Review of Current UGV Efforts
This chapter discusses efforts to achieve unmanned ground vehicle (UGV) capabilities that may be applicable to the Future Combat Systems (FCS), including Army Science and Technology Objectives (STOs) and initiatives of the Department of Defense (DOD), Defense Advanced Research Projects Agency (DARPA), other government agencies/services, as well as foreign UGV activities.
ARMY SCIENCE AND TECHNOLOGY PROGRAM
The Army Science and Technology program is administered through a well established process for initiating and managing a series of initiatives described in terms of Science and Technology Objectives. In this section relevant Science and Technology Objectives are described briefly and analyzed in terms of capabilities and impact on the Future Combat Systems program. The Army FCS program as presently planned is in a pre-systems acquisition phase of ongoing activities in development of user needs, science and technology, and concept development. Critical decisions regarding which technologies will be available to be integrated into the baseline FCS system design and development will be made in fiscal years 2003 and 2006.
As described in the Army Science and Technology Master Plan (ASTMP), there are two primary STOs in support of UGV developments: (1) the Tank-Automotive Research, Development, and Engineering Center UGV follower Advanced Technology Demonstration and (2) the Army Research Laboratory Semiautonomous Robotics for the Future Combat Systems.
Robotic Follower Advanced Technology Demonstration (STO: III.GC.2000.04)
The goal of the robotic follower Advanced Technology Demonstration (ATD) is to develop, integrate, and demonstrate in a relevant environment an unmanned follower vehicular system capability for future land combat vehicles. The technology will support a wide variety of applications, such as transporting ruck sacks and logistics supplies and providing security for army rear areas, leading to a lethal capability for beyond line-of-sight fire. As with the FCS program, the ATD has been accelerated to meet FCS Block I and Block II objectives. The follower ATD and a sister Crew Integration and Automation Testbed ATD (see later section) have been combined under a Vetronics Technology contract to provide two systems as follows:
A follower UGV system capable of following a manned lead vehicle or dismounted soldier, both on-road and off-road
A crew integration and automated testbed (CAT) vehicle system consisting of a two-soldier crew, with multimission capabilities allowing the crew to accomplish “shoot,” “scout,” and “carrier” missions while coordinating and controlling unmanned ground and air systems, including the follower vehicle.
The program is structured to provide, in close consonance with the Army Research Laboratory Semiautonomous Robotics STO (see following section), the robotic technology to support the performance requirements for the Mule UGV and the Armed Reconnaissance Vehicle UGV, as described by the FCS lead system integrator team.
To support the Mule mission the ATD program is planned in FY03 to demonstrate leader-follower capability with up to 1-km separation, including limited semiautonomy to avoid obstacles. Capability for teleoperation will be included as a backup. For the armed reconnaissance vehicle (ARV) mission, two robotic followers (a Stryker and an updated experimental unmanned ground vehicle [XUV]) will be included in the demonstration. In FY06 the ATD program is scheduled to demonstrate multilane on-road operation as
well as off-road operation with up to 24-hour separation, with limited semiautonomy.
This description of the robotic follower ATD program, coupled with the CAT ATD program, provides the basis for the committee’s answer to Task Statement Question 2.a in Box 3-1.
Semiautonomous Robotics for Future Combat Systems (STO: IV.GC.2001.03)
The focus of this STO is to develop semiautonomous mobility technology critical to achieving the transformation envisioned for the FCS. A number of on-road and cross-country mobility experiments have successfully demonstrated initial capabilities in controlled environments (e.g., known terrain, daylight, and favorable weather). Continuing technical efforts are focused on perception and sensor technologies, intelligent vehicle control, tactical environment behaviors, and supervision of unmanned ground systems. Key technologies include obstacle avoidance, terrain characterization and classification, and fusion of data from multiple classifiers. Several technologies developed under this STO were integrated with the experimental unmanned vehicles (XUV) used for the DEMO II and DEMO III UGV demonstrations (see DOD Joint Robotics program).
A restructured program has been proposed to demonstrate during FY03 terrain-dependent semiautonomous navigation at ranges over 1 to 3 km and limited mission capabil
BOX 3-1 Task Statement Question 2.a
Question: Will the follower UGV Advanced Technology Demonstration (ATD) lead to a capability that will meet stated Army operational requirements in time to be integrated into the baseline FCS development program?
Answer: While it is likely that the follower UGV ATD objectives will be achieved, the program does not consider relevant supporting technologies or system-level requirements. Consequently, it is unlikely that all of the technologies needed for a follower UGV system will reach TRL 6 in the 2003–2006 time frame that is required for the baseline FCS development program.
The committee is aware that the follower ATD is being restructured (in consonance with the Army Research Laboratory STO and the Crew Integration and Automation Testbed STO at the Tank-Automotive Research and Development and Engineering Center (TARDEC)) to focus on Mule and Armed Reconnaissance Vehicle capabilities, consistent with the concepts requested of industry by the FCS lead system integrator. To the extent that this assists in defining system requirements, it is clearly a step in the right direction. It remains to be seen, however, whether this dual focus will accelerate development of UGV systems for FCS.
ity for such survivability functions as obscurant dispersal and remote sensing capabilities for self-protection missions. The continuing program has the goal of demonstrating limited scout capabilities by FY2006, based on improved LADAR (laser detection and ranging) and stereo vision sensing, and on advances in perception and intelligent vehicle control. An extensive set of field experiments in a variety of settings is an essential part of this effort, but additional funding will be required to support development and instrumentation of test-bed vehicles to support the field tests. This description of the Army Research Laboratory STO program provides the basis for the answer to Task Statement Question 2.b in Box 3-2.
Other Army R&D Activities Contributing to FCS
The robotic follower STO and the semiautonomous robotics for FCS STO are the principal science and technology efforts for developing unmanned capabilities for the FCS. In addition to these primary TARDEC and ARL STOs, a review of other Army STOs indicates that 19 additional STOs use the term “robotics” or “unmanned” in the description of the STO activity. Of these, the following seven are of principal interest to the further development of UGV technology and offer potential support and risk mitigation for achieving the overall goals of the FCS program.
Future Combat Systems (STO: III.GC.2000.03)
A single STO covers the collaborative DARPA/Army FCS program described in Chapter 2. It is directed toward lightweight, lethal, deployable, self-sustaining, and survivable combat systems, collectively known as the Future Combat Systems (FCS), for the 2008–2015 time frame. UGVs are considered to be integral to the FCS. The DARPA efforts under this STO are described in a separate section below.
BOX 3-2 Task Statement Question 2.b
Question: Will the Army Research Laboratory (ARL) STO program result in significant advances in UGV autonomy beyond that achieved in the follower UGV ATD, and what operational requirements would the resulting capability be able to address?
Answer: The ARL STO is being restructured in consonance with the follower ATD to focus on Mule and Armed Reconnaissance Vehicle capabilities, as described by the FCS lead system integrator. In light of this the ARL STO will not significantly advance UGV autonomy beyond that likely to be attained in the follower UGV ATD for those particular applications.
The original ARL STO addressed capabilities that would support autonomous requirements not addressed in the follower ATD, including planning, navigation, and human–robot interaction.
Crew Integration and Automation Testbed ATD (STO: III.GC.1999.02)
As described previously, this STO is contributory to and has been combined with the robotic follower ATD because as the technology is directed toward the reduction of the workload on the human crew and therefore is directly applicable to unmanned crew functions. Specific technologies include intelligent driving decision aids, the application of semiautonomous driving technology, and automated route planning, all of which are pertinent to both manned and unmanned vehicles. The complete development of driving technologies and decision aids is planned for FY03, including demonstration on a vehicle test bed. This STO will contribute technologies to the Mule and ARV UGV missions as defined by the FCS lead system integrator.
Obstacle Marking and Vehicle Guidance (STO: IV.EN.2000.02)
The focus of this effort is to dispense “smart” markers to transmit and receive critical navigation information through and around obstacles or minefields. Successful development of this technology could be applicable not only to manned vehicles and dismounted forces but also to the path planning and path following of unmanned vehicles. It is planned by FY03 to have a complete smart marker system that will be timely for evaluation with the robotic follower ATD and the Semiautonomous Robotic Vehicle.
Mobility Support for Objective Force Maneuver (COE-ERDC STO COE.2002.04)
The results of this research will provide the capabilities and algorithms to quantify mobility and physical agility parameters that are essential for characterizing unmanned systems. The work addresses current deficiencies in modeling the breaching and crossing of complex obstacles with lighter vehicles than are now in the inventory. By FY06 the current program plan projects that the technologies to quantify mobility in urban environments, assess and negotiate obstacles, and model reliable driving behaviors will be available. The products of this program will include vehicle performance assessment tools and measures of performance of physical agility.
Advanced Robotic Simulation (STRICOM STR-03)
The objective of this activity is to develop intelligent behaviors for robotic systems within an environment that will provide complex mission and coordination behaviors for real robots. The basic premise is to leverage the demonstrated abilities to create complex military behaviors in semiautomated forces simulation and extend the expertise to create and experiment with military behaviors in live robots. Development of tactics and procedures for employment of new robotics systems in the battlespace are limited at present. This advanced robotic simulation STO provides the capability to train and control unmanned forces in collaborative simulation environments. It will provide personnel with the capability to train with robotic systems and also provide for development and experimentation with manned and unmanned interfaces, for instance, to determine how many unmanned systems an FCS crew can control under a variety of scenarios.
Sensors for the Objective Force ATD (STO: III.IS.2001.02)
Unmanned networked sensors can provide remote monitoring and advanced warning for robotic system operation in and beyond friendly force lines. The goal is to develop sensor packages for UGVs using advanced sensor technologies integrated with a robust network architecture. A component of this program involves the use of virtual simulation and live experiments in operational environments to establish baseline architectures, address operational integration issues, and investigate new operational concepts. If successful, this technology will be available in FY05 and will be well positioned for technology insertion in the FY06 FCS system block upgrade. This generic type of low-cost, distributed sensor system could provide advanced self-protection for robotic components of the FCS.
Airborne Manned/Unmanned System Technology Demonstration (STO: III.AV.1999.01)
This effort will demonstrate through simulation and flight test the control, tactics, and procedures for the operation of manned and unmanned air vehicles. Technical barriers associated with manned and unmanned teaming will also be addressed. The resultant software products for manned and unmanned teaming will be available in FY03 and may provide valuable information to UGV developers by means of lessons learned.
Joint Robotics Program
The DOD is actively developing special-purpose robotic UGVs for such applications as range and mine clearance, force protection, breaching, neutralization of ordnance and explosives devices, and reconnaissance, among others (DOD, 2001). These special-purpose robotic vehicles represent a first step toward achieving functional capabilities for ground systems. The earliest DOD robotic systems employed teleoperation, making use of an operator-in-the-loop, but this is clearly not enough. According to the DOD Joint Robotics
Program Coordinator, “Some users can accept teleoperation initially, but all users want autonomy ultimately” (Toscano, 2001a).
There are a variety of robotic UGVs being developed under the auspices of the DOD Joint Robotics program (Toscano, 2001a,b,c). The mission of the Joint Robotics Program (JRP) is to develop and field a family of affordable and effective mobile ground systems, develop and transition technologies necessary to meet evolving user requirements, and serve as a catalyst for insertion of robotic systems and technologies into the force structure. The JRP is structured to field first-generation robotic systems, mature promising technologies, and upgrade these capabilities by means of an evolutionary strategy. In the near term the acquisition programs emphasize teleoperation, operation on diverse terrain, more autonomous functioning for structured environments, and extensive opportunities for users to operate UGVs.
The JRP oversees efforts for specific robotics programs in all services. Its work on UGVs is managed primarily by the Unmanned Ground Vehicles/Systems Joint Project Office, Redstone Arsenal, Alabama; the Air Force Research Laboratory, Tyndall Air Force Base, Florida; and the Army Tank-Automotive Research, Development and Engineering Center in Warren, Michigan.
Several programs managed under the Joint Robotics Program umbrella include technologies with a potential for FCS UGVs:
The Standardized Robotics System (SRS) is a kit with components that can be used to provide teleoperations to various fielded systems. The Vehicle Teleoperation occupational requirements document (ORD) was approved in 1997. Operational employment for the SRS includes obstacle/minefield breaching and route and area clearing. The vehicles currently planned include D7G, T3, and Deployable Universal Combat Earthmover (DUECE) bulldozers, and an upgrade to the M1 Abrams chassis of the M-60 Panther currently being employed by U.S. forces in Bosnia.
The Robotics Combat Support System (RCSS) is a short-range (300 meters) line-of-sight remote control vehicle that uses a set of interchangeable attachments to perform a variety of engineering missions including landmine neutralization, wire breaching, dispensing of obscurants, and demolition emplacement.
The Man-Portable Robotic System (MPRS) will provide lightweight, man-portable UGVs to support light forces and special operations missions, focusing on reconnaissance during military operations in urban terrain (MOUT). The Joint Robotics Program Office has developed and matured the necessary technology through the conduct of concept demonstrations, fielding Matilda (Man Portable Robot) UGVs to the National Guard and the active Army for special purposes, such as cave clearance operations, and monitoring the DARPA Tactical Mobile Robotics program.
The Basic UXO Gathering System (BUGS) employs a semiautonomous reconnaissance platform that controls several small man-portable, expendable UGVs to clear submunitions and other unexploded ordnance (UXO) from the battlefield. After the reconnaissance platform locates the UXO and downloads pertinent location information to smaller robots, the smaller robots pick up and remove, or conduct “blow in place” operations. BUGS is in the technology demonstration and evaluation phase.
The Remote Ordnance Neutralization System (RONS) was started by the JRP, transitioned to production, and then upgraded by the Services. RONS has been fielded to a number of explosive ordnance disposal (EOD) units and additional systems are currently being procured.
The Robotics for Agile Combat Support (RACS) program consists of several robotic systems that can perform different types of missions. The All-purpose Remote Transport System—Force Protection (ARTS-FP) allows an operator to investigate and disable suspicious packages and vehicles. The All-purpose Remote Transport System—Range Clearance (ARTS-RC) includes a frangible blade assembly that provides an initial shock to UXO so EOD personnel can safely move through a cleared path. The ARTS-FP/RC systems are currently being procured and fielded. The Automated Ordnance Excavator (AOE) uses a commercial excavator with an extended reach capability that can precisely locate itself and dig up and remotely grasp subsurface UXO. The Remote Crash Rescue Vehicle (RCRV) will provide an autonomous crash/fire rescue platform that can respond in aircraft accidents.
The Mobile Detection Assessment Response System—Interior (MDARS-I) is a mobile robotic security platform that can conduct random patrols inside warehouses and storage areas, as well as conduct electronic inventories. MDARS-I is currently in development. The Mobile Detection Assessment Response System—Exterior (MDARS-E) is the exterior version of MDARS-I. MDARS-E can conduct robotic security functions at large fixed installations, such as warehouses and ammunition storage sites. It can also conduct such physical security tasks as intruder detection, assessment, lock/barrier checks, and alarm response. This latter system is being designed for a single operator to simultaneously control up to 32 robots.
The Mobility Enhancement Program (MEP) is aimed at improving the mobility of small, man-portable unmanned systems in support of urban warfare, combat engineering, physical security and force protection missions. The program has two main thrusts: the Omni-Directional Inspection System (ODIS) and the T3 High Mobility Platform. ODIS is a small high-agility platform that can read license plates and be driven under and survey the underside of suspect vehicles. The T3 High Mobility Platform is a novel 6 × 6 platform that is being used to explore UGV mobility over rugged terrain.
Demo III Program
The Demo III program is a technology base effort begun under the DOD Joint Robotics Program being conducted by the Army Research Laboratory and its government and industry partners. The program has been focused on developing and demonstrating technology that can provide supervised autonomous mobility in an unstructured environment. The JRP transferred responsibility for the Demo III program to the Army at the beginning of FY2001. A successor effort to the Demo III program is the ARL Collaborative Technology Alliance (CTA) in Robotics involving many of the same industry participants.
The committee was invited to view the formal Demo III demonstration at Fort Indiantown Gap in November 2001. Key technologies demonstrated included: perception algorithms for the fusion of information from multiple mobility sensors (daylight video, FLIR, LADAR, radar), object classification, and active vision; semiautonomous controls; dynamic planning/replanning; tactical behaviors; and soldier-robot interactions (Shoemaker, 2001). Appendix C includes an analysis of the Demo III contribution to semiautonomous A-B mobility development.
DARPA Unmanned Vehicle Programs
In addition to working directly with the Army on the FCS conceptual design, DARPA has undertaken four advanced research programs with high potential to benefit the development of future Army UGV systems. These include the Unmanned Ground Combat Vehicle (UGCV), PerceptOR (Perception off-road) Tactical Mobile Robotics (TMR), and Organic Air Vehicle (OAV) programs.
The Unmanned Ground Combat Vehicle (UGCV) program has as an objective the development of prototypes to demonstrate advanced vehicle design and interaction to achieve new levels of mobility, endurance, and payload capacity. Two payload classes of vehicles are being developed, a 150-kg payload version and a 1,500-kg payload version. Missions for both classes of vehicles are expected to evolve with the overall concept of the Army’s FCS.
Performance benefits are expected to be harnessed in the UGCV program as a result of being unrestrained by conventional design parameters associated with accommodating onboard human crew. These parameters include:
Shock constraints (collision, rollover, blast)
Vibration constraints (absorbed power)
Life support in nuclear, biological, and chemical environments
Man–machine interface constraints (e.g., pedals, seats, steering wheels)
Human comfort constraints (temperature, humidity, cockpit dimensions)
Time constants associated with human reactions
Risk constraints (loss of life versus loss of vehicle)
Survivability constraints (signature, shot lines)
Geometric constraints (e.g., volume for humans, window sizes) and
Human vulnerability issues (e.g., gun gas, energy leaks).
Because the resulting vehicles are not required to accommodate crew-associated constraints, innovative methods of design as well as operational use can be considered.
Operational endurance in terms of the range of travel of the vehicle and the time duration between refueling is also a key parameter of the UGCV program. The UGCV is expected to execute missions over much longer resupply periods than its manned counterparts. When resupply is needed, it is expected to be limited to fuel drops that the UGCV can self-serve quickly and resume its mission. Fourteen-day duration and ranges of at least 450 km are considered objectives of the program.
Because the UGCV can be expected to operate with imperfect knowledge of the environment, it may occasionally crash or roll over. The program is seeking designs that can recover from impacts with trees, walls, and rocks and have the ability of self-recovery from rollover or operation in inverted mode.
Designs must consider operations in forested or urban environments (complex terrain). In general, narrow vehicle concepts (or novel mobility concepts) will have higher maneuverability in these confined environments. In earlier programs vehicle width proved to be a limiting factor in attempts to move through forested areas. Higher speed operations in open but broken terrain show that a large wheelbase and/or low center of gravity have obvious advantages for stability. Narrow vehicles (e.g., mono-tracks, motorcycles) are possible options for this confined access, but issues associated with lateral stability (which may be low if practical self recovery is addressed) continue to need to be addressed.
The PerceptOR program seeks to quantify and develop ground robot perception for off-road and urban mobility under a variety of terrain and environmental conditions. There
is a strong emphasis on experiments in real-world conditions. Inexpensive surrogate vehicles are being used. Experiments are to be conducted with various changes in spectral, thermal, and material compositions changes to allow true all-weather, day-and-night operations.
The DARPA Tactical Mobile Robotics (TMR) program investigated ways to penetrate denied areas and project operational influence in ways that humans cannot by using reliable semiautonomous robotic platforms. Its approach was to integrate sensors, locomotion, power, and communications with limited autonomy on a compact, man-portable platform capable of penetrating into denied areas and serving as an extension of the soldier.
Program technical challenges included close-to-the-ground mobility in cluttered and complex terrain, perception for obstacle negotiation, and autonomous fault recovery. Prospective users were special operations and early-entry forces of all Services. The program has been successful in developing and evaluating robust and agile small-robot platform prototypes and in achieving advances in onboard sensor integration and data processing, permitting increased autonomous behavior, specifically in route selection and navigation, obstacle detection, classification and avoidance, and fault recovery from communications failure and platform destabilization.
Four different types of TMR research platforms were provided with volunteers from the TMR contractor companies and academic institutions to assist the search-and-rescue response at the site of the destroyed World Trade Center in New York City. During the final year of the program the research results were integrated onto a semiautonomous Packbot platform prototype with selected mission package options. The capabilities developed are similar to the Searcher example system postulated by the committee and could provide the foundation for developing a soldier-portable robot as envisioned by the FCS LSI.
The DARPA Organic All-Weather Targeting Air Vehicle (OAV) program is a key FCS enabling technology program sponsored by DARPA and the Army. The OAV program merges technologies for small, vertical takeoff and landing (VTOL) unmanned air vehicles (UAVs) with autonomous capabilities in order to enhance the situational awareness and effectiveness of soldiers in a network-centric battlefield.
The OAV concept employs ducted fan configurations with duct diameters ranging from 6 to 36 inches with both hover and cruise flight capability. VTOL eliminates the need for a separate launcher or airfield from which to operate. The vehicle can land autonomously to provide continuous surveillance from the ground (or a building ledge) using sensor packages currently available or in development. Operators can remotely order the OAV to “perch and stare” or to move to other locations or to return to base, adapting the capability to changing battlefield conditions.
Such a relocatable sensor capability for FCS could be linked with network-centric autonomous ground vehicles, such as the Hunter-Killer, and provide continuing updates of intelligence for situational awareness in ground operations.
Air Force Initiatives
While UAVs are a primary development focus, the Air Force is also concerned with UGVs for special purposes.
Unmanned Combat Air Vehicles
The Unmanned Combat Air Vehicle (UCAV) program is a joint DARPA/Air Force ATD program that will demonstrate the technical feasibility of UCAV systems that can effectively and affordably prosecute twenty-first century SEAD/strike missions within the emerging global command and control architecture. The objective of the UCAV ATD, also called the UCAV Demonstration System (UDS), is to design, develop, integrate, and demonstrate critical and key technologies, processes, and system attributes pertaining to an operational UCAV system. The critical technology areas are adaptive autonomous control; advanced cognitive aids integration; secure robust command, control, and communication; and compatibility with integrated battlespace. Through its UDS the Air Force seeks to validate assertions from its prior studies that a future UCAV Operating System (UOS) is both effective and affordable.
The UCAV program, along with other Air Force programs to improve the early Predator and Global Hawk UAVs, will rely on many of the same technology developments in perception, planning, human–robot interaction, and communications that are needed by UGV systems.
Air Force Unmanned Ground Systems
Air Force UGV efforts, including those managed under the Joint Robotics Program, are centered at the Air Force Research Laboratory at Tyndall Air Force Base, Florida. The laboratory group has a mission to “conduct research and development of advanced robotic technologies and systems to protect, support and augment the war fighter in the accomplishment of dirty, dull, dangerous and impossible missions” (AFRL, 2002). To this end they are developing robotic systems that will provide a spectrum of devices for agile combat support (ACS) and have developed an impressive array of vehicles based on commercial off-the-shelf (COTS) piece parts employing the Joint Architecture for Unmanned Ground Systems (JAUGS) common architecture, focusing on modularity and interoperability. These units are teleoperated (rf and tethered fiber optic) and cover such tasks as explosive ordnance clearing from training ranges and other areas where unexploded ordnance is present. The units are robotized Caterpillar D-8 bulldozers, Caterpillar 325L
excavators, and an All-Purpose Remote Transport System (ARTS) that can employ a number of bolt-on attachments (laser ordnance neutralization system, articulated remote manipulator system II, water-cutting head, charge-setting device, flail, and firefighting nozzle) to process the ordnance uncovered by the larger units. The evolutionary path for this vehicle is to provide such additional capabilities as infrared imaging system, GPS, and inertial navigation, and increased autonomy through path planning and obstacle avoidance software.
The Advanced Robotic Modules and Systems project and the Autonomous Vehicles Technologies project are conducted with industry partners and the University of Florida. These programs currently address vehicle positioning sensors, path planning, vehicle control, and modular architecture development and integration. Coupled to the autonomous vehicle technologies are mission packages that will allow validation of common architecture and modularity of the robotic system. Applications envisioned are RSTA (reconnaissance, surveillance and target acquisition), mule, security, medical evacuation, marsupial missions, and multiple vehicle control and vehicle–vehicle interactions. The initial focus is on the development of low-cost IMU/GPS (inertial measurement unit/global positioning system) units for position sensing, obstacle avoidance through innovative data fusion from a host of sensors, and path planning both from terrain models and vision based sensors. The test bed approach is to integrate the technology into such COTs machines as the caterpillar and ARTS units through a common architecture bolt-on package (Malhiot, 2002).
Navy and Marine Corps Initiatives
The Navy is involved in developing unmanned systems for air, sea, and ground. It is developing a separate UCAV for carrier-based operations. The Marine Corps developed the lethal UGV for ground combat known as “Gladiator” as part of the JRP. The Naval Research Laboratory conducts research in artificial intelligence and other technology areas relevant to autonomous systems on all platforms.
The Navy UCAV would provide a versatile, multipurpose vehicle capable of performing surveillance/reconnaissance and strike missions. The Navy’s UCAV effort leverages the Air Force development efforts discussed above, but there are differences in the operational environments that must be considered. Naval technological development addresses additional constraints, such as design compatibility with catapult takeoffs, carrier approaches, and arrested recoveries, among others. The Navy considers that the UCAV must be well orchestrated into the mix of aircraft and that the coordination of manned and unmanned aircraft must be seamless. As in the case of the Air Force UCAV program, it is important that DOD ensure a close coordination between the Air Force, Navy, and Army UAV programs, as well as between the UGV programs.
Artificial Intelligence Research
The Navy Center for Applied Research in Artificial Intelligence at the Naval Research Laboratory (NRL) conducts research in adaptive and autonomous systems, intelligent decision aids, and other technology areas with direct applicability to future autonomous systems. While research in voice and hand gestures for human control of robots is in its infancy, the Samuel intelligent testing system designed for software-based systems is now ready to support Army systems development (Meyrowitz and Schultz, 2002).
The National Aeronautics and Space Administration has conducted research and development and has deployed UGVs (such as the Lunar Rover) for use in the exploration of celestial bodies. The program uses technologies directly applicable also to Earth-bound UGVs, including perception and sensors, path planning, navigation, mobility, communications, and power/energy.
Much of the NASA research and development is accomplished at the Jet Propulsion Laboratory (JPL) in Pasadena, California, and JPL has also been involved in defense programs in these technology areas. As indicated below, there is strong research cross-fertilization among these technologies due to the number of agencies and contractors working on the various NASA and military UGV programs, most of whom are involved with more than one program.
National Institute of Standards and Technology Initiative
The National Institute of Standards and Technology (NIST) developed a “4-D/RCS” reference model architecture for behavior generation, world modeling, sensory processing, and value judgment processes used in the Demo III program (Albus and Meystel, 2001). NIST is also teamed with industry to develop a common software architecture that can be applied to unmanned ground systems of all services. This effort is called the Joint Architecture for Unmanned Ground Systems (JAUGS). The objectives of JAUGS are to define and model the unmanned systems domain and then standardize the interfaces and behaviors among software components. JAUGS will be applied to all UGVs developed under the JRP.
Department of Transportation Initiatives
The U.S. Department of Transportation (USDOT) from time to time has sponsored research relevant to UGV issues since the mid-1960s. The DOT-sponsored research has been
applied to vehicles intended to operate on roads or special guideways, and has not been applied to off-road operations. Hence, its applicability to Army UGVs is associated primarily with on-road operations. However, some of the vehicle positioning and sensing research can be more broadly applicable. Projects have been sponsored by:
Federal Highway Administration (FHWA)
National Highway Traffic Safety Administration (NHTSA)
Federal Transit Administration (FTA)
Intelligent Transportation Systems Joint Program Office (ITS-JPO)
California Department of Transportation (Caltrans) and
Minnesota Department of Transportation (Minn-DOT).
These projects have been aimed at supporting the development of a variety of ground transportation services:
Collision warning systems
Collision avoidance systems
Automated guideway transit systems
Automated highway systems and
Bus rapid transit systems.
Note that none of the vehicles that would use these systems are intended to be “unmanned,” but some of them are intended to be driven under completely automatic control. Others use sensors and user interfaces to assist drivers, but the same sensors could also be applicable to UGVs.
The customers for these systems are typically private vehicle purchasers (individuals and corporate fleets) and local public agencies (cities, counties, and transit districts), rather than the federal or state departments of transportation. These systems are developed by private vendors, and the major share of development costs is assumed by these developers rather than by the USDOT. Thus, the USDOT funding should be considered only the “seed funding” to advance the technology to the feasibility demonstration stage, and does not approach the total investments being made on these systems as they approach deployability.
Systems that are developed for commercial trucks or private passenger vehicles are intended to be mass produced in large quantities at low unit costs. These offer excellent opportunities for gaining production economies of scale that could benefit UGV systems using similar or related technologies.
The key technologies from these road transportation applications that could be applicable to UGVs are
Sensing of vehicle position relative to roadway lanes
Sensing of absolute vehicle position (global positioning system [GPS] and inertial navigation system [INS] in combination with others)
Sensing of proximity to other vehicles (ranges up to 150 m)
Identification and hazard assessment of targets within sensor range
Vehicle–vehicle data communication
Vehicle–roadside data communication
Vehicle dynamics control
Road condition sensing and estimation
Automatic steering control and
Vehicle-following speed and spacing control.
Department of Energy Initiatives
The DOE has mobile robotics programs at several of its national laboratories as well as within the DOE University Research Program in Robotics (URPR). DOE applications include security, environmental remediation, materials storage and monitoring, and response and cleanup of accidents involving nuclear materials.
At Sandia National Laboratories the Intelligent Systems and Robotics Center has developed several mobile robotic platforms ranging in size from smaller than 1 cubic inch to as large as a military wheeled vehicle (HMMWV) (SNL, 2001). Extremely small platforms were developed to establish the current limits in autonomous micromechanical systems that can be applied to covert surveillance missions. The larger platform Accident Response Mobile Manipulator System was developed for DOE accident response to nuclear weapons or other hazardous materials. Some of the other platforms such as Fire Ant, Dixie, SARGE (Surveillance and Reconnaissance Ground Equipment), Gemini, and Sand-Dragon were developed for military agencies. For example, SARGE was developed as a production prototype for the DOD Joint Program Office for Unmanned Ground Vehicles/ Systems, Gemini was developed for the Special Operations Command, and Sand Dragon was developed for the Marine Corps Warfighting Laboratory. Other mobile platforms such as the RATLER vehicles and the hopping robots have been used in such DARPA programs as TMR, Distributed Robotics, Software for Distributed Robotics (SDR), Mobile Autonomous Robots Software (MARS), and Self-Healing Minefield. In addition to designing, building, and testing robotic platforms, Sandia has considerable experience in developing distributed cooperative controls for mobile robotics. They have developed and demonstrated decentralized control algorithms for formation following, perimeter surveillance, facility surround, building search, minefield reconfiguration and healing, and chemical plume localization missions.
At Oak Ridge National Laboratory the Center for Engineering Science Advanced Research is developing autonomous multirobot learning algorithms for inherently cooperative tasks. They have been “studying autonomous
multi-robot learning for inherently cooperative tasks and have developed two new approaches to learning in the domain of cooperative multi-robot observation of multiple moving targets. These new techniques now allow robots to build up memories of their experiences in the environment, evaluate the utility of alternative cooperative actions, and then select actions to take that increase the likelihood that the desired global team goals will be achieved through the individual robot decisions. These multi-robot learning techniques are the first in the field that enable robot teams to automatically learn new inherently cooperative control tasks, rather than having to be programmed explicitly. These capabilities facilitate the solution to a wide variety of applications, including environmental cleanup, space exploration, military applications, and industrial operations” (ORNL, 2002). Oak Ridge has also participated in such DARPA mobile robotic programs as TMR and MARS.
At Idaho National Engineering and Environmental Laboratory the Remote, Robotics, and Automated Systems group is developing large and small robotics and automated systems to simplify efforts in the protection of DOE workers and the environment (INEEL, 2002). They are working primarily on robotics for mixed waste operations, deactivation and decommissioning of underground storage tanks, chemical analysis automation, and cooperative telerobotic retrieval. A small group of staff has also been involved with such DARPA mobile robotic programs as SDR and MARS.
Within the DOE URPR the University of Michigan has worked on mobile robot navigation and radiation mapping (<http://www.urpr.org>). Their work includes innovative mobile robot design, obstacle avoidance, and advanced mobile robot positioning (UMICH, 2001). The principal investigator, Johann Borenstein, has also been involved in DARPA’s TMR program and developed obstacle avoidance technology that allows mobile robots to navigate cluttered indoor environments filled with dense smoke.
Collaboration Among UGV Programs
Collaboration is achieved between UGV efforts because there are relatively few defense projects involving robotics technologies in general, and the field of unmanned systems is limited to a relatively small number of universities and companies. The small circle of robotics experts from academia and industry participating in multiple programs facilitates a common awareness of UGV advances and requirements.
Over the last 10 years DOD has initiated many Multidisciplinary University Research Initiatives (MURIs). Some of these have involved frameworks, models, algorithms, and software in the areas of perception, navigation, learning, and decision making in uncertain environments. There is also a thriving, worldwide university community in robotics, and progress has been made in multiple areas applicable to unmanned systems, including intelligent controls, robotic vision, computational geometry for intelligent systems, soft computing for intelligence augmentation, terrain modeling, and communication and control.
Because of their diversity, many of these crucial investments in research may not be fully utilized in the Army UGV program as it now exists. Further, pathways for rapidly transferring basic research knowledge to advanced technology test beds do not exist unless directly related to the Army through the particular MURI program. The mechanisms for technology transfer are inconsistent at best, with technology exchange meetings involving the university groups tending to be very perfunctory. This gap in the R&D continuum must be bridged to concentrate the spectrum of efforts that will develop UGV technologies and systems.
Many of the same industry teams that have participated in the Joint Robotics, Demo III, and PerceptOR programs, for example, are also part of the recently established Robotics Collaborative Technology Alliance (CTA). The JPL has supported NASA, DARPA, and Army robotics programs, so collaboration is high. Similarly, NIST participated with the Army in the Demo III program.
The committee sensed the competition among prospective government and industry participants as they vied for the PerceptOR down-select. There are a very limited number of UGV contracts. An unwillingness to share proprietary information in particular technology areas could easily offset the collaborative advantages of having a small playing field.
While having the same individuals participate in multiple programs may have advantages, it could prove a significant weakness if the demand for expertise increased dramatically or if the same few players monopolized and wittingly or unwittingly discouraged new participants. This assessment of interrelationships among the principal government agencies involved with UGV development provides basis for the answer to Task Statement Question 2.c in Box 3-3.
Technological collaboration between multiple programs will become more important, even essential, if the Army decides to focus its energy on developing a UGV for the FCS. The committee believes that it will take a designated advocate to do this, and a principal function of such an advocate will be to leverage UGV developments and promote collaboration toward explicit goals.
Automotive Industry Developments
Considerable effort is being invested within the automotive industry and related transportation organizations to develop systems that will enhance driving safety, assist drivers in controlling their vehicles, and eventually automate the driving as well. These activities, which are truly international in scope, offer the potential for technology spin-offs that could benefit the Army’s UGVs by lowering costs and accelerating the availability of components and subsystems. These could include sensors, actuators, and possibly software, control, and communication systems as well.
BOX 3-3 Task Statement Question 2.c
Question: How do the Army UGV efforts interrelate with other government ground robotics initiatives (e.g., National Aeronautics and Space Administration [NASA] rovers, Department of Energy [DOE] programs, Defense Advanced Research Projects Agency [DARPA])?
Answer: With the exception of DARPA (the Army funds several of the DARPA robotics programs) interrelationships of other UGV efforts with those of the Army are informal and unstructured. The small size of the robotics industry and the small number of robotics experts tend to encourage technical collaborations. The Jet Propulsion Laboratory, for example, has supported NASA, DARPA, and Army programs, so collaboration is high. Similarly, NIST was part of the Demo III program. Collaboration in particular technology areas may be inhibited by intense competition for a limited number of UGV-related contracts.
Automotive Night Vision Sensors
General Motors introduced the first automotive night vision system on its Cadillac deVille in the 2000 model year, and has sold them as fast as its supplier Raytheon has been able to make them (6,000 per year) (Scientific American, 2001a). This is a passive infrared system that detects infrared (IR) emissions from objects in front of the vehicle and projects the IR image on a head-up display. The system is being sold for $2,250 retail, and considering typical automotive mark-ups, this implies that the system is being supplied from Raytheon to GM for a little more than $1,000.
More recently, DaimlerChrysler in Germany has announced that it is developing an active IR night vision system that depends on active IR illumination of the driving scene by the vehicle (Scientific American, 2001b). This is claimed to offer the ability to “see” more objects that are the same temperature as the background and at a greater distance.
Ultrasonic Proximity Sensors
Various manufacturers in the United States, Japan, and Europe are offering parking assistance systems that warn drivers when they are approaching too close to obstacles at very short range (up to 1.5 meters). These systems typically use arrays of up to eight ultrasonic ranging sensors surrounding the vehicle at bumper height, and retail for up to $2,000. To be detected the sensing requires active ultrasonic emissions; it is intended for very short ranges and is therefore only applicable for very low-speed maneuvering.
Intermediate-Range Ultra-Wideband Radar Sensors
Several automotive companies are developing ultra-wideband impulse radar sensors for use at intermediate sensing range (perhaps 15 meters) to detect other vehicles that could represent hazards. These are not yet on the market and do not yet have FCC approval, but if the sensors come to market they could represent an inexpensive way of achieving wide-angle detection of hazards surrounding a vehicle at intermediate range.
“Long-Range” Automotive Radar Sensors
The automotive industry is beginning to provide its customers with adaptive cruise control systems (cruise control that can adjust speed to follow another vehicle at a suitable distance) and forward collision warning systems. These are based on use of forward-looking IR laser or millimeter wave (24 or 77 GHz) radar sensors, which typically have a narrow field of view (perhaps 12 degrees) and a range of 100 to 150 meters. The signal processing of the sensor systems is designed to discriminate “other vehicle” targets in a road environment with considerable clutter (e.g., bridges, signs, roadside lighting fixtures, and vegetation). Considerable adaptation would be required for use in an off-road battlefield environment; however, they have the advantages of multiple suppliers and probably rapidly decreasing costs. Current fully integrated systems, with the interfaces to throttle and brake and the HMI, retail for up to $3,000 on luxury cars and heavy trucks, but the prices are expected to decrease significantly in the next few years.
Lane-Tracking Vision Sensors
Because road departure crashes are a serious cause of death and injury on our highways, there is considerable interest in developing systems to warn drivers of imminent road departures. The most common technology for detecting this is machine vision, based on use of a small chargecoupled device (CCD) camera capturing the image of the road ahead of the vehicle and identifying the lane markings. The technology could be applied to Army UGVs that are intended to follow roads or well-marked trails for supply missions but would not be transferable to the more general off-road environment. Two American companies, AssistWare and Iteris, are marketing systems for use on heavy trucks at prices in the $2,000 to $2,500 range, and other systems have been made available on high-end passenger cars in Japan.
Drive-by-Wire Actuation Systems
Automotive vehicle designers are gradually adopting drive-by-wire technology for actuator systems for reasons unrelated to vehicle automation. Throttle by wire is avail-
able on some high-end passenger cars, although the analogous capability is already standard on most modern heavyduty trucks. Electronically assisted power steering has been used on the Acura NSX sports car for several years and will soon be available on less expensive cars. Brake-by-wire are being introduced on new Mercedes-Benz automobiles, and other manufacturers are likely to follow. The motivations for these introductions have been associated with providing a higher degree of control, simplifying the installation of the systems in the vehicles, and saving energy. However, ancillary benefits can be gained by making the vehicles more amenable to automation. As the actuation systems are proven for automotive use, durability and robustness will become well established and prices will decline with volume production. This could make them promising candidates for use on Army UGVs.
Vehicle-to-Vehicle Wireless Communications
Interest is growing in the automotive world in the possibilities for wireless communication of data among vehicles to enhance driving safety and to provide new traveler information services. Recent work on standardization of vehicle–roadway communication (dedicated short-range communications [DSRC], at 5.9 GHz) has opened the door to inclusion of vehicle–vehicle communications within the same spectrum allocation and devices. The IEEE 802.11a R/A wireless standard is being adopted for vehicular use. This means that within the next few years the equipment for vehicle–vehicle communication could become very inexpensive and widely available. The DSRC standard will be applied to several different ranges of operation, perhaps extending as far as 1,000 meters, and with a sufficiently general geographic (i.e., GPS-based) addressing scheme it could be applied to any off-road battlefield environment for communications among UGVs and even between UGVs and nearby UAVs.
Automatic Vehicle-Following in Convoys
Several automotive research projects have developed approaches for enabling vehicles to follow each other automatically at close separations on highways. This could be applicable to Army supply convoys operating on-road or possibly even off-road. In the on-road applications it is necessary for the vehicles to operate very close together in order to produce the primary benefits (fuel savings and better utilization of highway infrastructure capacity) and to prevent nonequipped vehicles from cutting in front of the follower vehicles. In a battlefield environment it may be desirable to operate the vehicles much further apart, which introduces a contrasting set of system design requirements. The primary research on automatic vehicle-following for convoys of road vehicles has been performed by DaimlerChrysler in the CHAUFFEUR project in Germany (Schulze, 1997) and by the University of California PATH program (Rajamani and Shladover, 2001).
Foreign Government UGV Activity
A number of foreign governments sponsor research and development in robotics, including unmanned ground vehicle systems. The level of military interest ranges from high in some cases to no active interest. The military development paths include dedicated military research and development (R&D) programs as well as the application and exploitation of commercial robotic technologies. For many years commercial industries have developed robotic devices to relieve workers of repetitive and labor-intensive tasks; Japan is a notable example of commercial applications. Examples include such tasks as welding in difficult locations, spray painting, and repetitive assembly processes.
Potential military applications include a similar objective of relieving personnel of such repetitive and labor-intensive tasks as ammunition handling and loading, but more importantly the replacement of soldiers in hazardous tasks such as mine clearing, obstacle breaching, and disposal of unexploded ordnance. As semiautonomous and autonomous unmanned ground vehicles are further developed, it will become possible to replace personnel as well in such noncombat tasks as guard duty and logistic vehicle driving. Most importantly, as robotics technologies continue to develop both in commercial and military programs, the prospect of enhancing individual soldier performance becomes a real possibility.
The range of military mission areas under investigation in various countries is significant:
Reconnaissance and surveillance: France, Germany, Great Britain, Israel
Mine clearing and ordnance disposal: France, Germany, Great Britain, Israel
Logistics: Great Britain, France, Germany
Targeting: France, Germany, Great Britain, Israel
Unmanned weapons platforms: France, Germany, Great Britain
Camouflage, concealment, and detection: France, Germany, Great Britain.
It should be noted that commercial efforts in foreign countries, particularly in automotive industries such as Japan’s, continue to make significant advances that contribute to general progress in robotics technology. Presently, the United States has technology agreements with France, Canada, the United Kingdom, Israel, and Germany.
Examples of several foreign UGVs described below illustrate that the efforts are similar to those being investigated in the United States (Hutchinson, 2001).
There are a number of foreign teleoperated vehicles. The Israeli Pele is a tank-mounted mine-clearing and breeching
vehicle. Several vehicles by Great Britain include the Bison and Groundhog ordnance disposal vehicles, the Armored Vehicle Royal Engineers (AVRE) mounted on a Chieftain chassis and the Combat Engineer Tractor (CET). The British MARDI (Mobile Advanced Robotics Defense Initiative) test bed is teleoperated with optical fiber, intended for possible applications in RSTA and smokescreen operations.
The German PRIMUS (Program of Intelligent Mobile Unmanned Systems) mounted on the 4-ton Wiesel vehicle is teleoperated with some semiautonomous capabilities. A cooperative program is being conducted with the U.S. Army Research Laboratory on the auto-navigation system.
The French SYRANO (Systeme Robotise d’Acquisition pour la Neutralisation D’Objectifs) is teleoperated with optical fiber and with some semiautonomous capabilities. Possible applications include RSTA.
The foreign UGV activities are focused principally on platform-oriented R&D and on advanced concept demonstrations. Teleoperated vehicles are the most developed and are expected to be the principal applications in the next 5–15 years, with an increased application of semiautonomous ca
BOX 3-4 Task Statement Question 3.c
Question: Are there foreign UGV technology applications that are significantly more developed than those of the United States that, if acquired by the U.S. government or industry through cooperative venture, license, or sale, could positively affect the development process or schedule for Army UGV systems?
Answer: No. Based on the information available to the committee, there are no foreign UGV technology applications that are significantly more advanced than those of the United States.
pabilities, based on continuing development, particularly in computational power and in command and control capabilities. It is anticipated that smaller vehicles with limited intelligence will be utilized. Beyond 15 years increased autonomy is anticipated.
This section provides the basis for the answer to Task Statement Question 3.c in Box 3-4.