The Department of Defense's planned unmanned aerial vehicles, if fully developed and deployed, hold great potential for expanding the use of unmanned aircraft to collect intelligence. As noted in Chapter I , however, the programs to develop those UAVs involve various overlaps and problems.

The Congressional Budget Office has constructed five illustrative options to address some of the concerns that have been raised about those programs or to take greater advantage of the promise that UAVs appear to hold. All would use UAVs differently from the way DoD currently plans. However, the options by no means represent all of the possibilities for improving DoD's UAV programs. Nor are they motivated primarily by the desire to save money.

Option I focuses on DoD's highest priority for unmanned aerial vehicles: providing a UAV capability to the Army's brigade commanders. Within that option, the first alternative (Option IA) would provide that capability much more quickly than now planned by cancelling the Outrider tactical UAV and having brigade commanders use the existing Hunter systems instead. In addition, the Navy and Marine Corps would receive a new UAV more suitable for their maritime operations. The second alternative (Option IB) would keep Outrider but make it solely an Army system; the Navy and Marine Corps would continue to rely on their Pioneer tactical UAVs. Option II proposes a possible solution to the Army's concern that during a regional conflict its corps and division commanders would not receive support from Predator UAVs controlled by the Air Force. Specifically, Option II would give those commanders their own Hunters instead. Options III and IV examine the potential for trading off unmanned aerial vehicles for other aircraft: tilt-rotor UAVs for the Army's new Comanche reconnaissance helicopter (Option III), and Global Hawks for the Joint Surveillance Target Attack Radar System (Option IV), which the Quadrennial Defense Review has proposed reducing in number. Last, Option V addresses some Congressional concerns about overlaps in the UAV programs by cancelling Darkstar after it completes its Advanced Concept Technology Demonstration.

Many of the UAVs included in those options are still in the ACTD development phase, and DoD has not yet committed to buying them in quantity. As a result, the options cannot be compared with an overall DoD plan for unmanned aerial vehicles. Each option illustrates one possible alternative to one component of the department's UAV programs. Thus, it can only be compared with the plan for that particular option or the likely DoD plan once the UAVs finish development (assuming that the UAVs now under development--Outrider, Global Hawk, and Darkstar--are all ultimately successful and that the services buy and deploy them). Compared with those plans, two of the options would save money, two would cost money, and one would almost break even, based on both acquisition costs and operating and support costs over the assumed 15-year life of a UAV (see Table 5).
 

OPTION I: CANCEL OUTRIDER OR MAKE IT SOLELY AN ARMY SYSTEM

The Army has long had a need for a tactical UAV to support brigade commanders in combat operations. Indeed, the Joint Requirements Oversight Council considers the fielding of such a system its highest priority in the area of surveillance and reconnaissance. DoD would like to acquire 62 Outrider UAV systems, with the Army to get 38, the Navy nine, the Marine Corps 11, and four to be reserved for training. (The Navy systems are intended to have eight air vehicles each, whereas the Army and Marine systems would have four.) The Army is expected to procure one UAV system for each of its maneuver brigades--or three per division. A few divisions, the light ones, would receive four Outrider systems apiece. CBO estimates that acquiring those UAVs would cost about $860 million (in 1998 dollars), and operating and supporting them would cost an additional $930 million over 15 years.

Meeting DoD's goal has been difficult, however. Technical problems have delayed the Outrider development program, setting back the schedule for testing and producing the UAV. In addition, it is not clear that Outrider can satisfy the Navy's requirements for a tactical UAV, which include a vertical take-off and landing capability and a heavy-fuel engine to make shipboard operations easier. (The Navy is currently using Pioneer as a tactical UAV system that can be deployed on ships to support naval and littoral operations, but it had planned to phase out Pioneer in favor of Outrider.)

To address those problems with the Outrider program, CBO examined two alternative approaches. The first (Option IA) would cancel Outrider; instead, the Army would use Hunter for its brigade commanders, and the Navy and Marine Corps would procure a new tactical UAV. The second (Option IB) would keep Outrider but tailor it specifically to the Army. The Navy and Marine Corps would continue to use Pioneer.

Option IA: Use Hunter to Meet Army Brigade Requirements and Buy Other UAVs for the Navy and Marine Corps

This alternative would quickly give the Army a UAV to support its brigade commanders by cancelling the Outrider program and taking full advantage of a system that already exists: Hunter. Many Hunter air vehicles and their support equipment were purchased by the Army during the mid-1990s and are still in storage. At the same time, the Navy and Marine Corps would buy a new UAV with vertical take-off and landing capability to better meet their unique requirements. (The Navy would acquire nine of the new UAV systems and the Marines 11--the same numbers as planned for Outrider. Two additional systems would be purchased for training.) Compared with buying and operating 62 Outrider systems, Option IA would save $80 million in acquisition costs but add about $90 million to total operating and support costs (see Table 5).

Meeting the Army's Brigade Requirements with Hunter. The Army owns 56 Hunter air vehicles and 28 ground control stations as well as other support equipment. They were purchased in seven systems, with eight air vehicles each, as part of the low-rate production contract in 1994 and 1995. Currently, one Hunter system is stationed at Ft. Hood in Texas to support Task Force XXI--the Army's new digitized warfighting brigade. A second system (with four air vehicles) is stationed at Fort Huachuca in Arizona for training purposes. The other systems are in storage.

Equipping the Army's maneuver brigades with Hunter would require having an additional 18 air vehicles and 33 mobile UAV support team (MUST) packages, which consist of a modern ground control station and a smaller ground data terminal for immediate deployment. This alternative would also buy 82 air vehicles to allow for peacetime attrition (see Box 1). Those extra purchases--100 air vehicles and 33 MUST packages--are in addition to the Hunter equipment already in the Army's inventory. The current ground support equipment would be deployed with the division headquarters, and the new MUST packages would be assigned to the brigades.
 

This alternative would take advantage of what has been called the "pass-forward" method of operating and supporting unmanned aerial vehicles. As would be the case with Outrider, a Hunter system would be attached to a division's military intelligence battalion and would comprise six air vehicles, the principal ground control stations, and most other support equipment. Divisions are better able to support the logistics required by a Hunter system than individual brigades are. However, each brigade would have a MUST package with which it could take direct control of a Hunter air vehicle. The air vehicle would launch from the rear near the division headquarters and would be "passed forward" to the modern ground control station with the brigade. The brigade commander could then use the air vehicle for up to 10 hours before it needed to land for refueling and other support. He would have direct control of the air vehicle until he passed it back to the military intelligence battalion of the division, where its primary support base would be located.

Time to Full Operational Capability. Substituting Hunter for Outrider could shave several years off the time needed to give brigades their own UAV capability (see Table 6). According to the current schedule, Outrider should be deployed to all of the Army's maneuver brigades by 2004. However, it has experienced a number of development problems that the contractor may or may not be able to resolve. History suggests that developing UAVs involves a learning curve, and there may be more bumps in the road before Outrider is a mature system.
 

Hunter, by contrast, is already a mature, reliable system. The Army has used it successfully in various training exercises and in developing concepts of operation for tactical UAVs. The existing Hunter systems could be brought out of storage and deployed in much less time than waiting for Outrider. Those systems would need some upgrades and improvements, but that would not take long. The real bottleneck would be an insufficient number of trained crews. Currently, the Army can train two crews a year. It could expand that, but at some additional expense.

Capabilities. Hunter provides equal or greater capability to brigade commanders than Outrider would in four areas: dash speed, endurance, coverage, and payload. Speed is arguably the most important capability that a brigade commander wants in an air vehicle. Dash speed (the maximum speed at which a vehicle can travel for short periods) represents the responsiveness of the UAV assigned to a particular unit and determines how quickly it can provide intelligence at the right time and in the right place. Outrider and Hunter have essentially the same dash speed--around 200 kilometers per hour.

Endurance is another measure of capability in which this option exceeds the Army's plan. Based on the number of hours that the UAVs can operate at 200 kilometers (Outrider's intended radius), Hunter's endurance is more than twice that projected for Outrider--eight hours versus three hours. Thus, a brigade commander could study, track, or cover a particular target or area for a much longer time with Hunter than with Outrider. Another way of looking at that measure is the number of sorties required to provide continuous 24-hour imagery collection for each brigade in a division. The Army's likely plan for using Outrider would require eight sorties per brigade to provide nearly 24-hour operations. Option IA would require three Hunter sorties for the same capability.

Coverage is the amount of area a UAV system can cover with its sensor. That area depends on the air vehicle's altitude, camera, and endurance and on the number of vehicles that can be simultaneously controlled in the air by the available ground stations. This option provides a division's brigades with greater coverage than the expected Army plan--106 square kilometers per hour versus 61--largely because of the Hunter system's greater endurance and larger number of vehicles that can be controlled in the air.

Finally, Hunter's potential payload, 200 pounds, is more than three times that expected for Outrider. Hunter carries a similar sensor but also has room to carry additional payloads that are now being tested or developed, such as a laser to guide precision munitions to their targets. Outrider may be able to carry other payloads as well, but with a smaller air frame, it is not likely to do so as easily as Hunter.

Deployability. Critics of using Hunter to support brigade requirements have pointed to the system's large "logistics footprint" and deployment requirements. A full Hunter system of eight air vehicles and four ground stations originally required five-ton trucks and numerous sorties by C-130 aircraft to deploy it. By contrast, an Outrider system is supposed to be deployable in a single C-130 sortie (although in reality it will require slightly more than that).

The Hunter system that this option envisions for a division's brigades--six air vehicles, three ground control stations, accompanying shelters, data terminals, and three MUST packages--would take about eight C-130 sorties to transport to a theater of operations. Deploying a division's three Outrider systems (one for each of the maneuver brigades) is expected to require at least four C-130 sorties. Thus, the deployment requirements to support the three brigades of a division would be twice as great under this option as with the planned Outrider systems.

Overall Assessment. The option of using the existing Hunters in the Army's inventory, as well as purchasing a few more, would yield significantly greater UAV capability for brigade commanders faster and at a lower cost than the Army's likely plan. The chief disadvantage of this option is its greater deployability requirements. In addition, replacing air vehicles lost to attrition would be much more expensive because Hunter is a more capable platform. Overall, including those "attrition spares," Hunter would cost less to acquire but more to support than Outrider.

Meeting the Navy's and Marine Corps's UAV Requirements. Option IA would meet Navy and Marine requirements by purchasing a UAV with vertical take-off and landing capability and a heavy-fuel engine. The Navy has been examining several such systems for possible purchase. A typical example is Guardian, built by Canadier. The company has been flying a smaller-scale demonstrator model of Guardian for at least 10 years. The full-production model employs many of the same components as the demonstrator but has a larger air frame and rotors.

With a greater payload and range, such a UAV appears to be much better suited to meet the Navy's and Marine Corps's needs than Outrider, principally because of its vertical take-off and landing capability and its heavy-fuel engine. The principal disadvantages of such systems are that they have not yet had much flight-testing, and they are likely to be more expensive than Outrider.

Time to Full Operational Capability. In terms of development time, there is little to distinguish between this option and DoD's plan (see Table 7). The time needed to finish modifying a demonstrator model to bring it up to production specifications, integrate it with the tactical control system, and begin production is around two and a half years, at a minimum. Outrider is supposed to complete its development process and begin low-rate production in fiscal year 1999, which would give it the edge in development time, but its past problems lend little confidence to predicted schedules. Of course, a new UAV for the Navy and Marine Corps could experience delays as well. In addition, the services would then require several more years to buy and deploy the systems in the field.
 

Capabilities. With the exception of its ability to take off and land vertically, the capabilities of a typical maritime UAV are fairly similar to those of Outrider. CBO assumed that each naval task force or amphibious ready group would receive one UAV system--composed of a ground control station, four air vehicles for a Marine system or eight air vehicles for a Navy system, and associated data terminals and ground support equipment. Outrider's dash speed is expected to be faster than that of a typical UAV with vertical take-off and landing: around 200 kilometers per hour versus 160. (Some systems under development that have vertical take-off and landing could fly much faster than that, but they would probably be more expensive than the option CBO considered.) However, Outrider's endurance is expected to be slightly less; thus, eight Outrider sorties would be required for 24-hour operations, compared with about seven sorties for a UAV with vertical take-off and landing. CBO could not compare the coverage of the two systems because data for the prospective maritime UAV system were not available.

A significant area of difference between the two systems is likely to be their payloads. Whereas Outrider is expected to carry a 65-pound payload, a typical maritime UAV with vertical take-off and landing can carry 110 pounds. That gives it a somewhat greater potential to handle the type and quantity of payloads the nautical services may want their UAVs to carry.

Deployability. The deployability of either system is not an issue if the UAVs are stationed on board ships before a task force or amphibious ready group leaves port. Marine systems might need to be airlifted to an overseas location, however. Both an Outrider and a naval UAV system are expected to fit on a C-130 and would require at most two sorties for delivery to a theater.

Overall Assessment. The main concern about Outrider is its suitability for ship and fleet operations. The chief advantage of this option is that it provides an unmanned aerial vehicle system that is highly suitable to shipboard operations, primarily because of the UAV's vertical take-off and landing capability and heavy-fuel engine. Otherwise, the capabilities of the two systems are fairly similar. Outrider will almost certainly be the cheaper system, but money saved by using Hunter to fulfill the Army's brigade UAV requirements could offset the additional costs of a naval UAV.

Option IB: Buy Outrider Only for the Army

This alternative addresses the same issues as Option IA but in a way that favors the Army's UAV needs over those of the Navy and Marine Corps. It proposes buying the Outrider tactical UAV for Army use, while the Navy and the Marine Corps would continue relying on their Pioneer systems. The upgrades currently planned for Pioneer would continue, and replacements for air vehicles lost through attrition would be purchased as needed to maintain the existing systems.

Some of the problems that have confronted the Outrider program stem from the Navy's requirements for the UAV, which include a 200-kilometer range, a heavy-fuel engine, and the integration of extra components for shipboard operations.⁽¹⁾ Eliminating those requirements would leave a system capable of fulfilling most Army requirements. In fact, one high-ranking Army official has stated that Outrider--even with its problems--appears able to meet the Army's needs and should therefore be procured.

In terms of the Army, this option would be the same as the service's likely plan for Outrider. Thus, all of the relevant operational factors--time to full operational capability, deployability, and capabilities--would be the same as under that plan. For the Navy and Marine Corps, this option would represent no change from their current UAV situation. Pioneer is roughly as capable a system as Outrider, except that it is over 10 years old, requires a great deal of maintenance, and is more difficult to work with. Of course, there is no guarantee that Outrider would be able to overcome its development problems and emerge as a useful system for shipboard operations. This option would spend less on acquisition than either Option IA or DoD's plan. However, those savings would be partially offset by higher operating and support costs over 15 years (compared with DoD's plan). As a result, net savings from Option IB would be around $140 million compared with DoD's plan and $150 million compared with Option IA. In spite of those overall savings, the Army's Outrider systems would have a higher cost per system (unit cost) because this option would buy fewer of them.
 

OPTION II: USE HUNTER TO MEET THE ARMY'S DIVISION AND CORPS UAV REQUIREMENTS

Option II attempts to avoid the problems that could arise if the Army relied on Predator unmanned aerial vehicles controlled by the Air Force to meet its division and corps UAV requirements. With Hunter terminated, the Army proposes to rely on Predator--an Air Force system--to handle UAV missions at the division and corps level. However, the Air Force plans to buy just 12 Predator systems (each with four air vehicles and one ground control station) and to deploy only five of them to a regional conflict. The Air Force has stated that although it is willing to use Predator to support division and corps commanders, there may be higher priorities set by the theater commander or the national command authority that could require most, if not all, of the Predator assets. If two corps and seven divisions deployed to a regional conflict--as happened during the Gulf War--it seems unlikely that the average division commander would get a prompt response to his request for a Predator to perform a reconnaissance mission.

One possible solution to that problem is to provide each division and corps with its own UAV capability using the Hunter systems that the Army has in storage. This option would give a Hunter system with six air vehicles and three ground control stations to each corps and a Hunter system with four air vehicles and two ground control stations to each division--for a total of 64 air vehicles and 32 ground control stations. The Army already owns 56 air vehicles and 28 ground stations, so this option would require only a small purchase to fill out the force and to provide some extra systems for training and for replacing those lost through attrition.

Those Hunters are the same ones that would be given to the Army's brigade commanders under Option IA. As a result, Options IA and II could not be pursued simultaneously without buying substantially more Hunter systems.

Costs

Relying on the Air Force's Predators to provide imagery to corps and division commanders would cost the Army nothing. The purchase of those UAVs is already planned and included in the Air Force's budget, so giving them an extra mission to perform during a regional conflict essentially entails no added costs.

This option, by contrast, would increase the Army's costs. It would require a total of 72 air vehicles and 36 ground control stations. Bringing the existing Hunter systems out of storage would cost little, but buying 16 additional air vehicles and eight ground stations (plus attrition spares)--as this option envisions--would mean an additional $250 million in acquisition costs and about $500 million in operating and support costs over 15 years for all systems.

If the alternative were for the Army to procure its own Predators--something it has no plans to do--this option would save money in comparison. Buying a Predator system for every corps and division (14 additional systems in all) and operating them for 15 years would cost about $1.7 billion.

Capabilities

Both the Army plan and Option II have various strengths and weaknesses in the area of capability. The most pronounced strength of the Army's plan to use Predator is that system's 24-hour endurance. Only one Predator sortie is necessary to provide 24-hour coverage of a particular area or target. However, should a division commander ever need such lengthy coverage, he would probably be unlikely to get one of the few available Predators assigned to him for that long. By comparison, a Hunter system attached to a division would need to use three sorties for 24-hour operations (see Table 8).
 

The most pronounced strength of Option II is Hunter's dash speed. The Hunter systems assigned to a corps or a division could fly at almost 200 kilometers per hour, compared with 130 kilometers per hour for Predator. Furthermore, corps and division commanders would have more than one Hunter air vehicle at their disposal; thus, in a sense, the flexibility of their system would be multiplied by the number of vehicles they could put into the air simultaneously. Even if such commanders could automatically receive a Predator from the Air Force, they would still have much greater flexibility with their own Hunter system. Overall, those capabilities make Hunter more responsive to the immediate needs of its users.

In terms of payload, although a Predator air vehicle has much greater capacity than a Hunter, the total amount of payload available to unit commanders under the two alternatives would be more comparable. Under the Army plan, a corps or division commander would control a single Predator air vehicle and thus could use payloads of up to 450 pounds. But under this option, a corps commander would potentially have three air vehicles, each carrying 200 pounds, at his disposal. A division commander would have two air vehicles at 200 pounds each. Thus, although corps or division commanders could take advantage of a heavier payload under the Army's plan, under this option they would have the flexibility to put different payloads on different air vehicles. For example, because a corps commander could simultaneously control three air vehicles, one could contain a conventional imagery payload, another could have a laser designator, and a third could carry a countermine payload.

Deployability

The UAV systems under the Army's plan and this option are roughly comparable in terms of deployability. One Predator system requires five C-130 sorties to deploy. A Hunter system of six air vehicles requires about eight C-130 sorties, and one with four air vehicles needs five (see Table 8). Predator deployments are often described using a different measure, however: the number of C-141 aircraft. A Predator system requires two such sorties for deployment, whereas a Hunter system with six air vehicles would require about 3½ C-141 sorties, and a system with four would need less than 2½ sorties.

Delivering five Predator systems to support a major regional conflict, as the Air Force plans, would require 10 C-141 sorties. If the Army deployed two corps and seven divisions to the theater, as it did during the Gulf War, then the Hunter systems accompanying those units would need another 23 C-141 sorties. Those sorties would have to be counted as additional to the ones required for Predator because the Predator systems would almost certainly be deployed to the theater even if the Army had its own UAV capability for its corps and divisions.

Overall Assessment

Compared with the Army's plan, the most significant benefit of this option to corps or division commanders would be having their own unmanned aerial vehicle capability. They would not have to wait in line for a Predator to become available. Instead, with a Hunter system already available, they would be able to see what they wanted to see as soon as the air vehicle could reach the target area.

The chief disadvantages of this option relative to the Army's plan are the cost and the substantial additional requirements for deployment. Regardless of whether the Army began using Hunter, the Air Force would still want to procure Predator. Thus, the Hunter systems would be added to the equipment that would need to be delivered to a major regional conflict, requiring an extra 23 flights by C-141 aircraft. This option would also be more expensive for the Army. Even though many of the Hunter air vehicles and ground stations are already in the Army's inventory, pulling them out, making them ready for combat, and buying several more would entail costs to the service that the Air Force's procurement of Predator would not.
 

OPTION III: BUY TILT-ROTOR UAVs AND REDUCE THE ARMY'S PLANNED COMANCHE HELICOPTER FORCE

Many people would agree that unmanned aerial vehicles promise to enhance the fighting potential of U.S. forces on the battlefield by giving commanders immediate information about the disposition of enemy troops. After the Army's warfighting experiments in 1996 at the National Training Center, the Army Vice Chief of Staff described tactical UAVs as "major combat multiplier[s]" for a brigade commander.⁽²⁾ But compared with the combat power of an aircraft, for example, it is difficult to quantify the role that reconnaissance plays in warfare. Doubling the number of missiles that an aircraft can carry will double its combat power and, thus, the number of potential targets it can attack. But introducing new or better reconnaissance systems, although almost certainly making military forces better off, is not such a straightforward combat multiplier.

This option illustrates one way in which the military could make even more use of UAVs than it now plans. The option assumes that having many UAVs operating on a battlefield is indeed a major combat multiplier--in other words, that their reconnaissance capability will enhance the fighting power of U.S. forces. Consequently, this option would substitute Eagle-Eye--a tilt-rotor UAV developed by Bell Helicopter that the Navy has been examining--for some of the Comanche reconnaissance helicopters that the Army plans to buy. The Army's aviation forces would receive 369 tilt-rotor UAV systems (each with one ground control station and three air vehicles) in lieu of an equal number of Comanches. Taking into account attrition spares and "maintenance float" (extra aircraft bought to be used while others are undergoing maintenance), that would mean reducing the planned Comanche acquisition by 461 aircraft and purchasing a total of more than 1,900 Eagle-Eye air vehicles (see Table 9).
 

Several years ago, the Department of Defense ordered the Army to look into how Comanches and UAVs could be used together. The Army has a study under way to examine and test the concept of having a Comanche helicopter control a tactical UAV on the battlefield. But the service was also specifically asked to examine potential "trade-offs" between reconnaissance helicopters and unmanned aerial vehicles. It is not clear whether the Army is seriously examining that question. This option provides a brief look at the advantages, disadvantages, and costs of replacing reconnaissance helicopters with UAVs.

How Would It Work?

Substituting unmanned aerial vehicles for Comanches in future Army units could be done in one of two ways--both of which involve cavalry squadrons, regimental aviation squadrons, and corps target-acquisition and reconnaissance companies and platoons. The Army intends for the typical cavalry squadron attached to a division to include two troops of 12 RAH-66 Comanche helicopters.⁽³⁾ One way to carry out Option III would be to substitute one troop of 12 UAV systems for one troop of 12 Comanches. The UAVs would be controlled from the ground, and the helicopters and UAVs could operate separately or together. If they operated together, the helicopters could receive reconnaissance information from the UAVs via a communications link from the ground control stations, or the helicopters could themselves carry imagery data terminals that would allow them to see what the air vehicles see.

A second, more experimental way to implement this option would be to substitute six unmanned aerial vehicle systems for six helicopters in each troop. Then, each helicopter in a troop could control one UAV. The Army study that is looking at the technical challenges of controlling an unmanned aerial vehicle with a helicopter has not reached a conclusion about its feasibility. Having a helicopter pilot send out a UAV on a mission and receive nearly instantaneous imagery would avoid the need to operate through the communications link with the ground control station, which could save precious time in a tense combat situation. However, it also risks overloading the already busy helicopter pilot.

According to the Army doctrine, cavalry forces have two primary missions: reconnaissance and security. Reconnaissance includes route, area, and zone reconnaissance missions, and security includes screening, guarding, and covering other forces. Army officials say UAVs are capable of performing the reconnaissance missions. But in most cases, helicopters are better suited for supporting security missions, because such missions typically involve a much greater likelihood of combat.

Cavalry squadrons would not be the only forces affected by Option III. Target-acquisition and reconnaissance companies (or platoons) are assets of divisions and corps and are used mainly for supporting those units' artillery. They help identify targets, correct fire, and assess damage. In this option, tilt-rotor UAVs would also be substituted for all of the helicopters (a total of 93) assigned to those missions.

Costs

A principal advantage of Option III is cost. Tactical UAV systems, even the most sophisticated, are less expensive than manned aerial platforms, including Comanche. Substituting tilt-rotor UAV systems for Comanches would save the Army about $3.8 billion in acquisition costs. It would also save about $700 million over 15 years of operation and support, for a total savings over the life of the UAVs of about $4.5 billion.

Loss of Firepower

The biggest drawback of Option III is that unmanned aerial vehicles are not armed reconnaissance helicopters. To some degree, this option represents an apples-to-oranges comparison. The Eagle-Eye tilt-rotor UAV does not carry weapons, partly because it is unmanned and relatively cheap so losing it is not particularly worrisome.⁽⁴⁾ The more expensive Comanche, by contrast, has weapons for self-defense, as well as attack weapons such as Hellfire missiles. Thus, units that lost helicopters and gained UAV systems under this option would lose a substantial percentage of their firepower. For that reason alone, the Congress may not find the option feasible.

The trade-off CBO is emphasizing in this option is between platforms to conduct reconnaissance, surveillance, and target acquisition. The chief disadvantage of the UAV in that arena is that, unlike Comanche, it does not carry a pair of human eyes on board, which in some circumstances can be irreplaceable to confirm the identification of a particular target or reconnoiter a particular spot. Overall, however, the tilt-rotor UAV has many more advantages than disadvantages as a platform for conducting reconnaissance, surveillance, and target acquisition.

Capabilities

To better understand the other advantages and disadvantages of this option, CBO compared the reconnaissance capabilities of an air cavalry troop made of up of 12 tilt-rotor UAV systems and one made up of 12 Comanche helicopters. In the comparison, CBO used a typical mission profile for Comanche: flying to a radius of 200 kilometers and operating there for 40 minutes. The comparison assumed that all 12 helicopters and UAV systems would be available at any given moment. By several different measures, the UAVs had a number of advantages over the Comanches (see Table 10).
 

With respect to loiter time--the amount of time each aircraft can stay aloft over an area--the tilt-rotor UAV has roughly a 6-to-1 advantage over Comanche. It can provide more than four hours of loiter time, compared with approximately 40 minutes for Comanche.⁽⁵⁾ With respect to dash speed, the UAV is actually a little faster than the helicopter (370 kilometers per hour versus 315). Thus, it has a slightly greater ability to be in the right place at the right time.

Assuming that both the UAV and the Comanche troops flew out to a radius of 200 kilometers, both could observe 12 separate points or targets (one per platform)--something that resembles area reconnaissance. In making its comparison, CBO chose the most conservative assumption: substituting one UAV system with three air vehicles and a ground control station for one Comanche, rather than one air vehicle for one helicopter or some other trade-off. Observing multiple targets or areas simultaneously was deemed to be a crucial capability for a troop to have. In a real operation, of course, there may not always be 12 separate and distant targets to observe, or the 12 Comanche helicopters may operate separately.

Deployability

CBO did not conduct a detailed analysis of the deployability requirements of a cavalry squadron with two troops of Comanches versus one with a troop of Comanches and a troop of UAVs. Because cavalry squadrons contain a large, broad array of equipment other than helicopters (including tanks and artillery), the overall difference in the number of transport aircraft that the two types of units would need is likely to be marginal. That is particularly true since one Comanche and the air vehicles and mobile ground station of one tilt-rotor UAV system can each fit on one C-130. (The support equipment and the personnel to man and maintain both the helicopter and the UAV system would deploy separately.)

Overall Assessment

The primary effect of Option III would be to provide the Army with a substantial amount of reconnaissance capability at a much lower cost. The Army's reconnaissance helicopters, including Comanche, are chiefly intended to collect information. By spending extra money to acquire Comanche rather than a tilt-rotor UAV for that mission, the Army is getting a more versatile aircraft that can attack the enemy as well as observe it. However, it is also putting pilots at risk on each reconnaissance mission, even those that do not require a combat engagement.

Other factors could figure in any trade-off between unmanned aerial vehicles and reconnaissance helicopters. Helicopters are self-contained platforms in the sense that their pilots can see, report, and do whatever is necessary (in accordance with their commander's instructions). With UAVs, imagery is sent via a data link to a ground terminal where the commander can see it directly. But if the target under observation needs to be attacked, a second platform has to do the job, which can take time. In a critical situation, there may be no substitute for having an aircraft that can, for example, identify an advanced reconnaissance unit of the enemy and immediately destroy it.

UAVs, however, have the advantage of being able to operate in nuclear, biological, or chemical environments without jeopardizing the lives of their controllers. In that sense, UAVs are more versatile than helicopters. In addition, UAVs have the potential to carry many payloads other than video or infrared sensors. The payload capacity of the tilt-rotor UAV is large enough to accommodate signals intelligence packages, mine-detection equipment, and even potentially some types of armament.
 

OPTION IV: USE GLOBAL HAWK UAVs TO SUBSTITUTE FOR THE REDUCTION OF JSTARS

This option proposes taking greater advantage of unmanned aerial vehicles by purchasing a fleet of Global Hawks to give theater commanders more wide-area surveillance capability. That capability is now provided by the Joint Surveillance Target Attack Radar System (JSTARS)--a joint Army/Air Force reconnaissance system that combines a powerful multimode ground-surveillance radar with command-and-control systems on board a 707 aircraft. The purpose of the system is to detect mobile and stationary targets on the ground and transmit their locations to ground commanders and combat aircraft.

The Air Force had planned to buy 19 JSTARS aircraft in order to provide coverage for two combat theaters simultaneously. However, the recent Quadrennial Defense Review proposed reducing that purchase to 13 (plus one for testing). The Department of Defense argued that a fleet of 13 JSTARS would be able provide the round-the-clock coverage needed in a major theater war. In the event of a second conflict, some of the aircraft would have to redeployed to the second theater, possibly opening gaps in coverage. DoD plans to "explore the potential for supplementing radar coverage of enemy force movements from long-endurance unmanned aerial vehicles."⁽⁶⁾ This option reflects that idea.

There are several ways to achieve the same capability as JSTARS using high-altitude endurance UAVs, specifically Global Hawk. For example, one study commissioned by DoD proposed building advanced Global Hawks that would take advantage of existing technology to incorporate a radar system as capable as the one aboard JSTARS. CBO did not explore that option in great detail because such a choice assumes even greater technical risk than already exists in the development of Global Hawk. The current Global Hawk has been designed and built with one engine and has had only very limited testing so far. An advanced Global Hawk along the lines of the one in DoD's study would require two engines, substantial new development, and all of the accompanying technical risk that those entail.

A second alternative would be to use the Global Hawk as currently configured but to achieve the same capability as JSTARS by substituting processing power for electrical power and providing a more advanced radar.

The third alternative--which is the one that CBO explored in this option--is to provide the capability of a JSTARS by using several of the currently configured Global Hawks and their radars in the moving-target-indicator mode. According to CBO's analysis, that would require having at least three Global Hawks aloft simultaneously.

Costs

Assuming that the Air Force follows the Quadrennial Defense Review's recommendation to reduce the JSTARS fleet from 19 to 13 aircraft, this option would cost almost $1.2 billion more than the Air Force's plan in acquisition and 15-year operating and support costs. The Global Hawks bought under this option would be in addition to the planned purchases of Global Hawks for other reconnaissance and surveillance missions as well as the 13 JSTARS aircraft.

The reduction proposed by the Quadrennial Defense Review assumed that the NATO alliance would purchase a number of JSTARS, which would supplement the U.S. fleet in times of crisis, if necessary. But NATO has decided not to acquire those aircraft. As a result, many people inside and outside the Congress are suggesting that the Air Force buy the original 19 aircraft after all. In that event, this option would substitute 11 Global Hawks for the extra six JSTARS aircraft, saving a total of around $2.3 billion in acquisition and 15-year operating and support costs.

Capabilities

The radar systems in both JSTARS and Global Hawk include a moving-target indicator that detects moving vehicles. There are important differences, however, in the capability of those indicators. The major measures of their capability are ground-referenced coverage area, revisit (or update) time, minimal detectable velocity, range resolution, and azimuth accuracy (see Table 11).
 

Ground-Referenced Coverage Area. The ground-referenced coverage area is the area on the ground that the radar's moving-target indicator can cover. According to unclassified sources, that area for JSTARS is 150 kilometers by 180 kilometers, or 27,000 square kilometers--the notional area for which an Army corps has responsibility. (JSTARS's actual coverage capability is larger, although the exact figure is classified.)⁽⁷⁾ The ground-referenced coverage area of Global Hawk's radar, as it is currently configured, is about one-third of that. Thus, providing the same area coverage as one JSTARS would appear to require at least three Global Hawks.

Substituting three Global Hawks for a JSTARS could create some command-and-control problems, however. For example, it would require more coordination and possibly better data links with the units on the ground who would be receiving the images from the UAVs. Because three aircraft would be doing the job of one and there would necessarily be some overlap in coverage area, ground units might have to access the imagery from all three Global Hawks that were aloft and know which air vehicle was covering which part of the ground.

Revisit Time. The revisit (or update) time is the amount of time required for the radar to sweep over an area and provide the latest available imagery of moving targets. The revisit time for Global Hawk's radar over the ground-referenced coverage area is around 70 seconds. JSTARS has a faster revisit time--at least 60 seconds--although the actual number is classified.⁽⁸⁾

Revisit time is important for several reasons. The moving-target indicator only detects vehicles that are moving at a certain minimum speed. The radar image appears as lighted dots on a screen. (Tracked vehicles and wheeled vehicles can be displayed as two different colors.) The average person will not be able to make much sense of a moving-target-indicator image other than to know that there are moving vehicles within the sweep of the radar. A trained expert in such imagery, however, will be able to distinguish among the dots and often determine the size and type of the units. If the delay between sweeps of the radar is too great, it is more difficult to determine which dot is which from the previous sweep. Thus, tracking particular formations of vehicles may be more difficult if the lag time is too great.

In addition, because both JSTARS and Global Hawk are equipped with synthetic aperture radars, which can provide an image of a particular location, it is important to know when a vehicle or column of vehicles has stopped moving. When that happens the vehicles disappear from the moving-target indicator. If they are sufficiently interesting targets, however, a radar operator can use the synthetic aperture radar to determine what they are. But too long a lag time between sweeps of the radar may make it more difficult for the operator to find the target with the synthetic aperture radar after it has disappeared from the moving-target indicator.

The key issue is whether Global Hawk's slower revisit time is quick enough for the moving-target indicator to provide useful imagery. That appears to be the case, particularly because today the use of such indicators is geared toward detecting large formations of vehicles. A revisit time of 70 seconds will probably not have much effect on an analyst's ability to determine what vehicle formations the radar is picking up, although clearly it will have a slight negative impact (relative to JSTARS's revisit time). In the future, if moving-target indicators are used to try to identify more specific targets or vehicles, Global Hawk's radar may be at a disadvantage. But by then, there may also be more advanced radars with faster revisit times that Global Hawk could use.

Minimum Detectable Velocity. As its name suggests, minimum detectable velocity is the speed at which an object must be traveling to be detected by the moving-target indicator. However, only objects that are moving with some degree of perpendicularity to the radar signal will be picked up. Objects that are moving strictly parallel will not be detected. The reason is that as an object moves toward or away from the radar, it causes a shift in the radar signal and thus is detected. If an object is moving parallel to the radar, there is no shift in the signal or the shift is so slight that the object appears stationary--or, in effect, as though it were not there.

The minimum detectable velocity for JSTARS is classified. But for Global Hawk, that speed is four knots if the target is moving directly toward or away from the radar. If the object is traveling at an angle to the radar, it must be going at a higher speed to have the effect of moving toward or away from the radar at four knots. For example, at a slight angle from the perpendicular, an object may need to be moving at six to 10 knots to register. At a sharper angle, the speed would have to be much greater for the object to be detected.

Range Resolution. Another aspect of a radar's moving-target-indicator mode is the degree to which it can distinguish between two objects. In the most literal sense, a radar must "resolve" the imagery it receives. Different radars can do that to different extents, just as two people looking at an optometrist's eye chart may have different abilities to distinguish between the letters on the chart. If a convoy of vehicles is traveling along a road, those vehicles must be a certain distance apart for the radar to recognize them as separate objects. The closer together they can be and still be detected separately, the better range resolution the radar has.

Global Hawk's moving-target indicator is particularly good at resolving objects that are short distances apart, about 10 to 20 meters. An opponent would probably have a difficult time bunching its vehicles so close together that the radar would be unable to distinguish between them. The value of a good range resolution is most pronounced when a trained analyst of moving-target-indicator imagery is attempting to determine whether a particular set of vehicles represents a platoon, a company, or some other size unit. Being able to count--even roughly--the number of vehicles in an enemy unit may provide valuable information to a battlefield commander.

Azimuth Accuracy. Azimuth accuracy is relatively straightforward: it is how precisely a radar can determine the location of the objects it detects. By this measure, Global Hawk's moving-target indicator clearly has less capability than that of JSTARS. Global Hawk's indicator can place the location of a vehicle up to 350 meters away from its actual position, whereas JSTARS is expected to be more accurate. However, that weakness is mitigated somewhat by the fact that moving-target indicators are mostly concerned with identifying relatively large formations of vehicles rather than tracking individual vehicles. Thus, if the indicator picks up a convoy of vehicles traveling in a line 200 meters parallel to a road, the analyst can assume that the formation is probably on the road and that the 200-meter difference is within the radar's margin of error.

Other Capability Issues. A significant advantage of Global Hawk over JSTARS is that its moving-target indicator can provide "deep" (that is, long-distance) coverage without risking the lives of an aircrew. JSTARS is intended to operate at the forward line of U.S. troops and provide coverage over a range of 160 kilometers. At that distance, however, it would not detect whether enemy reinforcements were entering the theater and what kind of forces they were. To give a specific example, a JSTARS operating near the Kuwaiti border with Iraq would not be able to indicate the size and scope of any forces that might be moving into the theater from Basra or beyond. But Global Hawk, which is unmanned, could do so without risking an aircrew. Attacking enemy forces with precision munitions deeply (long before they are able to reach the forward line of U.S. troops) is an important element of future U.S. warfighting strategy. For that reason, it might make sense to buy even more Global Hawks to provide deep coverage or use some of the ones intended for coverage at the forward line of U.S. troops for deep coverage.

Deployability

In general, JSTARS and Global Hawk aircraft would self-deploy (fly to the theater of operations) in the event of a major regional conflict. But the exact deployment requirements of the Air Force's plan and Option IV are unclear, in part because it is unclear what the final JSTARS and Global Hawk programs will look like.

Overall Assessment

Assuming the Air Force follows the recommendation of the Quadrennial Defense Review to reduce the number of JSTARS aircraft, Option IV would give one or more theater commanders substantial additional capability to detect moving targets at a total cost of $1.2 billion. If instead the Air Force buys additional JSTARS aircraft, Option IV would save money in comparison. Global Hawk is not quite as capable as JSTARS, but it may be good enough for its intended missions and can be used without risking aircrews. Probably the most serious drawback of Option IV is the additional command-and-control efforts that would be required to make the option work.
 

OPTION V: END DARKSTAR AFTER THE ACTD AND RELY ON OTHER SYSTEMS

As noted in Chapter I, some Members of Congress have raised concerns about apparent overlaps between three of DoD's unmanned aerial vehicle programs: Darkstar, Global Hawk, and Predator. This option would address those concerns by cancelling the Darkstar program after its Advanced Concept Technology Demonstration phase ends and relying on other types of endurance UAVs instead. Darkstar is a high-altitude UAV that is expected to have low-observable (stealthy) characteristics. It is designed to carry out a particular mission: collecting imagery over highly defended targets before an enemy's air defenses have been suppressed. In addition, it may be particularly useful in supporting special-operations forces.

Aside from its stealthiness, Darkstar is expected to be less capable than Global Hawk, although more capable, for the most part, than Predator (see Table 12). The Defense Airborne Reconnaissance Office and the Air Force have described Global Hawk as a highly capable but moderately survivable UAV, whereas Darkstar is a highly survivable but moderately capable UAV. The chief advantage of buying Darkstar, therefore, is to buy stealthy reconnaissance capability.
 

The degree to which Darkstar really is a highly stealthy imagery-collection platform, however, is not clear--especially compared with Global Hawk and Predator. According to DoD, tests in May 1996 "validated" Darkstar's low-observable design.⁽⁹⁾ But the Darkstar program is still recovering from the crash of the first air vehicle, and much work remains before it will be successfully completed. A true test of Darkstar's stealth, including when its sensors are engaged, must wait until further development and military-utility demonstrations take place.

Although not stealthy, Global Hawk and Predator have some features that might help them survive in an environment in which enemy air defenses had not been suppressed. (For more on the issue of UAVs' survivability, see Box 2.) Global Hawk flies higher than all but the most capable of surface-to-air missiles (SAMs). Furthermore, it is equipped with an electronic countermeasures suite to thwart enemy SAMs in the event that they target the air vehicle. Predator has an all-composite air frame (which produces a smaller radar reflection than an air frame made of metal), can operate at night, and flies slower than enemy air-defense radars are typically programmed to detect.
 

Does that mean Global Hawk and Predator are as survivable as Darkstar? Probably not. It does suggest, however, that in light of the less threatening environment that the United States faces for the foreseeable future and the "niche" mission that Darkstar fulfills, it may be cost-effective to end the Darkstar program at the conclusion of its development process. Three air vehicles are expected to be left over from the Advanced Concept Technology Demonstration. They would still be available for the rare, high-value mission in which having a stealthy reconnaissance collector was especially desirable. Global Hawk and Predator could also be used for those infrequent missions if commanders were willing to accept the risk of higher attrition or the possibility that the imagery might not be collected.

Moreover, the types of systems that would probably pose the greatest threat to Global Hawk--the SA-10s and SA-12s, or "double-digit SAMs"--are relatively rare. And they would most likely be one of the first targets of U.S. air forces seeking to establish air superiority. Once those SAMs were destroyed, Global Hawk could operate more freely even if the rest of the enemy's air-defense systems had not yet been suppressed.

Not all Air Force officials or defense analysts would agree with the analysis underpinning this option. Some argue that the United States will face severer threats in the future because of the expected proliferation of high-quality surface-to-air missiles. Furthermore, one Air Force official argues, with the retirement of the SR-71 high-altitude reconnaissance plane, the United States no longer has the means to collect imagery over highly defended targets. Satellites cannot always be relied on to be in the right place at the right time, and other reconnaissance aircraft, such as the U-2, can be shot down by SAMs.⁽¹⁰⁾ Global Hawk and Predator will also be vulnerable.

Costs

Because this option would not buy anything to replace Darkstar, the savings stem directly from ending the program. Savings in acquisition costs alone would be more than $600 million compared with the Air Force's likely plan, plus another $400 million in 15-year operating and support costs.

Overall Assessment

In light of the fact that Darkstar does not provide any additional capability over Global Hawk besides stealth, the price of Darkstar is in fact the price of stealth in the UAV force structure. The Congress must decide how much it wants to pay for stealthy UAVs; right now, the price tag appears to be about $1 billion. Furthermore, the mission that Darkstar would fulfill is likely to be infrequent and could be accomplished by the three ACTD prototypes or by Global Hawk or Predator when necessary.

1. See, for example, General Accounting Office, Unmanned Aerial Vehicles: Outrider Demonstrations Will Be Inadequate to Justify Further Production, NSIAD-97-153 (September 1997), pp. 4-7.

2. Gen. Ronald H. Griffith, "Memorandum for Under Secretary of Defense for Acquisition and Technology on Tactical Aerial Vehicles" (May 8, 1997).

3. Under the Army's reorganization of its aviation assets--the Aviation Restructure Initiative--the two air cavalry troops per squadron will have either eight OH-58D Kiowa Warrior helicopters or eight AH-1 Cobra helicopters. When the Comanche is fully fielded, each troop is intended to have 12 of them. Thus, compared with the Army's current force structure, this option would reduce the number of armed helicopters only slightly and provide a large additional reconnaissance capability.

4. Bell Helicopter has plans in which Eagle-Eye could carry and deliver sensor-fuzed weapons. But in analyzing this option, CBO did not consider those plans feasible in the short or intermediate term because the UAV is still at the demonstrator stage and has not yet flown many hours and because the military's concepts of operation for using combat UAVs are still relatively undeveloped.

5. That figure assumes that Comanche is not using its external fuel tanks, which it is not likely to do most of the time.

6. Secretary of Defense William S. Cohen, Report of the Quadrennial Defense Review (May 1997), p. 45.

7. John Haystead, "JSTARS--Real-Time Warning and Control for Surface Warfare," Defense Electronics (July 1990), p. 39.

8. Joris Janssen Lok, "Joint STARS Gains vs Greater Radar Control," Jane's Defence Weekly, April 23, 1997, p. 30.

9. Department of Defense, Defense Airborne Reconnaissance Office, UAV Annual Report, FY 1996 (November 6, 1996), pp. 22-23.

10. It is worth noting that according to Defense News, the Administration has stated that there is no military requirement for the SR-71, a position the Air Force supports. See William Matthews, "USAF Again Retires SR-71 Reconnaissance Planes," Defense News, May 18, 1998, p. 32.

---