Technology Assessments and Forecasts
In the request that initiated the STAR study, the first item was to identify the advanced technologies most likely to be important to ground warfare in the twenty-first century. The STAR study included eight technology groups that focused on particular areas of technology. These groups assessed the state of the art and forecast the technology that was likely to be available within 10 to 15 years, so it could be included in Army systems by 2020. All advanced technologies with major Army applications were divided into the following eight technology groups:
Computer Science, Artificial Intelligence, and Robotics;
Electronics and Sensors;
Optics, Photonics, and Directed Energy;
Biotechnology and Biochemistry;
Propulsion and Power;
Advanced Manufacturing; and
Environmental and Atmospheric Sciences.
Each group reported its work in a Technology Forecast Assessment (TFA).
After the TFAs for the eight areas had been prepared, a panel drawn from the Science and Technology Subcommittee met to forecast potential long-term trends in research that might not produce useful technology until well after the 10 to 15-year time horizon of
the other TFAs. The eight area-specific TFAs and the Long-Term Forecast of Research are bound together as a separate volume of the STAR publications.
In this chapter the STAR Committee summarizes what it considers to be the key findings of the Long-Term Forecast Panel and the technology groups, with particular emphasis on their responsiveness to the STAR mandate. The summaries are organized by sections corresponding to the individual reports. For the sake of brevity, much supporting detail has been omitted; the STAR Committee urges readers interested in particular findings to study them in the context of the full report.
LONG-TERM FORECAST OF RESEARCH
Scope of the Long-Term Forecast
The Long-Term Forecast of Research represents the best guesses of a panel of experts on the directions in which technology of interest to the U.S. Army may progress during the next 30 years or more. The principal objective of this report was to highlight significant trends rather than forecast specific technological advances. The forecast panel identified 11 major trends that cut across the traditional boundaries between scientific or technical disciplines. These are discussed below as major multidisciplinary trends. In addition, a number of narrower discipline-specific trends within specific technology areas will have important consequences for future Army applications. In many cases these trends, which are summarized here, tie in with one or more of the major trends.
Management of Basic Research
The long-term forecast panel agreed that continued support of Army basic research (funding line 6.1) will be necessary if these research trends are to find fruition in Army-specific applications. Budgetary continuity and stability are crucial to achieving long-term objectives.
Major Multidisciplinary Trends
Trend 1: The Information Explosion
The flow of information in preparation for ground warfare and during battle will continue to increase as intelligent sensors, unmanned
systems, computer-based communications, and other information-intensive systems proliferate. Data bases and their management software will progress beyond even object-oriented data bases to third-generation data bases with new modes of indexing stored data and more intelligence in interacting with the human user of the data base. Mixed machine-human learning will team the learning capabilities of a person with the rapid data-processing and analysis capabilities of a computer.
The current limitations to practical application of artificial intelligence may be overcome if an adequate theory of representation creation can be developed and action-based semantics can be applied to the Army's battlefield information requirements. The in formation transmission bottleneck on the electronic battlefield calls for data compression techniques; semantics-based information compression would address this problem by assessing the value of information relative to the cost of transmitting or storing it.
Trend 2: Computer-Based Simulation and Visualization
Computer simulation of objects and processes, with graphical display of the computer-generated results, gives researchers a potent addition to the more traditional techniques of theory development and experimental evaluation. While computer simulation clearly depends on progress in computer hardware and mathematical algorithms, its growth also depends on understanding the basic principles governing the phenomena to be modeled. Long-term progress in integrating computation with science and engineering may require a broad-spectrum physical modeling language, rather than special-purpose simulation environments. Computer studies have already played a major role in modeling the behavior of nonlinear dynamic systems. This area of applied mathematics presents both limitations and opportunities for computer modeling of processes important for Army technology. For example, computer modeling will make possible detailed studies of how physical signals, such as light, radar, or sound, propagate in inhomogeneous media, such as the lower atmosphere or through forest canopies. In chemical research, the potential energy surface that characterizes a chemical reaction is a multidimensional mathematical function, which can be modeled and visualized for the researcher. But better methods are needed to approximate the relevant properties of complex molecular systems, and models are needed for reactions of
particular interest to the Army, such as combustion or detonation reactions at the surface of an explosive.
Trend 3: Control of Nanoscale Processes
As the features of microelectronic devices shrink to sizes measured in nanometers, new phenomena appear that alter how these devices behave. The particle-wave duality of this quantum world affects both physical and chemical behavior. For example, electron transport, which is essential to all electronic devices, becomes quantized at this scale. Structures no longer behave independently of neighboring structures; quantum mechanical phenomena such as quantum interference, tunneling, and ballistic transport occur. These changes set limits to the miniaturization of conventional semiconductor devices, but they also open opportunities for entirely new devices, such as atom clusters.
Natural biomolecules such as enzymes, or variations bioengineered from them, are likely to provide the first generation of molecular recognition devices. Such a device will detect a single molecule of a particular chemical species or with any of a class of molecules with specified structural similarities. Nanoscale chemistry will also control surface reactions, including surface catalysis, through the design and production of layers having an exact placement of component atoms, ions, and molecules.
These new ''nanoelectronic'' devices will operate at very low voltages and low currents; only a few electrons will suffice to differentiate between the 1 and 0 states of a binary digit. As the technology for quantum-based devices becomes available, subsequent steps will be to integrate them into "molecular" integrated circuits, then into monolithic integrated circuits (wafer-scale integration), which could conceivably have a trillion "devices" on a chip the size of a dime.
Trend 4: Chemical Synthesis by Design
This trend joins with trends 6 and 9 in an even more general trend: in the future, new materials will be designed at the molecular level for specific purposes, by designer-engineers using fundamental scientific relations between a structure and its functional capabilities. The realm of engineered chemicals will include both surface catalysts and enzyme-like catalytic molecules, whose specificity depends on their three-dimensional conformation. To support research into these structure-function relations, chemists will need to determine, by experiment and by derivation from quantum chemical theory, the three-dimensional
structure of complex molecules, including biomolecules. These structure determinations must be both rapid (on the order of hours or days) and at high resolution (on the order of angstroms).
Trend 5: Design Technology for Complex Heterogeneous Systems
If a system has many components and subsystems that vary markedly in physical and operational characteristics but must act as a functionally coherent whole, it can be considered a complex heterogeneous system. Modern combat vehicles, unmanned air vehicles carrying multiple smart sensors, and a theater air/missile defense system are all examples. At present, the design of such systems is largely a process of muddling through to an adequate result rather than a rational procedure derived from a testable theory. The mathematics of optimization theory can be improved but probably needs to be supplemented, or even supplanted, by other approaches. New approaches are needed for designing systems with robustness with respect to variation while taking into account the costs and benefits of marginal design information.
Statistical approaches that seek "least-sensitive" solutions for a complex design problem hold some promise. But, they currently lack a clear theoretical foundation and may not apply if the system's behavior is nonlinear over its operating range. A radical departure would be to model the design process itself, rather than attempting to model the system to be designed. Another area worth exploring is the use of nonlinear modes of control for systems whose functional dynamic range includes areas of nonlinear response.
Trend 6: Materials Design Through Computational Physics and Chemistry
This trend combines, within the field of materials science, two other trends: the growth of computer simulation (trend 2) and the design of useful products by application of fundamental relations between structure and function (trend 4). For materials design, these structure-function relations include interatomic forces, phase stability relations, and the reaction kinetics that determine how complex processes evolve. Possibilities of interest to the Army include light-weight (half the density of steel) ductile intermetallics, new energetic materials superior to current explosives and propellants in energy density and safety, materials harder than diamond, and tough polymers with working ranges extending to 500°C.
Trend 7: Use of Hybrid Materials
Also called composite materials, hybrid materials are especially attractive for Army applications because they can be designed for unique and special requirements. For example, the component phases of a hybrid can be altered, or the formation process can be modified, to improve performance in two or more dissimilar functions. The area of greatest technical novelty is that of smart structures. A network of sensors embedded in the structural phase of the composite acts like the sensory nerves of an animal's nervous system. A network of actuators allows properties of the structure to be altered, under the control of a microprocessor that reacts to the sensor signals, analogous to an animal brain.
Trend 8: Advanced Manufacturing and Processing
The above trends in designing materials, particularly hybrid materials, will be paralleled by trends in manufacturing fine-scale materials (at the scale of individual atoms) and thin-layer structures. Chemical synthesis methods such as sol-gel processing will be used, as will methods for controlling process energy precisely, such as laser processing. As nanoscale devices (trend 3) become available for sensors and actuators in hybrid materials, smart materials will be synthesized at a molecular level through application of principles such as self-assembly and molecular recognition. These principles were first studied in biological systems.
Trend 9: Exploiting Relations Between Biomolecular Structure and Function
The principles that relate the functions of biomolecules and tissue structural components to their molecular structure are now well enough understood to be used in designing materials. Among the potential applications are new battle gear for the soldier made from lighter and stronger fabrics, broad-spectrum vaccines and prophylactic medicines, sensors and diagnostic devices based on molecular recognition properties, and miniature motors and power supplies based on biological energy transduction mechanisms.
Trend 10: Applying Principles of Biological Information Processing
Biological systems receive, store, duplicate, respond to, and transmit information. The knowledge we have gained about the mecha-
nisms through which this information processing occurs will find practical applications. In the design of information systems, capabilities such as pattern recognition and selective abstraction of relevant data may use principles discovered from biological systems. Biological structures, natural or bioengineered, may be biocoupled with electromechanical and optoelectronic components (Figure 3-1). At even higher levels of information processing, a growing understanding of the biological basis for learning and memory may provide new models and techniques to improve training and performance for information-intensive tasks.
Trend 11: Environmental Protection
The Army will be affected by the general societal trend toward greater concern over environmental effects of toxic materials or disruptions of ecological balances. In the future, the Army will have increased responsibilities for ameliorating past environmental damage and minimizing new environmental contamination or degradation from its operations. Assessing the full impact of hazardous wastes, for example, will require development and verification of accurate models for the transport and fate of the target compounds in soil, air, water, and biota. Better methods to monitor and treat waste materials will be required.
In electronics, optics, and photonics, the directions for advanced sensor technology include conformal sensors and multispectral sensors, with onboard processors for data fusion and for mission-specific processing such as automatic target recognition. Future Army systems will use an integrated mixture of electronic, photonic, and acoustic devices to process both analog and digital output from a range of sensors gathering electromagnetic, acoustic, and magnetic signals (Figure 3-2). Active cancellation techniques will be used to reduce interfering background "noise" and unmask sources of interest. Extensive communications networking will require communication links with very wide bandwidths. Allied with the major trend in fine-structure manufacturing (see trend 8) will be advances in micropackaging and minifabrication of components, subassemblies, and entire nanoelectronic systems (trend 3). Methods for control of optical phenomena will provide faster, smaller, and more powerful architectures for digital data processing as optoelectronic technology expands.
In aeromechanics, computer simulations on new supercomputer
architectures will allow modeling of rotorcraft vehicles in their operating environment. This greater computing power, combined with advances in computational fluid dynamics, composite structural dynamics, and aeroelasticity will contribute to the goal of complete aerostructural simulation (another example of trend 2). Propulsion and control technologies will make hypervelocity projectiles and missiles possible. More knowledge will be needed of phenomena associated with hypersonic passage through the lower atmosphere, electromagnetic radiative characteristics of hypersonic vehicles, and the impact and penetration by hypervelocity projectiles against anticipated targets. If unmanned air vehicles become important means for transporting sensors and as brilliant weapons, the Army will require theoretical and experimental data on aerodynamics at low Reynolds numbers.
In molecular genetics, information deciphered from both human and nonhuman genes will have major implications of interest to the Army. The genetic blueprint information from nonhuman cells will be used in bioproduction of artificial products that mimic natural
materials and in the design and production of organisms with new or modified properties. Information about the human genome will yield new methods for preventing and treating diseases or the effects of CTBW agents. Artificial blood, skin, and bone, and perhaps even complex organs such as the liver or kidneys may be replaced by culturing an individual's own cells.
In clinical medicine, new instruments and sensors will be used in diagnostic and therapeutic equipment. The miniaturization of sensors (see major trend 3) and sensor data fusion will allow physicians to measure chemical and physiological events at the cellular and subcellular levels as they happen. Army applications include detection of CTBW agents in the field, monitoring of soldiers' physiological condition, and improved diagnosis and resuscitation of the wounded and sick while they are in transport.
In atmospheric sciences, high-resolution remote sensing of meteorological conditions will provide the data to initialize and validate computer models of the atmosphere on small spatial and temporal scales, for which the Army has special need. The validated computer model can then be used to improve sensor placement. By repeating this cycle, the sensor data-gathering and computer modeling activities will complement one another. The result should be increased understanding of small-scale weather conditions, including fog and cloud physics and more accurate representations of turbulence.
In terrain sciences, sensor technology and information processing are again important, for both automated extraction of information from multiple imaging and three-dimensional representation of terrain data. A key addition to existing terrain data capabilities will be a near-real-time system to analyze and map changes in terrain surface conditions and trafficability. Such a system would use sensor data on rainfall, soil moisture monitors, and computer modeling of soil properties based on hydrologic and atmospheric conditions.
COMPUTER SCIENCE, ARTIFICIAL INTELLIGENCE, AND ROBOTICS
The Computer Science, Artificial Intelligence, and Robotics Technology Group assessed the following technologies:
Integrated system development includes system development environments, design languages and compilers, problem-solving
strategies, simulation and optimization (in development), and the mathematics for representing and managing variation.
Knowledge representation and languages includes mathematical representations of information and special-purpose languages, such as battle control languages.
Network management concerns the management of multiple processors that pass digital data or other information (such as voice messages) to one another through interfaces.
Distributed processing is the execution of a computation (a program or a number of computationally independent programs) on two or more processors. Usually the processors are part of a network.
Human-machine interfaces include graphic displays, keyboards, control consoles, pointing devices, printers, audio outputs, and other means by which a computer or peripheral communicates to the human user or the user communicates to the machine.
Robotics includes stationary and mobile systems, airborne or ground-based, that are controlled by onboard computer programs. They may be (1) autonomous, (2) supervised by an operator but operating autonomously for routine operations, or (3) under continual operator control (tele-operated). Their mission may require sensors and communication capabilities only, or they may have advanced processing and even weapons capabilities.
Technologies to monitor are areas in which the Technology Group thought that nonmilitary R&D would lead the way and the Army could profitably use the results without funding research itself. These areas include machine learning and neural nets, data base management systems, ultra-high-performance serial and parallel computing, planning, manipulator design and control, knowledge-based systems, and natural language and speech.
The battlefield of 2020 will use millions of computer systems and components. These systems, ranging from tiny microprocessors embedded in weapons to mobile command-and-control centers, will be ubiquitous, critical, and essential. They will be interlinked by a wide range of communications media.
The effectiveness of individual soldiers in the future will be enhanced by computational tools that give them constant access to command-and-control centers, help them navigate, monitor their
physical condition, and provide an instant source of up-to-date knowledge in the form of "smart manuals."
Data bases, nearly instantaneous communications and analysis, intelligent decision aids, and multisensory information displays will provide commanders with an unprecedented awareness of the battlefield (Figure 3-3). While the potential for a good commander to affect the outcome will be multiplied, so will the potential for command errors to prove disastrous.
For logisticians, the most significant changes will be in the planning and control of logistics operations.
For strategic planning, warfare fought with computers and unmanned systems may become at once more common and more threatening. The publics of advanced nations may find war more acceptable if the number of casualties can be kept low. The rewards of aggression may be higher because of the aggressor's ability to exploit a temporary advantage in system sophistication.
Computers, data bases, and software will themselves become targets in warfare. Our computational resources must be protected while exploiting any vulnerability in the opponent's systems. This computer
warfare will involve at least four components: information security; injection of, and protection against, electronic viruses; sabotage; and exploiting the computational predictability of the opponent's systems.
Integrated System Development
The Army's development problems dwarf those of any other U.S. organization, governmental or private sector. For example, the Army will need to learn how to structure the simultaneous development of (1) the systems themselves, (2) the civilian surge capacity to produce them in large numbers without major peacetime investment, and (3) the doctrine to use their often revolutionary capabilities.
If the individual technical areas that contribute to integrated system development are considered separately (e.g., software engineering, electronic systems, or mechanical design), the Army could follow the lead of private sector developments. However, where these technologies relate to one another and where they affect tactics and training, the Army will need to lead.
Civilian developments can accelerate the introduction of technology into Army applications and reduce costs. Opinions vary on how much of the computer and systems hardware produced for civilian markets is too fragile for direct Army use. However, some civilian items will not meet minimum functional specifications, so it will be necessary to analyze carefully the effective use of components and systems developed for the commercial market. In some cases it may be necessary to redevelop them to military specifications; in many others (e.g., microcomputers), a preferable route is to provide an environment in which commercial items work well enough.
The 2-year cycle for computer obsolescence and the vulnerability of unmanned systems to countermeasures will make it infeasible to maintain constant fielded superiority over every potential threat. Systems to meet potential threats will need to be designed but left unfielded unless the corresponding threat materializes.1
The successful use of high-level design languages, including compilers to translate high-level design into detailed descriptions, will depend on satisfactory answers to two questions: Does the description
accurately reflect what the designer wanted it to? Is the high-level description correctly translated into an acceptable implementation?
The Technology Group forecasts that software design will remain a difficult problem in 2020, but very-high-level languages will have shifted most software development out of the hands of programmers and into those of subject-area experts (Figure 3-4). The Group also forecasts that by 2020 the Army will have the infrastructure, design tools, languages, and silicon foundry engineering to deploy to forward maintenance depots a silicon compiler for automated design and production of VLSI chips.
Problem solving strategies in artificial intelligence are beginning to progress beyond heuristic classification, which seeks a valid solution from a fixed set of possibilities, to heuristic construction, which creates a complex solution from incremental subsolutions. To apply this emerging technology to the Army's complex design problems will require further advances in (1) conceptual modeling of application domains, (2) abstraction hierarchies, (3) representations for partial design information that capture dependencies, and (4) problem-solving strategies for high and low levels of abstraction.
In the area of design simulation and visualization tools, by 2020 high-resolution simulations will be calculable from first-principle nonlinear equations and physical relationships. Real-time, interactive simulations of complex systems and operational environments will be achievable by networking thousands of individual simulators.
Optimization programs will be available to vary the parameters of a simulated design to seek the best set of design parameters. However, this optimization approach will still depend on having an initial design to simulate; the form in which a design problem is represented can control its solutions.
System developers must reason about sets of objects under sets of conditions. Mathematical representations of variation, both probabilistic and nonprobabilistic, will therefore be needed for integrated system development (as well as for other application areas). The Army will need to participate in this research because the battlefield, as a source of deliberately antagonistic variation, is unlike the environment faced by private sector developers.
Knowledge Representation and Languages
Advances in knowledge representation will increase the reliability of software by providing formal structures and mathematics to describe key information about the battlefield, such as terrain, the degree of certainty about the enemy's forces and intentions, and sets of potential outcomes.
The search for new mathematical representations of knowledge is being driven by the computer. Often, the needed representations do not exist yet.
The Technology Group envisions a battle control language that will give commanders control of computational power analogous to the control that current spreadsheet packages give users to ''program'' their own calculations and tables. The high-level language will use statements that look like operations, orders, unit TOEs (Tables of Organization and Equipment), and map graphics. The language
will let commanders control, interrogate, and understand a nearly instantaneous information flow about unit status and logistics. Incoming intelligence will be correlated and displayed in seconds. Continuous simulations, which will run in the background, will be used to test alternatives. Broad mission orders from the commander will automatically generate implementing instructions to units.
The networks of 2020 will carry voice and data at high data rates. Connection into the network will be available anywhere in the field. However, some of the unsolved problems of network management today are likely to continue. These problems include delay, redundancy, and priorities.
Managing communications bottlenecks within networks and across networks connected by gateways will be a critical task. Problems arise when different networks with different access schemes, protocols, and security levels communicate with one another or pass message traffic on to other networks.
The Army has a useful role to play in solving these network management problems. It can offer prototype and experimental environments where new approaches can be stress tested.
Today, the level of distributed processing has advanced to the point that a new application process needs to interact only with the operating system or network protocol, one step above interaction at the hardware level. The Technology Group forecasts that this level of required interaction will advance by 2020 so that applications will interact at a level of abstraction (meaning) that is far above the hardware level. This level of interaction will enable more powerful application programs for sensor fusion, situation assessment, operations planning, and sophisticated modeling.
The technology to support this future level of distributed processing will include massively distributed worldwide networks; dynamic, real-time protocols for distributed operating systems; and distributed object data base management systems.
The Army's use of distributed processing will face more difficult obstacles than occur in most civilian environments: (1) continually varying prioritization of processes; (2) robustness when large parts of the distributed system disappear without warning; (3) a vast range
of data types, inference types, and hardware; and (4) accommodation of several levels of security and access.
Given the potential for information overload, a crucial requirement will be to provide the human users of a system, particularly commanders, with the information they need without overwhelming them. The difference often depends on good interface technology.
In the area of human factors, technological advances will occur in visual display techniques, force-feedback controls, information presentation optimized for low data transmission rate, and workload optimization for control of multiple systems by a single operator.
Heads-up displays, which project an image from a lipstick-size tube onto eyeglasses, are just appearing as commercial products. By 2020 stereo heads-up displays will be standard.
The Technology Group projects that by 2020 interface media and modes will be customized to the user's job, expertise, and personal preferences. One operator will be able to control multiple systems simultaneously and efficiently. Controls will switch between pure program control, operator monitoring, and operator control, according to circumstances. Structured voice will be the standard mode of entering responses to option menus. Other input modes may include analysis of facial and body gestures from TV images or direct monitoring of physiological responses (although the semantic content of such input will be far less sophisticated than linguistic expression).
In the hypermedia area, users will be able to navigate consistently through many different kinds of information, including drawings, photographs, video, synthesized voice, and diverse textual formats. The technology and human factors expertise for hypermedia is just beginning to emerge.
The Technology Group forecasts that users of map information in 2020 will be able to switch freely between, or superimpose, symbolic data and simulated or real scenes. Synthesized speech will supplement visual displays. Computer-aided drawing will convey nonverbal information.
The core weapon of land war in the twentieth century has been the tank. The core weapon in the twenty-first century may well be the unmanned system, operating mostly under computer control with human supervision. Robot systems may be classified according to the
level of continuous operator control during the system's operation. Fully autonomous robots perform their tasks with no human interaction following mission assignment (i.e., after they are programmed for a task). Supervised robots can perform most of their assigned task autonomously but require interaction with their human operator from time to time or when special situations arise beyond their programmed capability. Operator-controlled robots require interaction with a human controller at frequent intervals. While one operator can control several (perhaps many) supervised robots, an operator-controlled robot in practice requires the full attention of its human controller. The term "tele-operated" may apply to either supervised or operator-controlled robots, as distinguished here.
The Army's requirements for successful battlefield robotics are unparalleled in the private commercial sector. Robots are feasible on production lines because variability in the environment and range of stimuli can be contained. However, variability on the battlefield is uncontainable to the extent that the enemy can affect it. Even by 2020, unmanned systems will probably still be less capable than manned systems. They will be useful because, if properly designed, they can be far more numerous than manned systems. The highest payoffs from battlefield robots will come from putting large numbers of sensors in places where soldiers should not go and from integrating the information from the sensors into a coherent picture.
The Technology Group forecasts that a more relevant model for conceiving of battlefield robots is the land mine rather than the human soldier (Figure 3-5). That is, military robots will evolve as "smart mines," with increasingly sophisticated sensors, weapons, and modes of propulsion, rather than as "mechanical foot soldiers." The sensors, weapons, and propulsion methods used by these robot mines will differ greatly from those used by soldiers.
To defeat enemy attempts at deception, battlefield robots will have to integrate a wide variety of sensor information. The Army of 2020 will have vast requirements for signal processing from a single sensor, sensor fusion, and sensor integration.
For robot weapons, the most practical concept may be explosively propelled projectiles that achieve armor-piercing velocity with low weight and cost. By 2020 single-missile robot tubes, hidden in ground cover, are likely to be more secure than missile batteries mounted on vehicles.
The trade-offs among range, cost, and flight time of robot weapons, whether from airborne or ground-based launchers, imply that a mix of systems is preferable to a more complex, all-purpose system. A typical unmanned system should be specialized. Properly con-
ceived, battlefield robots can be inexpensive and quickly developed, yet they will remain vulnerable to countermeasures and can rapidly become obsolete.
The mechanical issues of robot vehicle mobility are quite different for air and ground systems. Large robot air vehicles, whether autonomous (e.g. cruise missiles) or tele-operated (drones and remotely piloted vehicles), are well established already. The cost and bulk for terrain mapping and related navigational computation for airborne vehicles will continue falling. By 2020 the Technology Group foresees the possibility of building actuators, sensors, and computers on a single silicon chip.
Mobility for small ground systems over natural terrain is a major development challenge. The simplest mode, mechanically and computationally, is leaping followed by reorientation. The use of mechanical legs for walking or running motion will be more difficult, although the Technology Group expects both running and walking legged robots to be in use by 2020. Other options are wheeled or tracked vehicles.
The packaging of huge numbers of components into a mission-specific configuration is another key technological requirement for battlefield robotics to succeed. The Army may need to play an active role in supporting or conducting research in packaging.
Technologies to Monitor
Machine learning. Recent research in this area has produced a large number of software systems, most of which are tailored to one learning paradigm. During the next 30 years, integrated learning systems, designed for general learning problems, will gain more research attention. These systems will be able to adopt different learning strategies, depending on the problem at hand. However, there is still no general theory of learning. If the strategy is poorly suited to the problem, unwanted or incorrect generalizations can result. Another difficulty is that learning systems may be difficult to debug or to modify for changed circumstances.
Neural nets. The Technology Group views neural nets as a particular mechanism for machine learning. Their architecture seems closer to biological computation (animal and human brains) than do conventional programmed architectures. But the Group sees no convincing argument to conclude that this resemblance will make neural nets superior in "intelligence." They have, however, achieved some spectacular successes in pattern recognition, which could make them useful to the Army.
Data base management systems (DBMSs). Existing relational DBMSs will be superseded by knowledge-based, hyperdocument DBMSs, which will be able to represent complex data structures such as tables, large text documents, images, and maps. Also important will be a move to fully distributed DBMSs with automatic updating, maintenance, and dynamic optimization of storage location.
Ultra-high-performance parallel and serial computing. The modeling and simulation needs of the Army will continue to make use of the latest and fastest computers. The two major applications are scientific simulation as part of the development process and simulation of combat activities or wargames. Parallel architectures seem to be the only promising route for continuing the past rate of growth in computing power. Whether the conventional supercomputer design, which is based on vectorization of the computational problem, or radical alternatives such as massively parallel architectures is optimal appears at present to depend on a case-by-case fit of the problem to be solved, the software program that solves it, and the machine architecture.
Planning. By 2020 planning technology will be capable of producing complex plans in complex domains. The most dramatic
progress will be in the breadth of knowledge that is brought to bear in generating or revising a plan. Memory of past events, including plan successes and failures, will be used. Contingency alternatives will be explored to greater depth before an optimal course of action is selected. Revision and adaptation will be faster.
Manipulators. By 2020 manipulators will have more than 10 degrees of freedom, with capacity-to-weight ratios 10 times the current state of the art. They will be fully modular and mission-configurable.
Knowledge-based systems. By 2020 knowledge-based systems will begin to approach a general problem-solving capability, although they will still be restricted to one class of problems. They will be able to handle broader, more general knowledge representations; model time more richly; perform inferences under real-time constraints; and perform problem solving that is distributed across multiple processes, processors, and physical sites.
Natural language and speech. Although natural language processing will be able to understand report-length texts or long messages in well-defined domains, reliability will remain the essential issue. Natural language technology is most likely to be applied where error detection (by human review of the machine product) is possible. It may find use in "pretranslation" systems or in watchdog systems that scan large volumes of material for items to be brought to an operator's attention.
ELECTRONICS AND SENSORS
The Technology Group on Electronics and Sensors assessed the following areas of technology:
Electronic devices include advances in monolithic microwave integrated circuits, superconductive electronics, vacuum micro devices, computer memories, application-specific integrated circuits, analog-to-digital converters, digital signal processing microcomputer chips, and wafer-scale technology.
Data processors include electronic subsystems that act as signal processors and target recognizers. The emerging technologies here are multiprocessor computing and neural networks.
Communication systems include communications satellites and
other platforms for communications nodes. They also include technologies to provide communications security and robustness.
Sensor systems include UAV detection radar for surveillance of moving ground targets, airborne detection and recognition radars for stationary targets, acoustic array sensors, magnetic sensors, air defense radars, and space-based surveillance and target recognition radars.
The findings reported in the Electronics and Sensors TFA include general findings, those specific to the technology areas specified above, and summary findings on three high-impact electronic technologies.
Land warfare is likely to evolve, as air and naval combat have already, toward long-range weapons, increased depth of combat, and increased reliance on stealth, electronic countermeasures, and mobility. The key to this evolution is the development of electronic sensing and target recognition systems that can operate beyond the visual horizon.
Silicon will remain the bulk semiconductor material of choice for most applications for the foreseeable future, if only because of the high industry concentration of development resources and manufacturing capability committed to it. Although some significant evolutionary improvements in silicon semiconductors will continue, the next decade will begin to see the performance of silicon devices limited by intrinsic properties of the material.
In special areas, greater performance gains will occur because other materials are far superior to silicon. For example, gallium arsenide (GaAs) is becoming the material of choice for high-frequency transistors. Silicon carbide or diamond may emerge as the semiconductor materials of choice for devices operating at high temperature and high power.
In the area of thin-layer semiconductors, the well-established technology for chemical vapor deposition of thin silicon layers on silicon substrates is being augmented by new techniques for growing single-crystal layers from the vapor phase. These new techniques allow deposition of highly uniform layers of solid solutions (such as
GaA1As) as well as elements and binary compounds. It is now possible to fabricate extremely complex multilayer structures whose properties differ dramatically from any bulk material. The long-term impact of these techniques will be significant improvement in device performance.
The recent discovery of materials that are superconducting above 77 K has renewed interest in superconducting thin films for high-frequency analog and digital circuits. Films with sharp superconducting transitions, high critical-current densities, and low microwave losses have been obtained in the laboratory. Superconducting thin films enable major performance advances in a variety of microwave, radio frequency, and digital logic applications.
Monolithic microwave integrated circuits (MMICs) have been made possible by the developments described above in high-quality semiconductor materials, new thin-film deposition techniques, and improved lithography. MMICs provide small-signal amplifier and power amplifier components for applications in the range from 10 to 100 GHz (Figure 3-6). They will enable phased-array radars, signal intercept systems, and communications terminals to be built with far
smaller size, weight, and cost compared with conventional hybrid technology. This new technology will be crucial in developing the following systems: UAVs with multi-use apertures for both ground surveillance radar and electronic intelligence (ELINT) receivers; space-based imaging radars; covert, beyond-line-of-sight communications "manpack" terminals; and extremely-high-frequency (EHF) terminals for air-to-air and air-to-satellite links.
Micron-size vacuum transistors have become possible with the development of reliable cold cathodes with high current densities. Vacuum transistors, which operate on the same principles as traditional vacuum tubes, will have high-frequency and high-power capabilities beyond those of semiconductor transistors. They can be used to develop radar and communications systems at frequencies and power levels not attainable with current solid state technology. They also could replace the traveling wave tubes now used, with substantial reductions in the size, weight, and power consumption of the system. A principal advantage of these devices is their robustness to damage from electromagnetic pulse (EMP) associated with nuclear blasts or video-pulse directed energy beams.
The advances described above in semiconductor materials, thin film techniques, and circuit miniaturization will all contribute to the emergence of advanced electronic devices, which are still in the concept stage. For example, these evolutionary advances in microelectronics will aid in developing the optoelectronic circuitry needed for revolutionary advances in optical computing and neural networks (see Optics, Photonics, and Directed Energy). Similarly, they will be needed in the field of bioelectronics, including biosensor coupling and, ultimately, biocomputing systems (see Biotechnology and Biochemistry).
Computer memory chips will continue to increase in capacity; in silicon technology, both direct random access memories (DRAMs) and static random access memories (SRAMs) are projected to gain nearly two orders of magnitude in bits per chip during the period 1989–2000 (Figure 3-7). Cost will also continue to decrease. In addition, high-speed memories (2 nanoseconds compared with 10 nanoseconds for CMOS SRAMs) based on GaAs or silicon carbide, which are now in development, should show a corresponding rise in capacity.
An application-specific integrated circuit (ASIC) is a single-chip substitute for a subsystem previously assembled from a number of simpler, standard chips. The use of ASICs in place of subsystem assemblies increases reliability while reducing the number of components and lowering the production cost, weight, and power required. Generally, however, the system development cost increases. The least costly type of ASIC, the programmable logic device, is already well
established. It can be programmed in the laboratory in a few hours. Gate-array or standard-cell ASICs, which must be programmed at the mask level during circuit manufacture, require 6 weeks to 6 months to produce and cost about $100,000. In the next 10 years the capability limit for gate-array ASICs will grow from about 10,000 gates to around 500,000 gates at a 50-MHz clock speed.
At the upper end in both cost and capability is WST (wafer-scale technology), which can implement an entire system on a single substrate. The cost of a WST chip is about $1 million, and production time is about 1 year. WST offers the advantage of eliminating many of the separate fabrication steps required to implement a digital system. The potential of WST can be indicated by designs achieved or in development now (Figure 3-8). A fast Fourier transform unit on a wafer 7.5 cm in diameter, demonstrated in 1986, had a throughput of 300 million operations per second (MOPS). A WST design on a 12.5-cm wafer, under development in 1990, is expected to achieve 2 billion
operations per second. By the mid-1990s, increased wafer size, reduced feature size, and faster clock rate will together increase WST computation throughput to 50 to 100 billion operations per second. For potential military applications, more relevant measures are the computations per unit size, weight, or power; WST seems capable of achieving 100 MOPS/cm3, 200 MOPS/g, or 3,000 MOPS/W.
Analog-to-digital converters (ADCs) provide the connection between the analog world of sensors and the digital world of data processing. As sensor technology expands, either in bandwidth (as in radars) or focal plane size (for optoelectronic systems), the requirements also expand for wideband ADCs with high dynamic range. ADC technology is being rapidly advanced in frequency and precision capabilities by commercial sector interest, particularly for high-definition television. However, military applications also require high dynamic range to accommodate their wide range of signal levels. Research is under way, with the support of the Strategic Defense Initiative Organization (SDIO), to develop monolithic (single-substrate) ADCs suitable for military use, including radiation hardening.
A digital signal processing microprocessor (DSP) is an integrated circuit similar to the microprocessor in a high-performance microcomputer but designed to be optimal for signal processing applications such as filtering, spectral analysis, and convolution. A new development in this area is the DSP microcomputer, which can sustain an average computation rate close to the peak rate required by a typical DSP task. These single-chip devices can be integrated into compact systems.
The Technology Group expects the DSP microcomputer to attain faster clock rates and smaller feature size over the next five years (Figure 3-9). Designs to exploit parallelism, such as multiple processors on a single substrate, will appear. In ten years GaAs technology will enable 100-MIPS DSP microcomputers. Fiber optics will provide gigabit interprocessor communications. Arrays with up to 1,000 parallel processors will become available, although software design methodologies to support this level of parallelism will also need to be developed. Among the implications of DSP microcomputer chips
for ground warfare will be increased sophistication, while reducing the size, of systems that rely on signal processing. Examples include compact smart weapons with onboard target recognition, communications systems with advanced low-probability-of-intercept technology, speech recognition in command and control systems, and radar and sonar systems with sufficient sophistication to detect stealthy aircraft and quiet submarines.
Major advances are occurring in electronic design automation. The key requirement is to have the data output from one computerized design or fabrication step in the multistep process be directly interpretable by the next step in the process. Two data format standards for this purpose are the VLSIC (very-large-scale integrated circuit) Hardware Description Language (VHDL), developed under DOD auspices, and the Electronic Design Interchange Format (EDIF), which is widely used in the commercial electronics industry. Software programs that synthesize data paths and entire circuits are available, although they provide quick turnaround at some cost in performance and silicon ''real estate'' efficiency. Other important tools are logic and circuit simulators. Much work is being done on developing integrated sets of tools for electronic design automation. (This need for integrated system design environments is also addressed in the Computer Science, Artificial Intelligence, and Robotics TFA; see section above.)
Increases in computing power in the near term will result from a new generation of VLSICs, multiprocessor computer architectures (parallel-processing supercomputers), increased use of GaAs and other high-speed semiconductor materials, and the technology for reduced instruction set computing (RISC). These changes will increase both numerical computing power (i.e., millions of floating point operations per second) and symbolic computing power (Figures 3-10 and 3-11).
A neural network is a computing architecture that performs highly parallel processing with a large number of simple processing elements (called the neurons). The neurons may be sparsely or densely interconnected. The potential advantages of neural networks include high-speed processing through parallelism, robustness to individual element failures, and compact hardware implementation of entire networks as VLSI chips.
Current realizations of neural nets are almost entirely in the form of simulations on a standard digital computer. Hardware realizations are still experimental, although rapidly maturing to the point
where neural net chips will be included in commercial information-processing systems. With sufficient commitment to their hardware implementation, neural nets could bring revolutionary changes to military systems. Their advantages of high-speed processing by rugged, compact hardware with little dependence on software would be significant for such applications as brilliant weapons, autonomous systems (UAVs and UGVs), automatic processing of sensor data, image processing, and adaptive signal processing and control. (Neural network realizations using photonic and optoelectronic hardware were discussed by the Optics, Photonics, and Directed Energy Technology Group; see abstract below.)
The Military Satellite Communications (MILSATCOM) architecture for the period after the year 2000 calls for communications satellites with the current ultra-high frequency (UHF) and super-high frequency (SHF) services, plus a robust/survivable segment and a complementary capability for augmentation and restoration.
The robust/survivable segment of this future MILSATCOM will operate in the extremely high frequency (EHF) range of 20 to 44 GHz. To support antijam, antiscintillation, and covert communications, it will use wideband, spread-spectrum techniques and autonomously adaptive uplink antennas. Low, medium, and high data rates must be supported, and agile uplink/downlink beams will be used to serve widely separated users concurrently. Satellite cross-links will provide worldwide connectivity without ground relays. The Technology Group identified the key electronics technologies required to accomplish these operational goals in a payload with substantially lower weight and power requirements than current technology. The new technologies include MMICs, VLSI processors, direct digital frequency synthesizers, and WST ASICs.
The augmentation/restoration satellites for the future MILSATCOM will be used to increase or replace critical coverage in a timely manner. Using the new technologies mentioned above, they can have many of the operational features of the robust/survivable segment mentioned above, except for supporting high data transmission rates.
The continuing development of smaller, more capable processors will benefit all radar systems but will be particularly significant for UAV-based radar systems. For example, UAV radar systems for the
detection of moving ground targets can be valuable adjuncts to the large systems, such as JSTARS (Joint Systems Target Acquisition Radar System), carried by manned aircraft. A current example is the DARPA-sponsored AMBER UAV, whose Ku-band radar includes a programmable processor to interpret raw radar data into moving-target reports. Moving targets the size of tanks or larger can be tracked out to a range of 15 km from the radar.
A UAV-based synthetic aperture radar could provide multiaspect information on stationary targets, sufficient to permit target detection and classification. Algorithms for automatic target cueing and recognition (ATC/ATR) are projected to improve considerably. The false-alarm density at a 50 percent detection probability may decrease by one or two orders of magnitude from the current performance of one per 10 km2. High-performance ATC/ATR algorithms, running on high-speed computers, can provide real-time detection of targets at surveillance rates (measured in square kilometers per second) that would overwhelm a human's imaging and decision capabilities. The underlying technologies for these advances will include neural nets, statistical pattern recognition, and model-based vision.
UAV-based radars could also be used for low-altitude air defense, overcoming the difficulties with terrain and foliage masking that hamper ground-based air defense systems.
Geographically dispersed networks of acoustic sensor arrays can be used to detect, locate, and recognize aircraft, weapons that are firing, and ground vehicles. While single arrays can provide directional cueing, networks of arrays can locate weapons and track aircraft. Networked acoustic arrays would provide passive battlefield surveillance for a variety of targets and cueing for active sensor systems. The electronics needed for these networks include noise suppression for the arrays, small data processors deployed with each array, low-data-rate communications from all arrays in the network, and processors to apply interpretation algorithms to the data collected from the network. The Technology Group forecasts that major capabilities of this type might be achieved within 5 years (Table 3-1).
Superconducting quantum interference device (SQUID) magneto-meters are sensitive enough to detect the magnetic field perturbation generated by a moving tank at near range (Figure 3-12). However, ambient temporal variations in the earth's magnetic field are four to five orders of magnitude larger, so background noise is likely to obscure the tank's signature. Whether advanced signal processing could distinguish tank signatures is a question requiring further field measurements and research.
Ground-based radar systems constitute the principal surveillance-
TABLE 3-1 Current and Projected Capabilities of Acoustic Array Sensor Networks
Current Capability (experimental 5-m array)
Projected Capability (smaller 2-m array)
Detection range of 5 to 20 km (10 km avg.)
Low wind and quiet background noise
High winds and high background noise (battlefield conditions)
Direction finding of 2° to 3° accuracy and 15° resolution
3 loudest targets under same "quiet" conditions
3 to 5 loudest airborne targets and several loudest weapons under battlefield conditions
Target location within 50 to 1,000 m
Depending on network geometry and source motion
Depending on network geometry and source motion
Multitarget location weapon
1 airborne target per array; unknown for transients (weapons) and ground vehicles
1 to 3 airborne targets per array; several weapon firings per second per array
Single helicopters in quiet background
Helicopters in multi-target, noisy environment; recognition and aid for other aircraft
and-tracking sensors for surface-to-air missile systems. Current systems, including HAWK and Patriot, are severely strained by newer and potential threats: tactical ballistic missiles, low-observable aircraft, cruise missiles, and modern electronic countermeasures. Advanced technologies for improving these systems are in development but were not detailed by the Technology Group.
The Technology Group summarized the performance parameters and technology requirements for space-based radar platforms for surveillance and target recognition. This application would require light-weight phased array radar antennas. The Technology Group assessed the advantages of corporate-fed phased array antennas for this purpose and for other applications, such as long-range, low-observable airborne radar. The transmit/receive functions of array antennas are an excellent application of the MMIC technology described above.
Summary Findings: High-Impact Electronic Technologies
Terahertz electronic devices are those that can operate at frequencies up to 1012 Hz. They will be needed to amplify and process analog signals with frequencies extending to this limit; they also will pro-
vide the basis for digital logic that can be switched within time intervals on the order of a picosecond (10-12 s). Terahertz devices will be used as the fundamental components in advanced radar, communications, electronic intercept, and weapon guidance/seeking systems. The electronics technologies that are potential candidates for terahertz performance include devices based on compound semiconductors (e.g., GaAs and InP), superconductive devices, vacuum microdevices, and optoelectronic devices.
Teraflop computers are high-speed computers capable of performing 1012 floating point operations per second. Unless logic devices with switching speeds of less than 10-14 s (two orders of magnitude faster than the terahertz devices discussed above) can be implemented, which
seems unlikely, a teraflop computer will require a large number of slower processors operating in parallel. For example, a single processor based on terahertz devices may be able to achieve a computing power of 1010 operations per second. So a teraflop computer based on terahertz devices would require about 100 such processors operating in parallel.
High-resolution imaging radar sensors will use the terahertz devices and teraflop processors described above for radar sensor suites capable of finding and recognizing targets at long range. Targets could be either stationary or moving; fixed targets will be the more difficult detection-and-recognition task, which will probably require the teraflop processing capability. The ability to locate and identify surface targets with precision at long range, combined with advanced communications, navigation, and command systems, will enable such targets to be attacked successfully from distances well beyond the range of enemy weapons.
OPTICS, PHOTONICS, AND DIRECTED ENERGY
This Technology Group assessed the following areas:
Optical sensor and display technologies receive optical radiation and interpret it for imaging displays to the user. The technological advances in this area include laser radar; multidomain sensors; sensor fusion; infrared search, track, and identification; focal planes with massively parallel processing; and helmet-mounted or similar heads-up displays.
Photonics and optoelectronic technology. Photonics comprises the science and technology to use photons to transmit, store, or process information. Optoelectronic technology couples electronic data-processing elements with optical elements. This area includes fiber optics, diode laser arrays, optoelectronic integrated circuits, optical neural networks, acousto-optical signal processing, and various other technologies that process optically transmitted information.
Directed energy devices are intended to generate highly concentrated radiation as a means of directing a high level of energy—which may be very short in duration—on a small target area. The radiation may be from the optical portion of the electromagnetic spectrum (as in lasers), from the radio frequency portion (microwaves), or from accelerated charged particles. Directed energy technology was considered by the Power and Propulsion Technology Group as well as the Optics, Photonics, and Directed Energy Technology Group.
Optical sensor technology and photonics provide basic building blocks for advanced integrated sensors and high-speed processors. Directed energy devices provide the long-range, speed-of-light capability to degrade or destroy hostile smart systems.
Essential for advanced system design in these technology areas is a computer-aided design environment that allows the integration of detailed information on sensors, processors, and the basic properties of their components. In addition to the initial design and development effort, such an environment could be used interactively to respond to evolving threats.
Optical Sensor and Display Technologies
Laser radar provides high-resolution target imaging, target discrimination, and detection of low-observable targets. The current technology includes systems based on carbon dioxide lasers, solid state lasers that use diode-pumped neodymium, and titanium sapphire lasers.
Solid state laser technology is being extended to an average power of several hundred watts, which will enable laser radars to have very long ranges (depending on the wavelength and atmospheric attenuation). Further development of both carbon dioxide and solid state laser technology should provide the peak and average power needed for various laser radar applications, while the size of the system will decrease significantly.
Multidomain smart sensors will combine a laser radar with one or more other sensor systems. A laser radar working with a wide-area surveillance sensor, such as microwave radar or a passive infrared search-and-track (IRST) sensor, enhances target detection and identification while reducing false alarms from clutter. Because the system can be configured so that the sensor components share the same physical optics, information across domains can be fused at the pixel level. This can provide a multidimensional information space for subsequent sensor fusion processes. The richness of this information can allow a human observer to detect targets in motion and stationary targets—even those concealed by camouflage or ground cover. One such system, for use on tanks and air defense weapons, combines a rangefinder and front-looking infrared laser radar.
Passive IRST systems can be used with laser radars for wide-area searches capable of detecting low-flying aircraft against terrain and
clouds. Passive IRST has the added advantage of being covert; it sends out no signal beam that can be targeted by enemy counter-measures. By operating in two bands simultaneously, passive IRST can make stealthy or camouflaged air vehicles more detectable.
A multidomain sensor system with laser radar and sensor fusion at either the pixel level or the image level will be part of an automatic target recognition system for detecting and classifying aircraft and missile threats on the tactical battlefield. Airborne systems currently under test for tactical target detection and identification use carbon dioxide and GaAs laser radars in combination with passive sensors in the visible and 8-to 12-µm region, plus 85-GHz millimeter-wave radar.
Another application of laser technology for multidomain smart sensors is differential absorption LIDAR, or DIAL. (LIDAR stands for light detection and ranging.) DIAL can be used to detect specific chemicals in atmospheric emissions by their absorption of light from one laser beam of a dual-beam system. Current systems are being developed to detect volatile solvents used in clandestine chemical-processing operations (drug processing in particular). The technology is extendable to the detection of other military targets, including CTBW production facilities, vehicles hidden in trees with their engines idling, fuel dumps, and perhaps ammunition dumps.
Further experience is needed in combining laser radar with passive IRST in a package suitable for Army applications. If a first-generation system can be field-evaluated within the next 5 years, full production of the system should be possible within 15 years. A passive IRST wide-area search system combined with laser radar for ranging and identification would have a major effect on low-altitude surveillance and defense against air targets. The variety of such targets would include conventional and stealthy manned aircraft, cruise missiles, UAVs, and tactical missiles.
Only an integrated approach to sensor fusion can satisfy the demands of the future battlefield for rapid integration and interrogation of signals from a multisensor suite (Figure 3-13). Successful response to incoming missiles, aircraft, smart weapons, and satellites will require completely autonomous target detection, recognition, and acquisition. The time requirement may not permit a man in the loop, so the system must provide 100 percent target validation. These time and reliability requirements necessitate the use of multidomain sensors and automatic processing of their images.
With respect to sensor fusion technology, a new concept is the integrated sensor. A high-capacity, optical-domain parallel processor—probably of a neural net design—would be directly interfaced to the high-
density focal plane of a multidomain sensor optics package. The output from the processor would feed directly to a fire control system.
Focal plane arrays are currently available for the infrared region, but integration into a monolithic, switchable structure has not yet been achieved. Optical parallel processing is under development. Signal processing from multiple acoustic receivers, whose output might go to the fusion unit through a secure fiber-optic channel, is another development requirement.
Smart focal planes are another concept for the rapid processing of data from sensor optics (Figure 3-14). An array of small-area detectors will share space on the focal plane with processing circuitry. An array of microlenses will direct the incoming radiation to the detector array. The focal plane image will he read out in a massively parallel manner to a sequence of optoelectronic processing planes beneath the focal plane. This parallel-processed readout from the focal plane will avoid the current bottleneck of serial readout and serial processing in conventional serial computers.
As conceived, the smart focal plane technology would allow image acquisition and processing rates greater than 5,000 frames per second. The output information will be highly processed already, so communication bandwidths can be reduced. The potential size reduction from current serial processing technology could yield advanced capabilities in a package small enough for use in smart missiles. However, much of the technology to implement the smart
focal plane concept remains to be developed. Suitable algorithms and processor architectures for optical processing of images are still in the research stage.
For infrared scanners, focal plane arrays of detectors based on Schottky-barrier materials can improve the photon collection efficiency of the entire sensor by five orders of magnitude, compared with conventional infrared scanners. Although Schottky-barrier materials have a lower quantum efficiency than other solid state detectors, a focal plane array in a staring format compensates for that disadvantage by using a large number of detectors. Arrays of 10,000 × 10,000 detectors should be available in the next two decades.
Schottky-barrier technology covers all optical wavelengths of interest to the tactical battlefield—visible, near-infrared, and mid-infrared. There are many options for combining them in multispectral arrays or tuning for a particular region. Arrays for the near-infrared region can use skyglow and thermal emission in the 1-to 2-µm range for night vision.
Smart sensors will also be applicable to the soldier's personal gear, as in the smart helmet concept (Figure 3-15). Future battlefields will require enhanced awareness by the field soldier, in response to increased use of camouflage and stealth techniques. Eye protection will be needed against antipersonnel lasers. The smart helmet incorporates advanced night vision sensors, sensor processing, and commu-
nications with eye protection, because the information seen by the eye will always be an indirect display.
Notable advances in display technology will include the integration of monolithic drive circuitry with lightweight, low-power display arrays. Displays will range from personal ''eyepiece'' viewers and helmet-mounted heads-up displays to large, multiviewer display screens. The direction of development in advanced displays will be toward allowing a human operator to have true telepresence in environments that are too dangerous or are physically inaccessible. Lightweight displays with a wide field of view will have major military applications.
Photonics and Optoelectronic Technologies
Photonic approaches to communications, such as fiber-optic cables, offer several advantages over electronic systems. They are relatively immune to electromagnetic interference and provide very large bandwidths (in the terahertz range). Computing applications of photonic systems can have higher clock rates and large-scale parallel processing.
For military communications network applications, a fiber-optic network will be supplemented by radio frequency links to mobile nodes: sensors, satellite relays, and users. The fiber-optic network will be more resistant to jamming, interference, and interception than the more vulnerable radio links. Military applications of fiber optics currently provide rapidly deployable links over distances from tens of meters to kilometers. Time-division multiplexing is used, which limits the bandwidth and compromises the robustness and flexibility of the network. Future military applications will combine wavelength-division multiplexing with time-division techniques, providing a combined peak network capacity in the range of 10 Gbit/s or more and servicing hundreds of users.
Fiber optics may also allow close integration of widebandwidth sensors with ground operations. In addition to advanced C3I sensors, this capability will also enable telepresence by passing high volumes of sensory data between the remote platform and the ground operator, via a connecting optical fiber.2 Because the high-powered signal
processing and computing capabilities needed for data reduction, interpretation, and display can be located at the ground controller's location, the remote platform can be much smaller, less expensive, and therefore more expendable. Among the possible applications are (1) advanced fiber-optic-guided missiles, (2) airborne surveillance platforms with multidomain smart sensors but minimal onboard signal processing, and (3) tele-operated ground vehicles for both reconnaissance and weapon delivery.
Guided optical-wave sensors are an area of fiber-optic technology in which changes in the amplitude or phase of optical waves in the fiber are used to sense vibration (acoustic or seismic sensors), temperature or pressure changes, rotation (gyroscopic sensors), and even electrical or magnetic fields. A notable current effort in this area is the fiber-optic gyroscope. Although it is less sensitive than mechanical or laser gyroscopes, it offers compactness, robustness, and low cost—qualities that suit it to a number of missile applications. This technology is just emerging.
Since the mid-1980s, important advances have occurred in solid state laser technology, largely through research supported by the Department of Energy, DARPA, SDIO, and the service laboratories. Their all-solid-state design provides the advantages of high reliability and low maintenance. Mass production techniques for solid state materials and for laser array pumps promise low cost as well. At present, designs with longitudinal pumping give the highest efficiencies, but transverse pumping of solid state laser slabs by two-dimensional diode laser arrays is better suited for higher power levels, albeit at modest efficiencies.
For example, the projected capability of a neodymium laser demonstrator, due by 1993, is 300-W average power, about 10 percent efficiency, and a lifetime of greater than 109 shots. Various wavelength conversion techniques will enable this demonstrator or similar devices to be wavelength-selectable from the visible to the mid-infrared at greater than 100 W. This average power level exceeds the minimum required for many space-based and tactical applications.
This solid state laser technology will provide eye-safe laser rangefinders and target designators that are more reliable than those now fielded. in addition, however, it will lead to new applications, such as the laser radars described earlier, active optical countermeasures (antisensor lasers), and high-bandwidth laser communication from satellites to theater and battlefield commanders.
A related area with recent advances is diode laser and laser array technology. While individual diode lasers are limited in power, coherent arrays will yield 10 W or more of output at greater than 40
percent efficiency. Scaling to over 100 W average power may be feasible, with power densities exceeding 100 W/cm2. Varying the semiconductor material will enable these arrays to operate in the visible region, at eye-safe wavelengths above 1.4 µm, and even in the 2-to 5-µm range that is used for laser-activated proximity fusing.
Optoelectronic integrated circuits combine electronic and optical microcomponents on a single semiconductor chip (Figure 3-16). The purpose of the chip may be to provide an information interface between the two technologies or to create a functional hybrid device. Commercial applications are driving the rapid development of this just-emerging integration of the two technologies. Within 15 years, optoelectronics will be mature for communications and computer interconnect applications. In 30 years, it will have a wide range of applications built on hybrid functionality, such as massively parallel optical processing and wavelength multiplexing.
The best-characterized materials for optoelectronics are GaAs and other semiconductors formed by combining Group III and Group V elements (III–V semiconductors). Ferro-electric liquid crystals are another possibility, particularly for light-modulating applications. Lithium niobate is currently the material of choice for volume holographic storage and interconnects.
A key point is that all the semiconductor optoelectronic technologies, even of the III–V semiconductors, are still very immature compared with silicon technology. A substantial and sustained investment will be needed for this technology to mature. In the near term, optoelectronic-processing applications will mostly use arrays in which the logic function is performed by electronic components, while the optical components provide the mechanism for highly parallel interconnections. In terms of combined speed, low power, and high spatial density, optoelectronic arrays based on III–V semiconductors will be difficult to surpass in the long term.
Neural networks constitute another information-processing technology in which photonics will play an increasing role. By analogy with biological neural systems, a neural network contains two types of processing elements: synapses and neurons. A synapse performs an operation on its single input; a neuron receives inputs from multiple synapses and combines them in some nonlinear way to produce an output. Although photonic or optoelectronic implementations of these elements are at present less developed than electronic alternatives, they offer the potential for far larger numbers of synapses and neurons per component. In addition, photonic elements can support more flexible connectivity patterns, including some that appear essential for neural net architectures to perform vision and image-
processing tasks. In the long term, the Technology Group forecasts that optical neural networks will be used for real-time automatic target recognition based on multidomain sensor inputs, for speech understanding (i.e., word and pattern recognition), and for complex signal processing.
The TFA includes a special chapter on existing R&D projects on optical neural networks.
Acousto-optics uses the crystal vibrational modes of a Bragg cell to encode or decode information carried in the modulation of light beams. The potential information-carrying capacity of modulated light and optical processors can be illustrated by an analog acousto-optical device called a time-integrating acousto-optical correlator. Current versions of this device can process the equivalent of 1013 operations per second, which is several orders of magnitude more than existing electronic devices and a factor of 10 above the goal of "terahertz" devices. In addition to this potential for higher information through-put, optical processor architectures will require less power and will be smaller and weigh less than digital electronic processors.
Acousto-optics is also applicable to advanced sensors. The characteristics that can be "read" from a signal intercepted from an emitting platform can be used to identify the specific signal type and determine the location or velocity of the platform. There is no limitation on the wavelength of the electromagnetic radiation that can be processed in this way.
Optical techniques with lasers can be used to control information carried in the amplitude, frequency, and phase modulations of micro-
wave radiation. In addition, optical fibers make excellent waveguides for distributing information-carrying microwaves; the available bandwidth can be on the order of hundreds of gigahertz. This technology appears promising for control of phased arrays, control of remote antennas, microwave communications requiring extremely high data rates, and secure communications. The Technology Group forecasts that within 15 years microwave and optical circuitry will be integrated on a single chip. In 30 years, sophisticated optical computing will be used for various adaptive antenna functions, such as beam shaping, null steering, and side-lobe suppression. Many microwave frequencies will be multiplexed over one fiber-optic network.
In adaptive optics, a wavefront sensor is used to measure aberrations in an incoming light signal. This information controls a deformable optical element that adjusts to compensate for the optical aberrations, which would otherwise limit the performance of the optoelectronic system. The advanced techniques for adaptive optics use nonlinear optical materials that perform both the sensing and the compensation functions.
Aberrations caused by optical system imperfections or the atmosphere result in substantial signal degradation or loss of laser coherence in nearly all present optical systems. Adaptive optics will become an essential part of future systems; they will substantially increase the operating range and improve the resolution of laser systems. For military applications, adaptive optics can improve performance of many optoelectronic systems, including antisensor lasers, passive battlefield imaging, active or passive space object imaging, auto-tracking, and optical jammers. Adaptive optics can also improve the projection of laser power from directed energy weapons by correcting for atmospheric aberrations between the beam source and relay mirrors.
Sophisticated countermeasures to laser antisensor threats can make use of applied nonlinear optics. This technology will also be important in implementing components required for all-optical processing and computing. The Technology Group forecasts its use for passive laser protection within 15 years and for advanced optical processing components in 30 years.
Binary optics is a technology for creating diffractive optical devices on a substrate by use of lithography and micromachining. It gives the optical circuit designer the capability to create novel elements as well as alternatives to more conventional refractive elements. Binary optics methodology builds on VLSI circuit technology; both are well suited to computerized circuit design and manufacturing. Low-cost mass production of binary optic designs is possible through replication, embossing, or molding of subassemblies. In addition, the potential
of diffraction devices to compensate for aberrations pushes the range of optical design further into the deep infrared and ultraviolet regions.
The transfer of binary optics technology to U.S. industry began only in 1988. Already, more than 30 optics and aerospace companies have acquired the knowledge and capability to produce binary optics. One important near-term Army application for binary optics is to correct chromatic aberration in infrared imaging systems. Binary optics also has the potential to simplify the production of optical systems for military applications. It should also make those systems cheaper, lighter, and lower in power consumption.
Directed Energy Devices
In a technical sense, even laser devices that are used for information functions (laser radar, rangefinders, target detectors) can be considered directed energy devices. However, as used here the term applies to energy beam technology primarily concerned with delivering a high-energy flux on a target.3
The Technology Group described the technologies needed for a conceptual ground-based laser antisatellite system. The directed energy weapon in this system would either be a free-electron or chemical laser, complemented with adaptive optics.
The free-electron laser (FEL) uses a high-energy accelerator to create an intense stream of electrons. The stream traverses a series of alternating magnetic fields, which causes them to emit coherent electromagnetic radiation at a wavelength tunable by the electron's energy and the magnetic field strength. The entire beam-generating process occurs in a high vacuum, which limits self-distortion found in crystal or gas lasers. The distinctive advantages of this high-energy laser include efficient production (greater than 25 percent efficiency) of high average-power output; broad, continuous tunability over a wide frequency range (in theory, at least, from long-wavelength microwave to short-wavelength x ray); excellent beam quality; and generation by electrical power, which simplifies logistical support.
The Army is pursuing two FEL approaches under SDIO funding. Both have demonstrated sufficient electron beam brightness for operation at high power and have been operated at a variety of wavelengths (although not at x-ray wavelengths). Peak output power in excess of 10
MW has been demonstrated, with average power for short durations in the range of tens of kilowatts, but excellent beam quality at high average power is yet to be demonstrated. Major testing is under way.
By 2020, ground-based—and possibly space-based—FEL systems might conceivably be able to intercept and destroy missiles during their boost phase. (A ground-based system would use space-based relay mirrors to reflect the beam onto targets.) To achieve this goal, difficult technical breakthroughs in beam generation and steering are required. For tactical applications, a high-power microwave beam, using an FEL source, should be available with multimegawatt power.
High-voltage, short-pulse electron beam accelerators can be used to drive conventional microwave sources (magnetrons, klystrons, backward wave oscillators) to create an intense, narrow-band, pulsed radio frequency energy beam. Peak power levels can be as high as several gigawatts, with energies per pulse greater than 200 J. A newer technology uses a solid state switch to produce a high-power wideband radio frequency beam (also called a video pulse), which can operate as a repetitive pulse (repeating with a frequency of 10 to 100 Hz).
Continued progress with these high-power, high-energy radio frequency sources will enable a new class of weapons, in which mission kill is accomplished by burning out electronic components or detonating electro-explosive devices in the target. Potential targets include smart munitions, antiradiation missiles, mines, aircraft, radar and infrared guided missiles, communications nodes, and UAVs. The Technology Group also summarized the projected capabilities of systems now under development.
FEL, high-power microwave, and charged-particle beam weapons all depend on the development of a compact accelerator. The basic concept is to alter the linear transport geometry of the traditional linear induction accelerator into a spiral or circular configuration. This would allow the same accelerator module to act repeatedly on a circling swarm of charged particles, until they reach the desired velocity. Compact accelerator development is being pursued by the Naval Research Laboratory and in two projects currently supported by DARPA.
BIOTECHNOLOGY AND BIOCHEMISTRY
The Technology Group on Biotechnology and Biochemistry characterized "biotechnology" as the application of scientific principles for clinical and industrial uses of biological systems to produce
goods and services. Living organisms or their parts are used to make or modify products or to develop organisms for specific purposes. Contributing technologies include molecular and cellular manipulation; enzyme definition, design, and production; and microbial techniques for growth and fermentation. The technologies assessed by this Group were divided into the following six categories:
Gene technologies include the methods to "touch the genome and modify it. (The genome is the cellular site of genetic material, which carries the information for biological inheritance.) The techniques include gene replication, splicing, modification, regulation, transportation, and expression.
Biomolecular engineering is the technology to design and produce biomolecules (structural proteins, enzymes, etc.) with specific, tailorable properties.
Bioproduction technologies use living cells to manufacture products in usable quantities. These methods of biosynthesis can range from fermentation to solid state molecular synthesis, multistage bioreactors, and methods still evolving.
Targeted delivery systems are composites of materials that are designed to concentrate the active agent(s) in the composite at specific sites in the body where its activity is desired.
Biocoupling is the linkage of biomolecules or biomolecular complexes to electronic, photonic, or mechanical systems. For example, highly sensitive and selective detector molecules (biosensors) would be bound to microelectronic (or optical) circuitry to produce a system able to detect a single molecule of a CTBW agent.
Bionics aims at methods of directly connecting the human neural system to electronic or mechanical systems, such that the nonhuman system functions in ways similar to human limbs or organs.
The technologies in the scope of this TFA are based on a wide range of scientific disciplines (molecular biology and biochemistry, physical and organic chemistry, medicine, manufacturing process technologies, and electronics). To exploit the potential of biotechnology for Army-specific applications, the Army will need to assemble multidisciplinary research teams with competence in physics, chemistry, biology, medicine, and engineering, rather than segregate staff by discipline, as traditionally done.
Collectively, the technologies are the newest, least mature of the STAR technology areas. They are expanding rapidly in terms of discoveries, applications, and inventions. Also changing rapidly are perceptions of their importance. The possibilities are high for amplification of results from relatively small increments in investment. The Technology Group believes that a stable funding base for biotechnology is essential to provide the continuity of research and application development required for its military potential to be realized.
The Technology Group believes that success in achieving biotechnology goals depends on program management with a fundamental appreciation of advanced molecular biology, especially nucleic acid and protein chemistry, immunology, and infectious disease (especially vector biology). An understanding of process engineering is also key to success. Program management for Army biotechnology programs should reside in domains such as the U.S. Army Materiel Command, the Medical Research and Development Command, and the Chemical Research Development and Engineering Command.
In some areas of direct interest to Army applications, foreign biotechnology is ahead of U.S. academic and nonmedical private sector efforts. The Technology Group concluded that the Army is well poised to use work from important foreign laboratories, because the Army Materiel Command has nascent but important programs in many of them. Various administrative, contractual, and legal restrictions must, however, be overcome to realize this potential. A specific recommendation is to establish with our allies more joint military technical working groups. The management of such working groups, or other forms of joint research teaming, should include currently active researchers as well as senior technical managers. The Technology Group believes the future of this area can best be foreseen by those at the research bench.
The Technology Group foresees CTBW as a growing military threat. Potential adversaries will continue to find the use, or merely the threat, of CTBW as an inexpensive weapon of aggression and defense from retaliation by the world community. The Group believes deterrence can best be achieved by a system of countermeasures, which would collectively eliminate the efficacy of the CTBW threat to Army contingency operations. The discussion in Chapter 2 of CTBW countermeasures, including (1) detection and identification, (2) physical protection, (3) medical prophylaxis and therapy, and (4) decontamination, was drawn primarily from an appendix in this TFA on CTBW. The Technology group forecasts that
biotechnology will prove pivotal in all four of these countermeasure categories.
Today's successes in biotechnology cluster in the areas of medicine and pharmaceuticals, agriculture, and bioproduction of specialty "natural-product" chemicals such as sweeteners and solvents. The Technology Group foresees future capabilities extending to large-scale bioproduction from generic feedstocks, design and synthesis of novel biomaterials; coupling of biomolecules with electronic, optical, and mechanical devices; selective improvement and modification of life forms; and environmental decontamination (bioremediation).
For the kinds of advanced capabilities needed by the future Army, biotechnology offers important advantages when compared with traditionally engineered and manufactured systems:
Biological systems perform complex, repetitive syntheses with few side products and few errors, compared with traditional chemical production methods. They are therefore well suited to routine, reliable production of complex substances in pure form.
Bioproduction can give these complex substances extremely specific recognition capabilities, making them ideal for selective detection or site-specific activity.
Biochemical reactions typically occur under milder conditions than analogous industrial chemical processes, so bioproduction can be less expensive, require lower energy inputs, have less critical operating conditions, and require simpler apparatus.
The cost of routine production of biotechnology products should usually be less than for alternatives produced by traditional processes. In most cases of biomolecular-engineered products, there will be no comparable alternative producible by nonbiological processes.
Biosystems are "engineered" at a molecular level, so such systems (a white blood cell, a microorganism, an eyeball, or a brain) are very compact relative to a traditionally manufactured electrical, optical, or mechanical system with similar functionality. To the extent that this engineering uses "components" already developed by nature, a large part of the initial R&D cost has already been paid.
The most important "system" in the future Army will continue to be the human soldier. Because the soldier is a biological system, biotechnology offers unique potential for enhancing the performance of this most complex, critical, and costly of the Army's systems.
A major obstacle to achieving biotechnological advances is the lack of adequately skilled personnel. The Technology Group addressed this issue in detail in a special chapter.
Recombinant DNA techniques can now be used to transfer the characteristic of one or several specific genes to a different cell or organism. As knowledge of specific genes and the mechanisms by which they interact increases, the techniques of recombinant DNA, cell fusion, and gene splicing will permit the transfer of multigene, complex characteristics into cells and organisms.
Cell fusion, or hybridoma, technology involves the fusion of two cells, each with desired characteristics. For instance, a cell that produces a specific antibody can be fused with a cell easily grown in cell cultures. The hybrid cell retains the ability to produce the desired antibody hut can be easily cultured in quantity. A recently announced hybridoma technique allows the production of monoclonal antibodies in days, rather than the months formerly required. Similar ''quantum-leap'' advances will continue for at least the next three decades.
Gene technologies enable production of new substances, or even new organisms, with applications to medical and nonmedical interests of the Army, such as these:
substances for discrete recognition of an organism (including identification of individual persons) or a substance (DNA probes, receptors and antibodies for specific molecular conformations);
diagnostics for disease and CTBW threat detection;
new or altered materials, with improved structural, functional, or renewable characteristics, produced by genetically altered biological systems;
medicinal drugs and therapeutic agents;
vaccines and multivalent vaccine delivery systems;
physiologically active compounds that modify biological response;
artificial body fluids and prosthetic materials;
new foods and food production processes;
decontamination, detoxification, and bioremediation processes;
new or improved materials for adsorbing or neutralizing hazards and for purifying water, food, or production feedstocks.
Gene technologies will be used in all seven of the high-payoff opportunities described below.
Our current ability to relate the structure of biomolecules to their function is limited for all but the smallest of these molecules. There
are still many surprises and predictive failures, even in areas where predictive methods are most advanced. At present we lack the ability to design de novo a biomolecule for a reasonably complex function, such as radar nonreflectivity. However, the scientific disciplines to pursue such a capability do exist.
Progress in biomolecular engineering will depend on advances in two contributing areas: (1) prediction of the biomolecular structures required to achieve a desired function and (2) methods to design, construct, and produce molecules or composites that meet specific functional requirements. The multidisciplinary research teams needed for this work must combine expertise in structure-function physical chemistry; physical biochemistry; computational methods for simulation, modeling, and display of biomolecules; analytical methods for determining the detailed structure of biomolecules; biophysics and chemistry of molecular biopolymer synthesis; and the biochemistry and molecular genetics of the genome.
Of the seven high-payoff opportunities identified by the Technology Group (see below), biomolecular engineering will be applicable to five: deployable bioproduction of military supplies, biosensor systems, novel materials, extended human performance, and antimateriel products.
The bioproduction techniques and resources already available include bioreactors; cell culture and fermentation techniques; cell growth media and factors; established cell lines for mammalian, insect, bacterial, yeast, and algal cells; cell harvesting and processing techniques; chemical coupling techniques and processes for immobilizing (fixing) cells and proteins; and techniques for purification and isolation, such as affinity chromatography.
Further development of fermentation and cell culture techniques, cell lines, and bioreactors will be particularly important for efficient large-scale production. Bioproduction methods also need to be scaled up from laboratory size to industrial production scales.
Affinity chromatography is based on the covalent coupling of affinity ligands, enzymes, and other biomolecules with specific recognition characteristics to inert, solid support materials. The resulting technology will enable rapid, efficient purification and processing of ultrapure materials on a large scale. In one type of purification (mono-clonal antibodies), the older technology of column chromatography had a process yield of only 40 to 60 percent, gave a product that was 95 percent pure, and required 2 to 3 days. The new method based on
membrane affinity can process the same amount of material in 1 hour, giving a 90 to 96 percent yield and a product that is 99 percent pure.
Bioproduction technology will be applicable to five of the Technology Group's selected high-payoff opportunities: deployable bioproduction of military supplies, enhanced immunocompetence, novel materials, in-field medical diagnosis and treatment, and antimateriel products.
Targeted Delivery Systems
In a targeted delivery system, an active substance is encapsulated in a membrane or a matrix that permits controlled release when the capsule system reaches its intended site of action. The release may be slow, by diffusion out of the encapsulating material, or triggered by dissolution of the capsule. Thus, these systems permit the use of biosubstances that would otherwise be inactivated or degraded before they could be effective for their intended purpose.
New microencapsulation technology, using biomaterials that are biocompatible and biodegradable, will protect sensitive active substances from degradation or inactivation by light, chemical, or biological stresses. In medical applications, drugs, vaccines, peptides, and proteins will be administered with microencapsulation systems now under development. In nonmedical applications, field-deployable, stable capsule systems will be useful for intelligent biosensors, decontamination systems, and biocamouflage systems for signature suppression.
Among the potential applications of interest to the Army are drug and vaccine delivery systems for prophylaxis or treatment of infectious diseases or CTBW agents, energy-rich or performance-enhancing foods and supplements, decontamination methods, deployable purification kits, and regeneration or replacement of tissues and organs.
The Technology Group projects advances in this area that will produce self-regulating delivery systems. Specific triggering mechanisms for release of the active substance will be developed, such as triggers by pH, ionic strength, specific receptor/ligand binding, or specific frequencies of electromagnetic radiation.
Of the high-payoff opportunities for biotechnology, targeted delivery systems could play a role in enhanced immunocompetence, novel materials, in-field medical diagnosis and treatment, and antimateriel products.
For near-term biosensor applications and longer-term bioelectronics, it is necessary to develop techniques to couple the biocapture and recognition event (the response of a biomolecule to its target molecule or energy form) to the means for amplifying, transducing, and communicating that information into an electronic, optical, or mechanical signal. Development of antibody or bioreceptor molecules as biosensors is in progress. The coupling technology is less advanced, receives less attention, and will be more difficult.
Bioelectronics refers to the use of biomolecules or biosensor systems within an electronic data-processing system—for example, a "microchip" integrated circuit that incorporates biosensor elements into a computer memory "biochip." The development of this technology depends not only on biocoupling advances but also on biomolecular elements with a binary signal response.
As with biomolecular engineering, to reap the potential of biocoupling technology will require multidisciplinary teams competent in many specialties, including molecular genetics, receptor physiology, and pharmacology; physical chemistry of macromolecules; the physics and chemistry of signal trapping and recognition; engineering adaptation of unit-event signals into systems with integrated outputs; and engineering to adapt the environment required by the biosensor to the sampled environment.
The Technology Group recommends that biocoupling be pursued in parallel with biosensor development, because biocoupling methods may determine which biomolecular mechanisms are feasible as biosensors within the larger system to which they are coupled.
The projected applications for biocoupling include the following areas of interest to the Army:
deployable remote detection and analysis systems (with telemetry) to assess the presence and status of hostile troops and equipment, disease and CTBW threats, or environmental parameters;
rapid diagnosis and identification of disease and CTBW threats in the field;
terrain and perimeter monitoring;
monitoring of critical personnel performance;
performance modification (see also bionics biotechnology); and
A successful bionic device replicates both qualitatively and semiquantitatively the function of the living physiological system it imitates. The technology has been most successful in copying neural systems that couple an environmental signal to receptors that act as high-gain analog-to-digital converters and otherwise process the signal as information.
Passive bionics, in which properties of a biological material are emulated for a single objective, is a maturing technology already being exploited by the Army. Potential developments include fibers for personal armor, nonlinear optical polymers for eye protection from lasers, and artificial sense organs for robotics applications. Bioelastomers, antifoulants, biolubricants, and bioadhesives are under study in important current programs. The Technology Group anticipates some striking successes in this area but no surprises.
The alternative to single-property, passive bionics is multicomponent cybernetic systems that model the neurally modulated "smart" behavior of animals. Such systems could be possible in 30 years. The Technology Group expects that progress in this area will depend on the development of high-density neural nets with advanced artificial intelligence capabilities and on advances in biocoupling.
The Biotechnology and Biochemistry Technology Group identified seven areas of biotechnological application as having the highest payoff for meeting long-term Army needs. It then matched these application areas to the advanced systems concepts of the STAR system panels. On the basis of the resulting matrix, the Technology Group identified one or more key products that could be produced within the next 30 years in each of the seven areas.
For each high-payoff area, the Technology Group set up a road map, or R&D time line. These road maps show the timing of required Army investments in the enabling technologies, important milestones in the development process, and an approximate time frame for fielding the product.
The high-payoff areas and their road maps are summarized here. For full details, the reader should consult the Biotechnology and Biochemistry Technology Forecast Assessment in the appropriate volume of STAR reports.
Deployable bioproduction of military supplies will use bioproduction
methods to produce food, fuel, potable water, explosives, and perhaps ammunition components from indigenous feedstocks. Ultimately, the feedstocks could be as simple as air, water, a carbon source (such as biomass), and sunlight or another common energy source. These theater-based production units would significantly shorten the logistics tail of deployed forces.
The key product identified for this area was deployable bioproduction of fuels. The objective would be a portable unit capable of small-scale (e.g., a few gallons a day) production of an engine fuel. It would be used by small units isolated from regular logistics support, such as Special Forces operations. Near-term Army investments in gene technologies and biomolecular engineering could result by 2002 in laboratory-scale expression of a selected nonpetroleum-based, oxygen-rich fuel. Investment in bioproduction technology beginning in 2005 could result in deployable bioproduction by 2020.
Biosensor systems will include novel ways of coupling electronic or photonic components with biosensors (Figure 3-17). The key products selected for forecasting were a multithreat, deployable detector array, to be fielded after 2020, and integrated bioelectronics. Army investments would be required in the next decade in gene technologies, biomolecular engineering, and biocoupling.
Enhanced immunocompetence for personnel would manipulate the genome of the soldier's white blood cells to confer immunocompetence against diseases and CTBW threats. The response of white cells would be altered to provide not only specific recognition of antigens, as in current vaccination, but also responsiveness to classes of antigens and their potential variants. Troops would be "immunized" in this way just prior to deployment. Gene libraries for immunogens and immunocompetence enhancers would be developed for relevant diseases and known CTBW agents. The capability to immunize troops in this way by 2020 would require immediate Army investment in gene technologies and later (after 2000) investment in targeted delivery systems. Advances in biomolecular engineering would be relevant but would not require Army investment.
Novel materials will result from the use of biomolecular engineering to design new molecules rather than simply borrowing or adapting natural genes to produce naturally occurring molecules. The key products in this area were specialized lubricants, adhesives, and coatings; adaptive camouflage; and multithreat protective clothing against CTBW, electromagnetic radiation, and ballistic impact. The relevant technologies would include gene technologies, biomolecular engineering, bionics, and bioproduction technology. Army investment in biomolecular engineering by 1995 would allow for novel
molecules to be designed by 2003. Immediate Army investment in bionics would enable prototypes of systems and materials based on these novel substances by 2005. Investment in gene technologies after 2000 would enable efficient expression of the novel substances by 2010. The adaptive biocamouflage was projected for fielding around 2020; the other two products would be ready after that.
In-field diagnostic and therapeutic systems will reduce casualties due to disease and CTBW threats. Gene technologies, biomolecular engineering, bioproduction technology, targeted delivery systems, and biocoupling technology will all be required. The key products identified by the Technology Group were rapid, specific diagnosis of symptoms, to be fielded around 2020, and an even longer-term system that would include countermeasure selection and its bioproduction. Immediate Army investments would be needed in gene technologies and biomolecular engineering. Investment would also be needed, beginning in 2005, in targeted delivery systems.
Extended human performance refers to direct coupling of the human central nervous system to machines and other uses of bionics and orthopedics. The required Army investments would be in gene technologies, biocoupling, and bionics. Performance-enhancing compounds and procedures could be fieldable around 2020; bionically linked man-machine systems would become possible around 2030.
Antimateriel (soft kill) products will disable propulsion systems, change the characteristics of soil or vegetation, or degrade warfighting materiel. The Technology Group specified only materiel and terrain as targets for these weapons. Gene technologies, biomolecular engineering, targeted delivery systems, and bioproduction technology would be required. Products to target supplies could be fielded soon after 2010. Antimachine products would come later.
The Technology Group on Advanced Materials assessed the following areas of technology:
Resin matrix composites use a polymer-forming organic resin, such as a thermoplastic, in which other structural materials are embedded.
Ceramics technologies produce reaction-formed ceramics, cellular ceramic materials, ductile-phase toughened ceramics, fiber-reinforced ceramics, and diamondlike coatings.
Metals technologies work with special steels (e.g., high-strength, laminated, and steel-based composites); light metal alloys; metal matrix composites; and heavy metals (e.g., tantalum, uranium, and tungsten alloys).
Energetic materials are the basis of explosives, propellants, and pyrotechnics.
This Technology Group expects five materials technologies to have major importance for the Army and recommends them as strategic technologies for special funding consideration: affordable resin matrix composites, reaction-formed structural ceramics, light metal alloys and intermetallics, metal matrix composites, and energetic materials. Reasons for this selection are given under the individual headings below.
The combined benefits of resin matrix composites, advanced ceramics, and light metals will make possible a new breed of ground vehicles. They will be lighter, less costly, and transportable by air, while also being hardened against ballistic attack and more compatible with techniques for low observability.
The Technology Group also identified three pervasive trends in materials science and technology. These trends involve increased use of (1) supercomputers to design materials and to model performance, (2) technology demonstrators to ease the transfer of new materials or processing techniques from the laboratory to the field, and (3) materials and structures designed to serve more than one purpose where multiple layers of single-purpose materials had been used.
The TFA ends with a summary forecast of advances in armor materials available for systems in the near term (up to 15 years) and the far term (30 years and beyond). For this summary, armor applications were divided into three categories: protection of the individual soldier, protection of aircrew and critical aircraft components, and protection of combat vehicles. In the near term, the following advances are expected:
Ceramic armor will be cost efficient as well as ballistically efficient.
Composite structural armor will improve in performance, while manufacturing and maintenance costs decrease.
Advanced aluminum alloys and aluminum-matrix composites will increase performance and reduce armor weight for light-armored vehicles.
High-performance ceramic glasses will improve transparent armor and dilatant armor for defeat of shaped charges.
Microstructural texturing will increase the ballistic performance of steel and other metallic armors.
Polymeric fibers will provide enhanced protection for the individual soldier.
In the far term, the above improvements will be extended and the following new capabilities will emerge:
Armor materials will be designed for multiple functions (ballistic protection, signature reduction, etc.).
Biologically engineered fibers will enhance ballistic protection for the individual soldier.
Nonlinear models for dynamic systems will lead to major advances in the design of materials for ballistic protection.
Resin Matrix Composites
Recent major breakthroughs in processing will significantly reduce the cost of resin-based organic composites. Conventional processing involves the application of heat and pressure, typically through use of an autoclave. Heat transfer depends on relatively slow conduction and convection mechanisms. A nonconventional alternative is electromagnetic curing, which uses microwave-absorbing features of the matrix resin (pendant groups attached to the polymeric backbone) to generate heat evenly and immediately throughout the material. The heating to set the resin can be done in minutes instead of hours. In addition, these advanced processing methods will allow complex parts to be made in one operation. Key Army applications are ground vehicles and in situ structural reinforcements.
Ordered polymers make use of the inherent strength of the carbon-carbon bond by increasing the density of such bonds in the material. If adjacent polymer chains can be more closely aligned, the matrix's mechanical properties can be substantially improved. The chemical backbone of the polymer can also be modified to make the chain itself more rigid. Incorporating these rigid-rod polymers within another host matrix gives a molecular-level reinforcement.
Composites that are reinforced in this way, at the molecular level, are easier to process and can be easily fabricated into components with complex geometries. Body armor for the individual soldier is one area of potential applications. Also, because organic polymers are generally transparent to electromagnetic radiation (e.g., from radars), a structure fabricated from polymer-reinforced composites has low-observable characteristics.
"Toughness" refers to the ability of a material to absorb energy with minimal damage and to resist crack propagation. In the development of organic matrix composites, high toughness has often conflicted with competing requirements for high strength and stiffness. Anticipated future advances in the molecular engineering of polymer
structure and in materials engineering of matrix composition will result in matrix materials that are both tough and strong (Figure 3-18). Thermoplastic matrices are tougher than thermoset matrices. In the composite material, the choice of reinforcement further affects toughness. The gap between the toughness of metals and that of organic matrix composites is decreasing; further research may yield organic composites with the toughness characteristic of metals.
The stability at high temperature of organic composites has also been improving. Molecular engineering of the polymer backbone has raised the continuous-use temperature from 175°C for epoxy to more than 400° for ladder polymers consisting of joined carbocyclic or heterocyclic rings (Figure 3-19). Biosynthesis may provide a cost-efficient production method for the resin monomers with the structure needed to form these polymers. The high-temperature composites made possible by these resins will be important for components
of vehicle power plants and weapons. In particular, substituting polymeric materials for metallic components in the hot areas of engines can reduce the heat signature of vehicles.
A smart composite has active or passive sensing elements embedded in the matrix. Passive sensor elements, such as optical fibers, can be used to detect temperature, pressure, and stress levels within the composite, either during its formation or when it is in use. During manufacture this information can be fed back to the processing control system. Interrogation of the embedded sensors during the structural lifetime of the composite can detect deterioration and combat damage. Active elements can be used to alter physical and structural properties of the composite during use. For example, vibrational damping characteristics of a composite beam can be varied by embedding a dispersed fluid that changes in viscosity in response to an electric current passing through embedded conducting fibers. Another potential use for active smart composites is in gas turbine engines, where properties of the compressor section could be modified to correct for wear, damage, or mission requirements.
Increased use of organic composites raises the question of bonding structural components of either the same or different composition. Adhesive bonding has major advantages over mechanical fastening when the bond is intended to endure for the life of the structure. Knowledge of the mechanisms for cohesive failure within the bonding layer is adequate to predict failure rate, but failure at the interface between layers is less understood. Further research in this area would benefit applications in battlefield repair of structures and reduce the time and cost of logistical support for maintenance and repair.
In reaction-formed ceramics, reactions among constituent materials during consolidation results in final structures with the same shape and dimension of the preconsolidation mix. So the process step that creates the ceramic material also shapes it to its near-final form. In addition, fiber-reinforced ceramics are easier to produce, because shrinkage of the matrix away from the reinforcing material is eliminated. This ability to preform ceramics and ceramic-composite articles will greatly reduce costs for ceramic applications in armor, antiarmor, and diesel engines.
For the near term (to 15 years out), the most important issue for reaction-formed ceramics will be control of the reaction rate. Controlling this rate requires continued research on reaction mechanisms
and accumulation of experiential data. The intrinsic reaction time of mechanisms now in use ranges from slow reactions that take days (e.g., chemical vapor infiltration) to near-explosive reactions performed in seconds (e.g., self-propagating high-temperature synthesis, or SHS). The techniques in the intermediate range, which take minutes to hours, appear most promising for the manufacture of complex-shaped components.
The Technology Group forecasts that in the far term (30 years out), reaction-formed techniques will begin replacing conventional sintering technology even for well-established, low-cost items. This technology will affect military applications such as armor, power and propulsion, gun barrels, missile guidance, and the packaging of electronic systems.
Cellular ceramics are porous rather than monolithic; they have a foamlike structure. In assessing the properties of a cellular ceramic, the appropriate comparison should be with materials it might replace rather than with the pure, dense form of the same ceramic. For example, a ceramic foam may have superior characteristics to a structural organic polymer, including greater tolerance for high temperature and chemical exposure. In some cases cellular ceramics may even have greater specific strength than a pure monolith of the same material.
For the near term, the Technology Group forecasts improved understanding of how to design a cellular ceramic with optimum properties, a growing body of data on these materials, development of processes for engineering the cellular microstructure, new coatings for cellular ceramics, and initial use of cellular ceramics in structural and electronic applications. For the far term, from 30 years out, the Group foresees cellular ceramics replacing dense structural ceramics and metals for both land-based and space applications. Their low dielectric constant will make them well suited for electronics packaging. Military uses may include heat shields in engines, high-temperature structural materials, thermal management in high-power electronics systems, and as a technique for low observability.
Ceramics can be toughened by the inclusion of a dispersed ductile phase, such as a metal or metal alloy. New processing methods, which make manufacturing easier, can be applied to a wide range of ceramic-metal combinations. While these ''cermets'' are naturally well suited for low-temperature applications, further work is needed on high-temperature composites. For the far term, the Technology Group expects cermets to be widely used in large-scale structural applications. Military applications, which will make use of the low manufacturing cost, light weight, scale-up potential, and unique properties
of cermets, will be in such areas as armor, gun tubes, structures for high-power electronics, and moderate-temperature engine components.
In fiber-reinforced ceramics, either fibers or whiskers of a ceramic (silicon carbide) or carbon are embedded in a ceramic matrix to improve the strength, fracture toughness, modulus, or thermomechanical properties of the matrix. Also, the composite often shows "graceful," or gradual, fracture rather than catastrophic failure. Either reaction-forming (see above) or pressure methods of manufacture are necessary to prevent shrinkage of the matrix away from the reinforcing fibers. Another problem is susceptibility to oxidation at high temperature.
For the near term, the Technology Group expects fiber-reinforced glass and glass-ceramic composites to be used for relatively small components. Developments will focus on improving the fiber materials and using fiber coatings to overcome high-temperature oxidation and tailor interfacial mechanical properties. In the far term, larger structures will become feasible. Near-net shape-processing methods (such as reaction forming) will predominate, multidirectional fiber weaves will be available, and joining techniques will mature. Fiber-reinforced ceramics will find military applications in armor, propulsion and power, metal-cutting and metal-forming tools, gun barrel technology, and spacecraft structures.
Diamond and diamondlike coatings are thin films deposited on a bulk substrate. The coatings give hardness, high thermal conductivity, infrared transparency, and other properties associated with diamond. Among the many potential military applications are abrasion-resistant coatings for sensor windows and other optics, high-power transistors and optically activated switches for high-power microwave or millimeter-wave sources, wear-resistant coatings for bearings and journals, substrates and insulating films for high-power electronics, and inert wear-resistant coatings for medical implants.
Even though ferrous metal technology is relatively mature, significant advances in improving properties are possible. Research into the relations between structure and particular properties, such as tensile strength and toughness, can open the way to process changes. A recent example is research into the role of sulfide inclusions in toughness; advanced techniques for producing low-sulfur steels have been able to increase toughness at a given strength. The Technology Group foresees further developments, following this pattern, with respect to mechanisms of failure ahead of a crack tip, further work on toughness, decreased susceptibility to hydrogen embrittlement in gear and
bearing steels, and surface treatments to improve resistance to wear and corrosion.
Laminated steels offer the possibility of combining the high wear resistance and strength of ultra-high-carbon steels with the high impact-resistance (toughness) of lower-carbon steel or other materials. The processing characteristics of such laminates also appear favorable.
Recently, the slow, evolutionary rate of advance in light metal alloys has been revolutionized by the development of an aluminum-lithium alloy with greater specific strength and stiffness than steel. Processing innovations have also raised the commercial potential of improved and new light-metal alloys. For instance, powder metallurgy for rapidly solidified (PM/RS) aluminum alloys offers potential for materials that will compete with titanium alloys for high elastic modulus and strength at high temperature. Although problems in areas such as fatigue and fracture toughness still need resolution, these alloys are likely to find important applications in engines and robot vehicles.
PM/RS has also increased the prospects for magnesium-based alloys. The improvements in alloy properties are similar to those of PM/RS aluminum alloys. Applications for structural members in helicopters and other lightweight vehicles appear to be their major uses. Similarly, new lighter intermetallics will eventually replace heavier nickel or cobalt-based alloys.
Advances in knowledge of composites based on steel and aluminum matrices will benefit the entire area of metal matrix composites. The Technology Group forecasts significant expansion of this area in the next 10 to 20 years, with revolutionary effects on the properties of steels, titanium, magnesium, intermetallics, copper, and heavy metals, as well as aluminum.
In addition to the advances in ferrous metal techniques described above, properties of steel such as wear resistance and tolerance of high temperatures may improve through the use of steel composites . Particles of a wear-resistant compound, such as TiC or TiN, are added to the steel. Research in this area is needed, as the effects of such particulates on high-temperature properties are not known. Stainless steels, which resist oxidation at high temperature, may be significantly strengthened by addition of these high-strength particulates. Applications with military significance include gas turbine engine components, transmission housings, and advanced gun systems.
Similarly, aluminum-based composites have ceramic particulates or whiskers added to the aluminum matrix. Addition of these reinforcements significantly increases stiffness while reducing the coefficients of thermal expansion and thermal conductivity. Composites with the stiffness of titanium are projected. Here, too, further research will be
needed on the effects of composites on other properties of the matrix metal. Other reinforcing materials and processing techniques also should be investigated. Some recent work indicates significant improvement of toughness in aluminum matrix composites by controlling the microstructure, but further investigation is needed, as is research on strain rate.
Particulate ceramics also offer great potential for reinforcing other metal and intermetallic matrices. The Technology Group cites one new process that appears usable for the addition of carbide, nitride, or boride ceramic particles into aluminum, copper, titanium aluminide, or nickel aluminide matrices. The resulting composites have specific properties superior to the superalloys, based on nickel or iron, now in use; they also offer a 50 percent weight reduction. Another advantage of metal matrix composites is their potential for advanced forming techniques, such as superplastic forming. The Technology Group forecasts applications of these composites in lightweight armor, missile components, rotating structures, gun barrels, and electrically powered guns.
The Technology Group focused on two areas of heavy metals technology: tantalum warheads for shaped-charge penetrators and alloys of depleted uranium or tungsten for use in kinetic energy projectiles. Materials research on tantalum penetrators focuses on metallurgy to produce penetrators that respond well to the explosive deformation processes that form them and do not break apart before reaching the target. Code-research groups are working on mathematical models to bridge the gap between the microscopic results of metallurgy and macroscopic experiments in shock-wave physics.
Most of the current research on depleted uranium alloys aims at improving the methods of thermal and mechanical processing of this highly anisotropic metal. Two promising areas are new alloy casting techniques and the use of PM/RS technology (described above for light metal alloys). Tungsten alloys continue to underperform depleted uranium alloys as kinetic energy projectiles against heavy armor. However, tungsten alloys work as well as depleted uranium against light armor; further research may improve their performance against heavy armor to the point where they are fully acceptable as an alternative to depleted uranium.
The Technology Group identifies two emerging technologies that are likely to change Army energetic materials significantly. New organic cage compounds will have higher material densities, and there-
fore more explosive power, than conventional organic explosives such as HMX and RDX (cyclotetramethylene tetranitramine and cyclotrimethylene trinitramine, respectively). Entirely new energetic materials are also possible by separating thin layers of inorganic reactants with inert barrier layers (Figure 3-20). The layer thicknesses would be on a micron scale.
Research also continues in methods to make energetic materials less sensitive to uncontrolled stimuli (fire, ballistic impact, or explosion of nearby munitions) without sacrificing their performance. The Technology Group forecasts progress in understanding the phenomena of sensitivity during the next decade, including computer modeling capability to design insensitive materials with high energy density before synthesizing them in the laboratory. Within 20 years, processing technology will allow smaller particle size and better control of the interaction between energetic material and the matrix of binder and plasticizers.
In the processing of energetic materials, biotechnology is likely to become important not only for the biodegradation of hazardous waste products but also for the synthesis of energetic molecules.
PROPULSION AND POWER
The Technology Group on Propulsion and Power has assessed the following areas of technology:
High-power directed energy covers the technologies for high-energy lasers (beamed radiation at optical frequencies) and directed beams of radio frequency, microwave, and millimeter-wave radiation.
Propulsion technologies. The Technology Group divides propulsion technologies into four categories: missile propulsion, air vehicle propulsion, surface mobility propulsion, and gun or tube projectile propulsion.
Battle zone electric power covers generators of continuous electric power, generators for pulsed and short-duration power, and energy storage and recovery for ultimate use as electric power.
High-Power Directed Energy
High-power lasers, microwaves, and directed energy weapons offer the opportunity to disable or destroy enemy systems. Four promising laser types currently under investigation for this purpose are chemical lasers, free-electron lasers, ionic solid state lasers, and coherent diode laser arrays. Potential laser applications include ground-based ballistic missile and antisatellite defense, air defense, antisensor, and antipersonnel use. High-power microwave (HPM) technology offers the opportunity to physically damage enemy systems in the same types of applications as optical-wavelength lasers. As an energy source, beamed microwave power could also be used to keep remotely piloted aircraft aloft indefinitely. In the findings below, these various directed energy beam technologies will be referred to generically as high-power directed energy (HDE).
The Technology Group selected five HDE technologies as high-leverage areas for Army support: (1) ionic solid state laser arrays; (2) coherent diode laser arrays; (3) phase conjugation for high-energy lasers; (4) millimeter-wave generators (high-power); and (5) multiple-beam, klystron HPM technology. Each received a detailed assessment and a figure-of-merit scoring.
In considering high-energy lasers, the Technology Group focused on requirements in four areas of application:
For ground-based ballistic missile defense, the laser source produces a beam that is reflected by space-based relay mirrors onto the target missile while it is still in its boost phase. Lasers for this application must have very high power to deliver sufficient kill power in a short time over a long range. Wavelength selection is important to minimize power loss to the atmosphere.
Antisatellite laser requirements are similar to those for ballistic missile defense. For targets in low-to-medium orbit, the power required may be an order of magnitude lower, because the targets are softer and distances are shorter.
Air defense against aircraft and cruise-type missiles requires a laser that is effective at less than 10 km. The power requirement is an order of magnitude less than for antisatellite weapons. To be practical, the laser system must be transportable, perhaps even highly mobile. It must also be able to fire repeatedly at multiple incoming targets (100 to 200 shots at 2 to 3 s each) at multiple incoming targets before fuel reloading is needed. (In the terminology of conventional ballistic weapons, it must have a large magazine capacity.)
Antisensor lasers must be capable of crazing or destroying the optics, detectors, or other elements of sensor systems. The power requirement is moderate. The laser wavelength may be either in or out of the operating bandwidth of the threat sensor, but the source, together with its beam control and fire control subsystems, must be in a package suitable for mobile operation at or near the forward line. The requirements for antipersonnel systems are the same.
Ionic solid state laser arrays were selected as a high-leverage technology for the Army. Their output takes the form of pulses with high peak power. They can be built for either low pulse rate and higher peak energy (glass host) or high pulse rate and lower peak energy (crystalline host). At present, they have efficiencies of 2 to 5 percent when flashlamp pumped and 10 to 15 percent when pumped by (incoherent) diode laser arrays. They are currently used for range finding, target designation, remote sensing, and communications.
For weapon applications, ionic solid state lasers must be scaled well beyond their current performance levels. The Technology Group forecasts improvements to possibly 30 percent efficiency and drastic cost reduction through pumping with diode laser arrays. Complete tunability from the mid-infrared to the ultraviolet region is also possible, although R&D work is required. Also, to make ionic solid state lasers acceptable for tactical applications, the cost of the diode laser arrays used to pump them must be reduced by one or two orders of magnitude. The Technology Group forecasts that crystalline-host lasers could achieve average powers greater than 500 W per aperture, delivering 1 to 4 J of energy at pulse rates of 200 to 500 Hz. Glass-host lasers could achieve outputs of more than 1 kJ at repetition rates of 1 to 3 Hz. Coherent coupling could raise outputs to levels 10 to 30 times higher than that.
Coherent diode laser arrays, another high-leverage technology choice, are tunable by adjusting the composition and temperature of the semiconductor whose bandgap transition is the radiation source. The Technology Group forecasts that power levels of 10 to 1,000 W per modular unit, with energy fluxes as high as 1 kW/cm2, can be achieved in the next 5 to 10 years. Efficiencies of 50 percent appear achievable. The main development issues for extending this technology to weapon system power levels of tens to hundreds of kilowatts are (1) extending the mechanisms for phase locking hundreds or thousands of individual diode laser modules into an extended coherent array and (2) managing the waste heat generated by high-power operation.
The Air Force is currently conducting research on coherent diode laser arrays. The Technology Group recommends that the Army not only monitor the Air Force's projects but also pursue complementary work.
In addition to these two high-leverage laser technologies, the Technology Group reviewed two other laser source technologies: chemical lasers and free-electron lasers (FELs).
The most highly developed chemical laser is hydrogen fluoride/deuterium fluoride (HF/DF). Its major advantage is that electric power is not required for lasing, which is produced from direct chemical reaction. However, to be operated in continuous mode, the laser cavity must be at low pressure; in pulsed mode it may be operable at atmospheric pressure. A major development forecast by the Technology Group is short-wavelength (visible or ultraviolet) chemical lasers.
The mode of operation of FELs, as well as ongoing Army research on them, has been discussed above for the TFA on Optics, Photonics, and Directed Energy. The main future use of FELs will be in ground-based ballistic missile defense and antisatellite weapons. This Technology Group does not consider them practical for mobile hard-kill weapons, although they may prove usable for antisensor countermeasures and soft-kill weapons.
Among the laser beam control technologies reviewed by the Technology Group, phase conjugation was selected as a high-leverage technology for the Army. It is a relatively new, nonlinear optical process that can be used to correct dynamic beam aberrations in real time. It can be used for beam cleanup, beam combination, and array phasing. It is simpler to integrate into existing systems and has a faster time response than conventional adaptive optics. The capabilities of phase conjugation have been demonstrated on pulsed and continuous-wave lasers using a variety of source methods. There are two processes in
use, one suitable for high-power lasers, the other for beams with low or moderate power. By relaxing optical tolerances and replacing conventional adaptive optics, phase conjugation can reduce both the cost and complexity of moderate-power and high-power lasers.
As a method for combining beams, phase conjugation can conceivably allow large increases in laser output power. An n-by-n array of semiconductor diode lasers, each of power P, can in principle be phase-conjugated to produce an output beam with intensity n2P. Phase conjugation may also be applicable to optical countermeasures; a threat beam can be phase conjugated, amplified, and returned against its source, causing the source's destruction. By 2020, phase conjugation cells to track targets automatically and point directed energy beams could be well developed.
Radio frequency, HPM, and millimeter-wave weapons. The advanced electronics technology of the modern battlefield—advanced sensors, smart weapons, and autonomous systems—typically use radio frequency, microwave, or millimeter-wave signals ranging from 100 MHz to 300 GHz. The primary tactical significance of high-power electromagnetic pulse energy beams that operate in this same region lies in their potential to damage the electronics necessary for hostile sensors and systems to function.
Of eight technology options in this frequency regime that the Technology Group reviewed, it selected two as high-leverage technologies for the Army: coherent multiple-beam HPM energetic pulse klystrons and millimeter-wave sweeping-frequency generators.
An HPM pulse kills its target by coupling to electronic circuitry that it accesses through either functional openings or cracks in the target's shielding. The thermal energy from this coupling burns out device junctions. Although this nonselective mechanism does not require tuning to the threat system's operating frequencies, it does require a relatively high power density at the target. Together with the desired long range of the HPM weapon, this implies a high peak-power requirement for the beam source.
The klystron tube is basically an amplifier technology; a microwave-modulated beam of high-current electrons generates output microwave power with as much as a 50-dB gain over the input power used to modulate the beam initially. A device with a single electron beam at 500 kV and 300 A can produce 60-MW peak microwave power in an output beam that pulses at more than 100 Hz. Current research is focusing on the use of multiple-beam klystrons. This research is being aided by simulation codes, which allow comparison of design alternatives prior to selecting the best design for demonstration.
Assuming that adequate funding and technical manpower are avail-
able, the Technology Group forecasts that a multiple-beam klystron HPM weapon with peak power of several gigawatts can be fielded by 2000 (Figure 3-21). The peak power could be in tens of gigawatts by 2010. Achieving these goals will require substantial development beyond the current state of the art for components such as power conditioning systems, modulators, and thermionic electron cathode sources, as well as the klystrons. The forecast includes an outline and funding levels for the development effort needed to achieve fielded systems by the projected dates.
A millimeter-wave sweeping-frequency generator would produce tunable-frequency pulses of radiation in the range of 15 to 300 GHz. The kill mechanism is to engage target apertures and sensors to cause electromagnetic damage. Although the required energy delivered on target can be much less than in the HPM mechanism, the beam frequency does need to move through the operating range of the aperture or sensor. Multiple gyrotron oscillators can be used as beam sources to cover the range from 15 to 50 GHz, while FEL lasers can be used in the range of 50 to 300 GHz. With a high-power narrow beam, the effective range of this weapon can be hundreds of kilometers. Other requirements of an effective system are a heavy-duty cycle, at least 1-MW power to defeat smart weapons with millimeter-wave imaging or infrared guidance, and a frequency sweep range of 30 to 300 GHz.
The Technology Group forecasts that gyrotron technology will produce megawatt-level continuous-wave power by about 1995, but it will be at fixed frequencies (Figure 3-22). In the same time frame, FEL technology will provide 1-MW power levels with limited tuning capability and pulse rate. The FEL technology for tunable frequency
sweep across the entire 30 to 300 GHz range with continuous-wave 1-MW power can be achieved by 2005. These projections assume a program funded at the levels recommended by the Technology Group.
The Technology Group assessed multiple technology options in each of the four propulsion categories, as shown in Table 3-2. It recommends four options, one from each category, as high-leverage technologies for special consideration by the Army. These options are shown in bold and underlined in the table.
The Army's future missile propulsion technology must be adequate for new generations of smart-to-brilliant missiles. To achieve the range variation, minimum flyout times, targeting flexibility, and accuracy required for both offensive and defensive battle zone missions, these missiles must be capable of extremes of energy management and maneuverability. Therefore, the propulsion technology must provide the missile designer with broad options for peak thrust levels, thrust-time profiling, and multiple restart with pulse durations as short as tens of milliseconds.
TABLE 3-2 Propulsion Technology Options
Surface Mobility Propulsion
Air Vehicle Propulsion
Gun or Tube Propulsion
Gas turbine jet
Gas turbine engine
Electromagnetic Hydrogen cannon
Of the rocket propulsion technologies assessed by the Technology Group, gel propellants offered the most potential for new technology. They are generally less sensitive and safer to handle and store than either liquid or solid propellants. Under high pressure, they shear like liquids, so they have the superior energy management characteristics of liquid propellants (multiple stops and starts, proportional throttling). Their performance characteristics may be even higher than liquids, which in turn are generally higher than solids. Gels also aid in signature management, because chlorine and carbon products can be eliminated from the motor exhaust.
Solid propellants will see evolutionary improvement in specific impulse and other performance parameters, as new systems such as glycidal azides or high-strained carbon bond fuels substitute for the current nitrate-ester-plasticized polyethane (NEPE). However, energy management will remain difficult to implement with solid propellants.
Among the air-breathing missile propulsion technologies that the Technology Group assessed, turbine engines and ducted or air-augmented rockets showed the most promise for significance to the Army in 2020. The Technology Group expects that progress in these technologies will follow as a matter of course from various programs already in place among the military services. Substantial decreases in specific fuel consumption and increases in thrust per unit airflow are expected for future gas turbines.
For aircraft propulsion, foreseeable developments in gas turbines offer the potential to double current performance by increasing the
power-to-weight ratio and thrust-to-airflow ratio and by reducing fuel consumption. Advanced materials will reduce the weight of structures and components. (See Advanced Materials TFA for details.) Current programs will improve the performance of the Army's rotary wing and fixed wing manned aircraft. The Technology Group recommends that the Army rely largely on the ongoing Integrated High-Performance Turbine Engine Technology (IHPTET) program.
The high-leverage air propulsion technology selected by the Technology Group is the use of HPM (high-power microwave) to power high-altitude surveillance UAVs from ground transmitter sources (Figure 3-23). The components most in need of Army-specific development are the ''rectenna,'' which is located on the underside of the vehicle's wing to capture microwave radiation and convert it to electric current, and the ground-based transmitter. To lessen the UAV's signature when it is in hostile airspace, it can carry rechargeable batteries and use supplementary photovoltaic cells on the wings' up-
per surface. Periodically, it would fly back to safer airspace to have its batteries recharged from an HPM transmitter station.4
In the propulsion category of surface mobility, the Technology Group assessed primary power production (engines), methods of power transmission or distribution, and mechanical subsystems and components. It also assessed three general conceptual approaches to surface vehicle propulsion. Two of these general concepts, the Integrated Propulsion System (IPS) and hybrid-electric propulsion systems, received special consideration as aspects of the Group's overall high-leverage technology for surface mobility: a system designed under an IPS approach, having an advanced diesel or gas turbine engine and either all-electric or hybrid-electric power distribution.
The third general mobility concept, ground-effect machines, was not evaluated in detail by this Technology Group. (The technology is discussed at some length in the Mobility Systems Panel report.)
The likelihood that Army combat vehicles will in the future face enemies that have numerical superiority places a premium on vehicle survivability. One response is to improve armor; another is to produce more mobile, agile vehicles that are lighter in weight, have smaller silhouettes, and incorporate low-observable technology. The technology to enable the second response depends on lightweight low-volume propulsion systems (Figure 3-24).
For primary power, the Technology Group forecasts evolutionary progress in diesel engine performance through turbocompounding and stratified-charge combustion. In turbocompounding the engine exhaust gases are used to drive an input-air compressor (turbocharging) that is also linked to the flywheel. Stratified-charge combustion uses a spark-ignited precombustion chamber to ignite a leaner fuel-air mixture in the main chamber. The other promising option for future engines is an advanced gas turbine, pursuing the evolutionary line represented in the M-1 class Abrams tank engine.
The Technology Group foresees electric drive as the most promising power distribution system for the far term (circa 2020), despite negative assessments of it by the Army in previous decades. The Group's support for either an all-electric or hybrid-electric system takes into account dramatic improvements in all areas of electric-drive technology during the 10 years since the last Army review. The advantages
relative to other power transmission approaches include (1) improved weight distribution, since components are modular; (2) individually driven wheels or track drive sprockets, eliminating complex transmission/differential drive trains; and (3) a common power distribution system for vehicle drive, electrically powered weapon systems (such as an electrically energized hypervelocity gun), and power storage.
A hybrid-electric propulsion system includes a mechanical transmission link between the engine and the wheels or track drive sprockets, in addition to the generator and traction motor subsystem for the electric drive.
IPS (integrated propulsion system) is a conceptual approach for applying integrated systems design to propulsion systems, rather than a propulsion method (Figure 3-25). Under IPS all aspects of the
system—engine, transmission, cooling system, air filtration, auxiliary power, inlet and exhaust ducts, diagnostics, signature reduction, and even maintenance—are designed interactively, rather than by small teams designing components in isolation. Currently, IPS is being applied by the Tank-Automotive Command for the design of a diesel-based system and a gas turbine system. These designs should increase power density by at least 50 percent, with a similar increase in horsepower per unit weight.
The Technology Group recommends that the IPS approach also be applied to long-term consideration of a propulsion alternative that incorporates electric drive. For example, a hybrid electric-drive propulsion system offers major gains in total battle zone effectiveness, gains that will be enhanced by electrical energy storage systems with far higher electrical power density than current technology. Two such systems that offer particular promise are advanced batteries (perhaps a fivefold improvement over lead-acid batteries) and flywheels (fourfold improvement or more).
For propulsion of projectiles from guns or tubes, the Technology Group reviewed chemically energized propulsion, electrically energized propulsion, and hydrogen cannon (hot gas propulsion). The chemical propulsion options reviewed were modular charge, rocket-assisted projectiles, traveling charge, liquid propellants, and ram guns. The Group selected chemical propulsion by liquid propellants and elec-
trically energized guns as the technologies having the most potential to augment gun capability for the future Army.
Liquid propellants are an evolutionary approach to gun propulsion technology. In the ongoing Army program, state-of-the-art gun barrel and recoil mechanisms can be used; only the breech must be redesigned. Liquid guns with 25-, 30-, and 105-mm rounds have been successfully demonstrated. Liquid propellants are relatively insensitive to shock, are stable in long-term storage, and offer a number of performance benefits. Further evolutionary improvements to the current technology are possible.
Two new types of guns will use electrical energy, in whole or in part. In the electrothermal chemical (ETC) gun, as in a gun using conventional propellants, the projectile is accelerated by the high-pressure gases created in the gun tube. In an ETC gun these gases are created by the combustion of normally inert materials at the very high temperature of a plasma, which is created by a pulse of electrical power supplied to the gun breech. Provided the pressures can be tolerated, the ETC gun can achieve projectile velocities that are perhaps 30 to 40 percent higher than can be efficiently achieved with conventional propellants. This is certainly as high as is needed for field artillery systems.
In the electromagnetic (EM) gun, the projectile is accelerated by electromagnetic pressure instead of gas pressure. EM guns have been demonstrated with very small projectiles at velocities of 7 km/s. With projectiles of the mass needed for air defense and antiarmor roles, velocities well over 2 km/s have been achieved. Currently, the pulse power conditioning unit for the EM gun is too bulky for compact ground vehicles, but it is expected to be reduced substantially in the next few years.
On current evidence, both ETC and EM guns may have a place in the set of future Army armaments.
Battle Zone Electric Power
The future Army will require electrical power in the battle zone at levels from tens of watts for surveillance and communication to hundreds of megawatts for directed energy weapons. Mobility will be essential. For mobile continuous-power generation, the key to substantial weight reduction is to increase the generating and distribution frequency from the current standard of 60 Hz to 400 Hz or higher.
Internal combustion turboshaft (gas turbine) engines, which are already in use for mobile electric power, offer more potential for the future than the alternatives (internal or external combustion piston
engines and fuel cells) for mobile continuous-power generation. A turbine running at 24,000 rpm can drive a 400-Hz alternator directly, without the heavy gearing now needed to drive 60-Hz alternators. Continued Army support for the IHPTET program (recommended above for aircraft propulsion technology) can realize the potential of this technology, when coupled with an aggressive effort to advance the technology for high-frequency, lightweight alternators, power conditioners, and distribution system. This effort could raise the power-to-weight ratio for mobile electric power units from the range of 0.05 kWe/kg (kilowatts electric per kilogram) for current 60-Hz gas turbine units to more than 3 kWe/kg.
In power generation and distribution, the use of high voltages can also decrease weight; the conductor weight required for a given wattage decreases as the square of the voltage. Improvements in high-voltage semiconductor devices would allow an increase from the current limit of about 1 kV to levels at which a power-to-weight ratio of 5 kWe/kg would be possible for a mobile electric unit.
As a primary power source, fuel cells would become practical only if a breakthrough occurs that would allow liquid hydrocarbons and air to fuel them.
Directed energy devices and other electrically energized high-power systems of the future Army will require generators for pulsed and short-duration power whose average power for the duration of output ranges to hundreds of megawatts. In both the mass and bulk (volume), generators in this class are half prime power unit and half power conditioning unit. For the prime power unit, the technologies with the most promise for the Army were judged to be gas turbine engines (for energy production) and flywheels (for energy storage). For power conditioning, new, molecularly tailored solid state devices and improved methods of heat removal should make possible an order-of-magnitude reduction in weight.
For power conditioning in pulsed or short-term generators, the Technology Group sees the development of high-temperature, high-power electronics as a crucial area. In particular, continued evolution along present lines must be pursued for capacitors, inverters, switches, and transformers. For each of these component types, the combination of high voltage, high frequency, and high power requires technology that is beyond the current state of the art but not out of reach.
Energy Storage and Recovery
Reducing the observable signature of power generation units in the battle zone will become increasingly important. Technologies that allow storage of power in low-signature devices, such as secondary
batteries or flywheels, will become critical for the short but intense conflicts on future battlefields. For example, mobile systems may move into position using internal combustion engines for locomotion; under battle conditions they would switch to their onboard energy storage devices. Of the storage device technologies reviewed by the Technology Group, rechargeable batteries and mobile (vehicular) flywheels were selected for their broad applicability to Army needs in 2020.
Rechargeable (secondary) batteries capable of a large number of discharge/recharge cycles probably will play a far greater role in the future Army than they have in the past. The current state-of-the-art lead-acid battery needs to be replaced with an innovative technology. The Technology Group forecasts an increase in energy density by a factor of four or five and of power density by two or three for a new battery technology relative to current lead-acid batteries. It projected future (2020 time frame) performance parameters for five battery technologies now in the research stage.
The anticipated advances in flywheel technology will come primarily from new composite materials with high ratios of tensile strength to weight (Figure 3-26). These materials will increase energy density
by an order of magnitude over flywheels made with high-strength steels. The cost of fabrication for composite flywheels should also decline dramatically over the next 30 years.5
The Technology Group for Advanced Manufacturing focused on the systems aspects of manufacturing rather than individual process technologies. Specifically, the Group reviewed the following topics:
Key technologies include intelligent processing equipment, microfabrication and nanofabrication, flexible computer-integrated manufacturing, and systems management.
Applications important to the Army include distributed and forward production facilities, rapid reaction to operational requirements, and parts copying.
Issues in manufacturing technology include sources of supply, availability of materials and components, military versus civilian R&D industrial preparedness, capital investment and facilities, flexible production schedules, design for manufacturability, and environmental and legal issues.
The focus for technological advances in the next generation of manufacturing will be on the inclusion of information systems with the energy systems and material management systems introduced by previous generations of technology. Instead of having only persons and machines involved in a manufacturing process, automated machinery now includes some form of computer control based on feedback from sensors.
Adding information systems to manufacturing results in major improvements in accuracy, reliability, and quality. For example, auto-
mated machinery can (1) reduce the cost for increased functionality; (2) enable civilian specifications and quality standards high enough that separate military specifications are not necessary (which implies that civilian production facilities can be used for military production); and (3) significantly decrease the time from concept to deployed system.
Because production lines can be tailored quickly, order quantities no longer must be large to be economical. Variations in products will neither add cost nor reduce reliability. Therefore, it becomes possible to customize weapons and support gear to fit a specific intended use in a specific environment, rather than requiring an item to fit a broad category of conditions for a design lifetime of 10 to 15 years.
By changing response times for resupply from years to weeks, with the ability to customize for current needs, the inventory that must be maintained is sharply reduced.
Technological advances in material transformation processes combine new scientific understanding of the underlying transformations with automated control systems to monitor and control the process. These changes in processing technology will accelerate three trends: (1) the ability to specify the attributes of a material ("designer materials") will broaden to include the ability to design and fabricate "designer parts"; (2) the information subsystems component of larger systems will increase; and (3) the reproducibility of processes and control information will increase the ability to model variations in process variables and predict system performance.
Intelligent processing equipment can sense (i.e., monitor with appropriate sensors) important properties of the material that are altered by the process, and it has the intelligence to control changes in these properties. Although industrial robots are the most visible component of this technology, to perform they must be coupled with sensor systems and intelligent control systems.
Microfabrication and nanofabrication involve manipulating and fabricating materials at the microscopic or atomic level, respectively. The next generation of integrated circuit chips will require these techniques (see TFA on Electronics and Sensors, above, and Basic Sciences, below). Microscopically applied films and surface treatments are used not only in microelectronics fabrication but also in metallurgy for low-friction bearings and other special characteristics. The potential for low cost and high sensitivity in new devices with microscopic dimensions will make possible microsensors for measur-
ing flow, pressure, chemical concentrations, and other parameters in mechanical, medical, and environmental applications.
Flexible computer-integrated manufacturing (CIM) applies information systems technology to the levels of manufacturing integration above the level of intelligent processing equipment, which applies information control to a single process or workstation. A group of workstations, constituting a factory cell, can be organized around a set of related tasks or functions. Cells are combined into factory centers, which manage subassembly and system assembly operations. At each level, information systems coordinate the manufacturing elements. The CIM system as a whole oversees all the factory's operations, from workstations to cells to centers.
To implement flexible CIM, the factory control systems are supplemented with associated tools and technologies, including simulation models, computer-aided design, computer-aided engineering, group technology, computer-aided process planning, and factory scheduling tools.
Systems management applies information systems at the enterprise level (within or between enterprises) rather than at the level of controlling a specific manufacturing operation (as in CIM). Product data exchange allows business units to exchange computer information generated from their different computer-aided design and computer-aided manufacturing systems. Data-driven management information systems contain the kinds of design, inventory/order, and machine capability information needed to design and manage flexible CIM operations.
Applications Important to the Army
Distributed and forward production facilities consist of manufacturing modules stored together with product subassemblies, raw material, and the electronic knowledge of how to complete the manufacture of finished products to order. The "facilities" are put in place before they are needed. The approach can be applied to simple products, such as clothing, food, and equipment, that can be produced to specific sizes or packaging preference as needed. It is also applicable to larger weapon systems, which can be stored as modules. Upgrades can be made by replacing modules before assembly rather than by retrofitting.
Rapid reaction to operational requirements uses advanced design and manufacturing technology to shorten the cycle from specification to product delivery. For the Army, this application could support specifications sent directly from the field to the manufacturer.
Parts copying uses three-dimensional sensor technology to provide measurements of an existing part, coupled with technologies to etch or sinter raw material to the specified dimensions. Among the potential benefits are storing just one part as a master for copying, making parts without an engineering description, replicating a replacement from pieces of a damaged part, and scaling up or down from one existing size. While this capability is crude and limited at present, it should be applicable to a broad spectrum of parts by 2020.
Issues in Manufacturing Technology
The Advanced Manufacturing Technology Group raised eight issues related to manufacturing for the Army:
Sources of Supply. The current practice of requiring multiple sources of major components or dual-source manufacturing may needlessly increase the cost of defense system acquisition.
Material and Component Availability. U.S. manufacturers are becoming increasingly dependent on foreign sources of basic materials, processes, and components. A number of actions can be taken to ensure that critical materials, skills, and equipment are available if needed by defense forces. Actions may also be needed to limit loss of control over the cost of critical defense materials or systems.
Military versus Civilian R&D. Many of the manufacturing technologies to produce military items will be developed for civilian production first. However, some differences in standards will continue, and some areas of manufacturing will remain unique to the military.
Industrial Preparedness. The ability of the U.S. industrial base to respond to a major mobilization is severely limited. Flexible advanced manufacturing facilities that are capable of rapid conversion from commercial to defense production appear to be the best solution.6
Capital Investment and Facilities. Flexible manufacturing systems can help to reduce the investment risk that current procurement practices have placed on industry. Some special contracting arrangements will also be needed. Otherwise, inadequate investment in cost-
effective production facilities will only drive up the eventual cost of defense systems.
Flexible Production Schedules. Where the same facilities can serve for both civilian and military production, the Army should consider arrangements for extended delivery schedules to allow military production during off-peak demand for civilian goods.
Design for Manufacturability. Manufacturability should be included as a design evaluation criterion from the outset of requirement formulation. The cost and availability of products will ultimately be driven by the ease and flexibility of production.
Environmental and Legal Issues. Environmental concerns and court decisions about them will continue to affect production facilities of interest to the military. Court decisions like those affecting Department of Energy nuclear material plants will force managers in the government and the private sector to take actions that will affect the cost and availability of defense materials and systems.
ENVIRONMENTAL AND ATMOSPHERIC SCIENCES
The Technology Group for Environmental and Atmospheric Sciences assessed the following areas of current and projected technology:
Terrain-related technologies include digital topography, terrain imaging sensors, and terrain surface dynamics.
Weather-related technologies include atmospheric sensing, weather modeling and forecasting, modeling of atmospheric transport and diffusion phenomena, and weather modification.
Military operations depend on information about terrain and weather at both large and small scales. As combat operations place increasing emphasis on force mobility and high-technology sensor-dependent systems, the Army will increasingly require a comprehensive information base adequate to:
characterize operationally significant environmental features (vegetation, soil condition, roads, bodies of water, etc.) of all the land masses of the globe;
determine how these environmental features are affected by global and local weather patterns; and
identify variability in ''local'' weather as a function of locale.
The Army will also need to provide its deployed forces with the capability to sense and interpret the current local conditions of the atmosphere and terrain.
Army units in the field require three categories of information about terrain: topography; environmental features less permanent than topography (roads, amount and type of vegetation, habitation, soil condition); and the capability to anticipate changes in soil condition that may result from weather or enemy action. Topographic data on relatively permanent features can be gathered well in advance of operations and maintained in digital data bases. Advances in digital imagery and in the methods of extracting and storing data from digital images have overcome many of the limitations of two-dimensional maps.
The critical need is for a global, three-dimensional terrain data base. Querying and data retrieval must be easy and fast. Yet quick updating of information must also be supported. Techniques are needed to give field commanders dynamic interrogation and viewing of local terrain, plus the ability to generate hard-copy maps for later reference. The enabling technology includes high-capacity optoelectronic storage media, a data base structure for storing three-dimensional data, software and hardware for rapid processing of large data sets, high-speed broadband communications links, multicolor map production from digitized data, microprocessor workstations as the local nodes in this terrain information network, and artificial intelligence to automate reasoning about the interaction of terrain features and other environmental factors, including the weather.7
In terrain sensing technology, a major breakthrough would be the direct recording of three-dimensional terrain data. An interim evolution is platform-based processing of raw sensor data. Neural network technology can be applied to automated feature extraction. Emerging
technology for the identification of features by their wide-spectrum signature (hyperspectral imagery) will also be applicable.
Technology for real-time terrain analysis will use computer modeling for which input data from the terrain data base are supplemented with current data from weather and soil sensors. Technological advances will be required in high-resolution terrain sensors, direct observation, and the processing of raw sensor data. Hyperspectral imagery can provide information on subsurface conditions as well as surface characteristics.
Weather prediction for operational areas requires atmospheric sensors in the area and the processing capability to synthesize the data into a timely and accurate picture of current conditions, as well as conditions 12 hours to 2 days ahead. The spatial scale for battlefield weather reporting can range from 20 to 200 km.
In the future, atmospheric sensors will be flown into forward battlefield areas with UAVs, where they may remain airborne or be dropped to the ground. To provide information on the lowest level of the atmosphere, terrain-following UAVs will be needed. It may also be possible to extract useful weather data from smart weapons. R&D is needed for passive sensing techniques, because the active methods now in use for atmospheric sensing provide targets for the enemy. For remote sensing, satellite LIDAR and radar systems will gather images in wavelengths from the ultraviolet to the microwave region.
With respect to weather data communications and processing technology, the multispectral data of the future will require broadband, high-capacity communications links. Data may be relayed to the processing center from local sensors via satellites. To lessen the data transmission load, signal preprocessing in the sensor platform will be applicable. Position location technology will be important in pinpointing the location of sensors transmitting data.
Improvements in civilian-oriented weather modeling and forecasting will continue, and the Army will draw upon this technology. In addition, meteorological models are under development for regional use. These have a smaller scale of resolution and rely on sensor data collected on a grid at the same scale. High-speed computers are needed for modeling future conditions from current data and for controlling interpretive displays of both current and predicted conditions. The meteorological community will continue to explore applications of artificial intelligence to weather forecasting. The Army will need to incorporate advances and improve on them
for its particular concerns with battlefield scale, effects of combat on local conditions, and weather effects on tactics and equipment performance.
The Army's interest in atmospheric transport and diffusion modeling stems from concern with the spread and dilution of CTBW threats, airborne nuclear radiation hazards, pollutants, and battlefield obscurants. Breakthroughs in methods for solving nonlinear stochastic and probability equations for physical, chemical, and meteorological phenomena will allow more realistic modeling of transport and diffusion processes. The projected increase in computing power will also make such modeling more readily available to field commanders.
Even a modest capability to modify weather on a local scale, such as lifting fog or initiating precipitation, could have important military consequences. The Technology Group found no particular progress in this area. It concluded that the status of research on weather modification does not merit Army investment at this time, although the Army should continue to monitor this field in case a breakthrough occurs.