AV communications systems and provide a powerful underlying physical basis for the Navy doctrine of Sea Strike, Sea Shield, and Sea Basing to project sovereign military power. As one example of the “art of the possible” for AVs in support of naval operations: it may be possible to develop a high-altitude, long-endurance, ship-based UAV having modest EO/IR ISR&T sensors and supporting multipoint communications relay that would be less expensive than the cost of shooting it down. As an “antenna farm,” this vehicle would not be particularly stealthy, but shooting down a relatively small U.S. asset at high altitude is intrinsically diffi-

OCR for page 20
Autonomous Vehicles in Support of Naval Operations cult and very dangerous. Such a UAV might have a very small spot factor, or ship deck parking area. It could have a payload bay that supported expensive payloads such as large-aperture optical sensors or weather-penetrating imaging radars, or cheaper payloads such as extra fuel tanks or joint direct attack munitions (JDAMs), which are precision-guided bombs. Carrying such different types of payloads, these UAVs could act as decoys for one another so that an adversary would not know which were the high-value targets. When weather is a problem, these UAVs could drop expendable micro-UAVs (possibly just gliders) to perform final target identification as required for weapons-release authority. Projected Autonomous Vehicle Capabilities The AVs available today are the systems that are actually flying or floating or driving and that can be ordered on the basis of the manufacturing time as the time limit. The AVs that will be available tomorrow are the systems now in development, with their development funding being established, and incorporating technologies available now. The AVs that will be available farther into the future are systems for which technologies have been conceived but must be developed and then incorporated into the systems. The Navy can speed up its acquisition and use of AVs by accelerating its procurement of some of today’s AVs (those of its own choice) while taking maximal advantage of existing AVs and AV developments of the other military Services. This latter effort might take the form of assigning Navy personnel to joint programs in which they would gain experience as operators and planners for the UAVs of other Services (e.g., Global Hawk and Predator with the Air Force) in field operations for tasking and data exploitation. In some cases the Naval Services do not need their own organic AVs because capabilities exist in other national and theater systems. In other cases, as exemplified below, the Naval Services clearly do need their own organic AVs and should not be forced to violate sound system design principles (e.g., tight feedback control) by relying on other Services for a core capability. The Navy has some unique requirements for AVs. UUVs are the obvious example of such naval-specific needs, but its requirements regarding UAVs are almost as unique, and need to be addressed directly in development. While the other Services have moved ahead in developing the basic elements of AV technologies, the Navy can benefit from these developments, adding to them unique naval requirements, including the capability of deck takeoff and landing, deck handling and operations, minimization of deck spot sizes, and attending to the premium for antenna real estate aboard ships. UAVs designed for land-based operation will not generally meet these requirements, and the expense and time spent attempting to modify and adapt them to shipboard application may not be justified or successful. Conversely, UAVs to be based on ships (carriers, destroy-

OCR for page 20
Autonomous Vehicles in Support of Naval Operations ers, amphibious ships, and the Littoral Combat Ship) need to be defined and procured for persistent ISR, SEAD, strike, communications relay, and so on. The Joint Unmanned Combat Air System under advanced development by DARPA, the Air Force, and the Navy addresses some but not all of these issues. It is clear that a single vehicle cannot satisfy all of the requirements. Where appropriate, the Navy can benefit from other Services’ developmental expenditures and lessons learned, reducing the cost of fielding its own UAVs. To make this process effective, the closest attention must be paid to proper system engineering based on the “whole” problem (including concept of operations, airspace deconfliction in mixed airspace, mobile networking and interoperability, launch and recovery, and staffing and logistics support). Combinations of Short-, Medium-, and Long-Range Sensing As previously mentioned, for ISR&T imaging, image quality and coverage are determined by a relatively simple set of physical and optical rules. The farther one wishes to see, the larger the aperture (typically the lens diameter) required. This general principle applies and scales to all wavelengths, including imaging radars (which also have a power component in the scaling parameters). For the detection and classification of a target, a minimum amount of information is required, which in turn places requirements on the imaging system. For example, it is generally accepted that to classify a target (e.g., to determine whether a vehicle on the ground is wheeled or tracked), an image must have 16 resolution elements (pixels) across the narrowest dimension of the target. This image quality typically provides about 90 percent correct classification. From high altitude, a high-quality, large-aperture imaging system is required. From a lower altitude, smaller apertures, and therefore a lighter-weight and lower-cost system, can be used. One effective operational combination is for ISR&T imaging to have a high-altitude detector/classifier cue a lower-altitude “examiner” to perform recognition and possibly identification. Since some lower-resolution radars can image through clouds whereas higher-resolution optical imagers do best in clearer air, this combination can be very effective. In practice this system solution requires that the lower-altitude UAV be able to get to the indicated location quickly, which can be a problem for low and slow ISR UAVs if they are not deployed nearby. As previously mentioned, one apparently attractive option is to have a larger, high-altitude ISR&T UAV deploy a small, expendable, low-altitude UAV to get close-up images, even in clouds or fog, for final target confirmation and human weapons-release authority as required by normal rules of engagement. Such systems have been successfully used in experiments in which the small “Finder” UAV developed by the Naval Research Laboratory was deployed from a long-endurance Predator UAV.

OCR for page 20
Autonomous Vehicles in Support of Naval Operations Sensing includes nonimaging methods such as electromagnetic sensing (EMS), SIGINT, and other means of cueing for imaging systems. One example is a payload that includes EMS (for detecting the carrier signal of a threat radar) that cues the imager to look in the source direction and send back the image and the location of a threat radar. In summary, higher-quality, broader-coverage imagery requires larger aircraft at higher altitudes, while smaller imaging systems can be used at lower altitudes to give extremely high resolution over small areas. In UUV optical imaging applications, short ranges are normal, being limited by the low optical transparency of the water. To remedy this limitation, imaging sonars are often used; with these as with other sensors and for the same physical reasons (wave properties), larger apertures are required to give acceptable resolution. Manned and Unmanned Systems Working Together While AVs are valuable as independent mission assets, one of the most promising modes of operation for AVs is to enhance manned mission capability by operating as wingman or adjunct. The Army, through the Future Combat System (FCS) program, has a doctrine indicating that increased awareness of enemy positions and numbers (information dominance) will allow the defeat of an enemy that has more units and heavier armor. According to this doctrine, FCS forces fire from cover, avoid battle, bypass enemy forces, and avoid ambush even as the FCS forces are just beginning to build up. The information dominance enabling such actions will be provided by UAVs and UGVs working in concert with manned systems. There are few or no well-thought-out concepts of operations for mixed manned and AV operations. Constructive, live, and virtual simulation could play an important role here. The Marine Corps can take full advantage of these developments where appropriate. It is noted that the Marine Corps has been one of the most visionary organizations with regard to AVs, supporting R&D efforts going back more than 20 years. The Marines developed the hand-launched UAV Dragon Eye and the small, teleoperated UGV Dragon Runner that have yielded important experience and lessons learned in Operation Iraqi Freedom. Defeating the larger enemy while using assets that can be quickly inserted by a few C-130s (as envisioned by FCS) means optimizing each component, including armored vehicle size, the number of personnel, and the mix of AVs. This same logic can be applied to the provision of forces and components arriving by ship and to the reviewing and optimizing of the mix of assets, with AVs included. It needs to be noted that this report does not address in any great detail the following questions: (1) How does the performance of an autonomous vehicle system degrade owing to communications bandwidth and latency if a human crew is put offboard? (2) Can fewer humans be used offboard than would be used

OCR for page 20
Autonomous Vehicles in Support of Naval Operations for the same number of manned aircraft (e.g., can a single, human crew “pilot” multiple AVs and still be highly effective)? Briefly, the answers to these questions are that the bandwidth currently needed to provide the remote human operator with the necessary situation awareness is very large. However, providing this bandwidth is actually not difficult based on physical law, as discussed in Appendix B. Automated systems can handle aircraft controls and communications latencies at least as well as humans, so it is possible to make UAVs and UGVs (communicating through a UAV network) highly capable by putting the crew offboard. The communications issues that constrain interoperability between AV systems are discussed in Chapter 7 in the subsection entitled “Communications Issues as Constraining Factors on Interoperability.” What is not straightforward is the question of having a single crew control many vehicles simultaneously—because during periods of peak operational tempo, the performance of all such systems are limited primarily by the ability of the crew to sense and assimilate information, even if all of the necessary information is delivered from the AV to the remote crew. Huge advances in computing and algorithms will be required to change this latter fact. An alternative approach to having a single crew pilot multiple vehicles is to time-share less than one crew per vehicle, allocating them only to those vehicles where the operational tempo is greatest and accepting the losses that result. This approach may be cost-efficient if the costs of the AVs can be made relatively low compared with the operational costs of maintaining a large number of crews. There is little specific work on how operator performance degrades with increased operations tempo and mission complexity, but psychological research suggests limits to performance, and more such research is required. The operator-to-vehicle ratio as a function of operations tempo and mission complexity is not known, but estimates for UGVs range from 2:1 to 1:5. Refining operator-to-vehicle ratio as a function of operations tempo and mission is important, and more research is needed.18 18   Private communications with Clinton W. Kelly, Science Applications International Corporation, May 2004.