Dynamical Systems and Control
Quantum and Non-Equilibrium Processes
Information, Decision and Complex Networks
Complex Material and Devices
Energy, Power and Propulsion
Basic Research Initiatives
Information, Decision and Complex Networks (RTC)
The Information, Decision, and Complex Networks Department is designed to support many U.S. Air Force priority areas including space situational awareness, autonomy, and cyber. The programs in the Department include Complex Networks, Computational and Machine Intelligence, Dynamic Data Driven Application Systems, Foundations of Information Systems, Information Operations and Security, Mathematical and Computational Cognition, Science of Information Computation and Fusion, Sensing Surveillance and Navigation, Systems and Software, and Trust and Influence.
Information: Critical challenges for the U.S. Air Force moving forward lie at the intersection of the ability to collect, mathematically analyze, and disseminate large quantities of information in a time critical fashion with assurances of operation and security from the infrastructure to the mission levels of the systems. The Science of Information, Computation and Fusion Program enables the ability to collect, disseminate and integrate information in such a way as to mathematically characterize and assess the most appropriate information for a range of mission critical tasks. The Sensing Surveillance and Navigation Program develop algorithms to collect and decompose critical sensing information and enables techniques that interface between the physical domains such as Electromagnetics and methods in navigation and geo-location. The Dynamic Data Driven Applications Program enables analysis of the interplay between physical systems such as fluid dynamical systems and software systems and architectures as in the case of aircraft flight systems. The Information Operations and Security Program looks at fundamental issues for assessing systems in terms of secure operations and mission assurance and the Systems and Software Program assesses these systems from a verification and validation standpoint to guarantee operations under a variety resource constraints.
Decision Making: This thrust focuses on the discovery of mathematical laws, foundational scientific principles, and new, reliable and robust algorithms, which underlie intelligent, mixed human-machine decision-making to achieve accurate real-time projection of expertise and knowledge into and out of the battle space. The Computational and Machine Intelligence as well as the Mathematical and Computational Cognition Programs focus on machine and human cognition and learning. The objective is to maximize the ability of machines to conduct higher level cognitive activities with quantifiable risk and accurate models of human decision makers. The Trust and Influence Program seeks to model and measure the way collections of individuals make decisions and are influenced both in small groups and culturally.
Complex Networks: Complex Networks consists of the Complex Networks Program and the Foundations of Information Systems Programs. Complex Networks is designed to mathematically represent networks of all kinds including communications and computation at all levels including content, protocol, and architecture. This mathematically unified representation is meant to measure, represent, resource, and secure critical infrastructures for U.S. Air Force and Department of Defense (DOD) applications. Additionally, the Foundations of Information Program is designed to use measurements and representations from the Complex Networks Program to verify and validate critical infrastructure performance.
Program Description: Network behavior is influenced at many levels by fundamental theories of information exchange in the network protocols and policies developed. The Complex Networks program seeks to understand mathematically how such fundamental approaches to information exchange influence overall network performance and behavior. From this analysis we wish to develop strategies to assess and influence the predictability and performance of heterogeneous types of U.S. Air Force networks that must provide reliable transfer of data in dynamic, hostile and high interference environments. Accordingly, we wish to develop approaches to describe information content, protocol, policy, structure, and dynamic behavior of a network by mathematically connecting observed network data to analytic and geometric representation. We can then exploit such mathematical tools in the formulation of network design and engineering approaches in areas such as information and communication theory, signal processing, optimization, and control theory. Examples of such tools might include methods derived from algebraic geometry, algebraic statistics, spectral graph theory, sparse approximation theory, random matrix theory, algebraic graph theory, random field theory, nonparametric estimation theory, algebraic topology, differential geometry, and dynamical systems theory, and quantum information theory. Advances in these mathematical methods will then enable specific ways to model, characterize, design, and manage U.S. Air Force networks and capture and predict the performance of these networks under many diverse conditions.
Basic Research Objectives: Thus methods of consideration in network modeling might include characterizing overall network performance by finding geometric descriptions of embedded parameters of network performance, specific analytic expressions for network behavior derived from inverse methods on network data, and divergence analysis of parameters characterizing one state of a network from another. Characterization of network behavior might include methods classify network behavior and structure through multi-scale vector space and convexity analysis, inference and estimation of networks through algebraic, graph theoretic, and Markov random field descriptions, and understanding of the robustness of given norms and metrics in representing network behavior. Design of networks might involve understanding the efficiency, scaling behavior, and robustness of methods of information exchange including those that use both self and mutual information paradigms. Management of networks may involve assessment of stability and convergence of network protocol and policy for various network dynamical conditions with such properties as curvature, homology class, or geometric flow. Approaches should have specific applicability to U.S. Air Force networking, communications, and architectural design problems but may be drawn from techniques in network analysis from a broad set of disciplines including quantum information systems, materials science and statistical mechanics, molecular and systems biology, wave propagation physics, decision, economics, and game theory to name just a few. From this we can conceive of new directions toward a science of networked systems.
Dr. Robert Bonneau AFOSR/RTC (703) 696-9545
DSN 426-9545; FAX (703) 696-7360
back to top
Computational and Machine Intelligence
Program Description: This program supports basic research in computational intelligence. This program supports innovative basic research on fundamental principles and methodologies of computational intelligence necessary to create robust intelligent autonomous systems. Robustness is defined as the ability to achieve high performance given at least some or all of the following factors: uncertainty, incompleteness or errors in knowledge; limitations on sensing; real-world complexity and dynamic change; adversarial factors; unexpected events including system faults; and out-of-scope requirements on system behavior. The vision of this program is that future computational intelligence systems will achieve high performance, adaptation, flexibility, self-repair, and other forms of intelligent behavior in the complex, uncertain, adversarial, and highly dynamic environments faced by the U.S. Air Force.
Basic Research Objectives: The program encourages research on building computational intelligence systems that derive from and/or integrate cognitive and biological models of human and animal intelligence. The investigative methodology may be theoretical, computational, or experimental, or a combination of thereof. Proposals to advances in the basic principles of machine intelligence for memory, reasoning, learning, action, and communication are desired insofar as these contribute directly towards robustness as defined above. Research proposals on computational reasoning methodologies of any type and combination, including algorithmic, heuristic, or evolutionary, are encouraged as long as the proof of success is the ability to act autonomously or in concert with human teammates to achieve robustness as defined above. Computational intelligence systems often act as human intelligence amplifiers in such areas as planning, sensing, situation assessment and projection; will monitor, diagnose, and control aircraft or spacecraft; and will directly interact with humans and the physical world through robotic devices. Therefore, research that that enables mixed-initiative interaction and teaming between autonomous systems and human individuals or teams is an important part of the program. Basic research that bridges the conceptual gaps between state-of-the-art statistical/machine learning algorithms and human cognition and performance are also welcomed. The program encourages multidisciplinary research teams, international collaborations, and multi-agency partnerships. This program is aggressive, accepts risk, and seeks to be a pathfinder for U.S. Air Force research in this area. Proposals that may lead to breakthroughs or highly disruptive results are especially encouraged.
AFOSR/RTC (703) (703) 696-9545
DSN 426-9545; FAX (703) 696-7360
back to top
Dynamic Data Driven Applications Systems (DDDAS)
Program Description: The DDDAS concept entails the ability to dynamically incorporate additional data into an executing application, and in reverse, the ability of an application to dynamically steer the measurement (instrumentation and control) components of the application system. DDDAS is a key concept for improving modeling of systems under dynamic conditions, more effective management of instrumentation systems, and is a key concept in architecting and controlling dynamic and heterogeneous resources, including, sensor networks, networks of embedded controllers, and other networked resources. DDDAS transformative advances in computational modeling of applications and in instrumentation and control systems (and in particular those that represent dynamic systems) require multidisciplinary research, and specifically need synergistic and systematic collaborations between applications domain researchers with researchers in mathematics and statistics, researchers computer sciences, and researchers involved in the design/ implementation of measurement and control systems (instruments, and instrumentation methods, and other sensors and embedded controllers).
Basic Research Objectives: Individual and multidisciplinary research, technology development, and cyberInfrastructure software frameworks needed for DDDAS applications and their environments are sought, along four key science and technology frontiers: Applications modeling: In DDDAS an application/simulation must be able to accept data at execution time and be dynamically steered by such dynamic data inputs. This requires research advances in application models that: describe the application system at different levels of detail and modalities; are able to dynamically invoke appropriate models as needed by the dynamically injected data into the application; and include interfaces of applications to measurements and other data systems. DDDAS will, for example, engender an integration of large scale simulation with traditional controls systems methods, thus provide an impetus of new directions to traditional controls methods. Advances in Mathematical and Statistical Algorithms include creating algorithms with stable and robust convergence properties under perturbations induced by dynamic data inputs: algorithmic stability under dynamic data injection/streaming; algorithmic tolerance to data perturbations; multiple scales and model reduction; enhanced asynchronous algorithms with stable convergence properties; multimodal, multiscale modeling and uncertainty quantification, and in cases where the multiple scales or modalities are invoked dynamically and there is need for fast methods of uncertainty quantification and uncertainty propagation across dynamically invoked models. Such aspects push to new levels of challenges the traditional computational math approaches. Application Measurement Systems and Methods include improvements and innovations in instrumentation platforms, and improvements in the means and methods for collecting data, focusing in a region of relevant measurements, controlling sampling rates, multiplexing, multisource information fusion, and determining the architecture of heterogeneous and distributed sensor networks and/or networks of embedded controllers. The advances here will create new instrumentation and control capabilities. Advances in Systems Software runtime support and infrastructures to support the execution of applications whose computational systems resource requirements are dynamically dependent on dynamic data inputs, and include: dynamic selection at runtime of application components embodying algorithms suitable for the kinds of solution approaches depending on the streamed data, and depending on the underlying resources, dynamic workflow driven systems, coupling domain specific workflow for interoperation with computational software, general execution workflow, software engineering techniques. The systems software environments required are those that can support execution in dynamically integrated platforms ranging from the high-end to the real-time data acquisition and control - cross-systems integrated. Software Infrastructures and other systems software (OS, data-management systems and other middleware) services to address the "real time" coupling of data and computations across a wide area heterogeneous dynamic resources and associated adaptations while ensuring application correctness and consistency, and satisfying time and policy constraints. Specific features include the ability to process large volume, high rate data from different sources including sensor systems, archives, other computations, instruments, etc. ; interfaces to physical devices (including sensor systems and actuators), and dynamic data management requirements.
Areas of interest to the AF, which can benefit from DDDAS transformative advances, include: areas driven by the AF Technology Horizons report, including: autonomous systems (e.g. swarms of unmanned or remotely piloted vehicles) ; autonomous mission planning; complex adaptive systems with resilient autonomy; collaborative/cooperative control; autonomous reasoning and learning; sensor-based processing; ad-hoc, agile networks; multi-scale simulation technologies and coupled multi-physics simulations; decision support systems with the accuracy of full scale models (e.g. high-performance aircraft health monitoring, materials stresses and degradation); embedded diagnostics and V&V for complex adaptive systems; automated software generation; cognitive modeling; cognitive performance augmentation; human-machine interfaces. Other examples include: Advanced electromagnetic sources with extremely high power densities, leading to a variety of phenomena such as plasma formation, require holistic approaches to very large data sets (>1TB) and incorporate nonlinear, multi-scale, and multi-physics effects in a common framework. DDDAS provides new approaches for combining computational, theoretical, and experimental data sets for high interactive testing of multiple physical hypotheses at once.
Programmatic activities that will be launched under this initiative will support research in individual areas, but mostly in the context of multidisciplinary research across two of the four components above, or at least two of the four.
Dr. Frederica Darema, AFOSR/RTC (703) 588-1926
DSN 426-7551; FAX (703) 696-8450
back to top
Foundations of Information Systems
Program Description: The Foundations of Information Systems program intends to foster fundamental research on new methods for analysis, management, and design of complex information systems. Traditional approaches to systems methods involve verification or equivalence model checking paradigms for software and hardware components, and are limited to analysis of individual-components in the design cycle. We seek to enable comprehensive system-level analysis, optimized performance, function, behavior, operation, fault-tolerance, robustness, adaptability and cyber-security among other properties. These approaches should be considered throughout the design, operation, and expansion of the system. Foundations of Information Systems seek to characterize the analysis of systems in multi-scale representations of sub-components and system-layers, derived from specifications, models and measurements. Because of the heterogeneous and dynamic nature of information systems, increasing emphasis on measurement-based performance analysis is necessary to develop the capabilities sought here. Therefore, we seek methods that allow integration between specifications-based methods and measurement-based methods which involve statistical analysis and dynamical systems theory to estimate the current true state and performance of the system as a whole. Such new methods should enable quantifiable, performance-driven systems-engineering and more powerful analysis capabilities for managing the design, operation, and scalability of systems that need to be adaptive and interoperable.
Basic Research Objectives: Fundamental strategies that integrate specification or model based methods with measurement based, statistics, risk, and dynamical system methods into a unified framework are thus of great interest. Of particular interest are multidisciplinary research efforts creating new approaches and methods that bridge across analytic, agent-based, graph-based, event-driven, and statistical Bayesian approaches, with techniques utilizing methods from model equivalency checking. Techniques in verification drawn from probabilistic process algebras, model checking, categorical logic theory, and algebraic representation theory are of interest as are methods in sparse approximation, parametric and nonparametric estimation, functional analysis, and geometric inference for system measurement and identification. Also of interest are entropy-based systems metrics, mean-field-theory, information-flow analysis and nonlinear optimization for risk assessment; operator and sheaf theoretic methods, computational homology, rigidity theory, and algebraic and category theoretic methods for invariant systems analysis. Any such theoretical approaches should be linked to compatible strategies which can involve techniques from systems analysis at multiple levels of abstraction, software and hardware modeling languages, software and system interfaces that improve component integration, and new methods for instrumentation and measurement. Application areas of interest, but not limited to: distributed, autonomous, and heterogeneous systems, distributed computational and cloud computing systems, information security applications, and efforts in dynamic resource management. Other related systems examples could be drawn from such diverse areas as quantum, biological, or sociological systems. These application areas should have relevancy to current U.S. Air Force needs.
Dr. Robert Bonneau AFOSR/RTC (703) 696-9545
DSN 426-9545; FAX (703) 696-7360
back to top
Information Operations and Security
Program Description: The goal of this program is to provide the science foundation to enable development of advanced cyber security methods, models, and algorithms to support future U.S. Air Force systems. Research is sought to meet the Information Operations challenges of Computer Network Defense Computer Network Attack and the management of the cyber security enterprise.
Basic Research Objectives: The development of a Science of Cyber Security is the holy grail of this program. The development of the mathematical foundations of system software, hardware, human users and attackers, and network architectures with respect to cyber security (implemented in policy), including key metrics, abstractions, and analytical tools is a critical issue. Security policy research is of high interest to this program. Formal modeling and understanding of the human users and attackers in these systems is of high interest. Developing the theory and methods to operate securely on distributed and cloud systems and systems that may not be secure is of high interest. New approaches for cyber forensics, active response, moving target, fight through and recovery related to cyber-attack is of high interest. Attack attribution is of particular interest. Basic research that predicts and anticipates the nature of future information system attacks is of high interest. Research that leads to methods to discover malicious code already imbedded in software or hardware is a high priority. The theory and modeling of covert and side channels is of high interest.
Dr. Robert L. Herklotz AFOSR/RTC (703) 696-6565
DSN 426-6565; FAX 703 696-7360
back to top
Mathematical and Computational Cognition
Program Description: This program supports innovative basic research on high-order cognitive processes that are responsible for human performance in complex problem solving and decision making tasks. The overall objective is to understand these processes by developing and empirically testing mathematical, statistical or computational models of human attention, memory, categorization, reasoning, problem solving, learning and motivation, and decision making. We are especially interested in the development and evaluation of formal cognitive models that provide an integrative and cumulative account of scientific progress, are truly predictive, as opposed to postdictive, and finally, are generalizable beyond controlled laboratory tasks to information-rich and dynamic real-world tasks
Basic Research Objectives: Research to elucidate core computational algorithms such as those that pertain to understanding of the mind and brain, often posed as finding solutions to well-formulated optimization or statistical estimation problems, has proven to be particularly valuable in providing a benchmark against which human cognitive performance can be measured. Selected examples of such algorithms include (the list is not exhaustive): (1) reinforcement- and machine-learning algorithms for planning and control in sequential decision making, where short and long term goals of an action are optimally balanced; (2) sequential sampling algorithms for trading between speed and accuracy in decision-making under time pressure, where optimal stopping rules take into consideration payoff for a prompt but inaccurate decision and cost for delaying it; (3) classification algorithms from supervised or semi-supervised learning, where optimal generalization from examples during categorization learning is achieved through regularizing the complexity of data-fitting models; (4) hierarchical or nonparametric Bayesian algorithms for reasoning, causal inference and prediction, where prior knowledge and data/evidence are optimally combined; (5) active learning algorithms for adaptive information sampling.
In relating formal models to human cognition and performance, research projects should not only ascertain their descriptive validity but also their predictive validity. To this end, the program welcomes the work that (1) creates a statistical and machine learning framework that semi-autonomously integrates model development, evaluation, selection, and revision; (2) bridges the gap between the fields of cognitive modeling and artificial general intelligence by simultaneously emphasizing important improvements to functionality and also explanatory evaluation against specific empirical results. The program also encourages the development and application of novel and innovative mathematical and neurocomputational approaches to tackle the fundamental mechanisms of the brain, that is, how cognitive behavior emerges from the complex interactions of individual neurobiological systems and neuronal circuits. Cross-disciplinary teams with cognitive scientists in collaboration with mathematicians, statisticians, computer scientists and engineers, operation and management science researchers, information scientists, econometricians and game theoreticians, etc. , are encouraged, especially when the research pertains to common issues and when collaboration is likely to generate bidirectional benefits.
AFOSR/RTC (703) 696-9545
DSN 426-9545; FAX (703) 696-7360
back to top
Robust Decision Making in Human-System Interface
Program Description: The need for mixed human-machine decision making appears at all levels of U.S. Air Force operations and pervades every stage of U.S. Air Force missions. However, new theoretical and empirical guidance is needed to prescribe maximally effective mixtures of human and machine decision making in environments that are becoming increasingly complex and demanding as a result of the high uncertainty, complexity, time urgency, and rapidly changing nature of military missions. Massive amounts of relevant data are now available from powerful sensing systems to inform these decisions; however, the task of quickly extracting knowledge to guide human actions from an overwhelming flow of information is daunting. Basic research is needed to produce cognitive systems that are capable of communicating with humans in a natural manner that builds trust, are proficient at condensing intensive streams of sensory data into useful conceptual information in an efficient, real-time manner, and are competent at making rapid, adaptive, and robust prescriptions for prediction, inference, decision, and planning. New computational and mathematical principles of cognition are needed to form a symbiosis between human and machine systems, which coordinates and allocates responsibility between these entities in an optimal collaborative manner, achieving comprehensive situation awareness and anticipatory command and control.
Basic Research Objectives: In the area of a) data collection, processing, and exploitation technologies, there is a need for (a.1) attention systems for optimally allocating sensor resources depending on current state of knowledge, (a.2) reasoning systems for fusing information and building actionable knowledge out of raw sensory data, (a.3) inference systems for real time accumulation of evidence from conflicting sources of information for recognition and identification. In the area of b) command and control technologies, there is a need for (b.1) prediction systems for anticipating future behavior of adversarial agents based on past experience and current conditions, (b.2) rapid decision systems with flexible mixtures of man and machine responsibilities for reactive decision making under high time pressure, (b.3) robust strategic planning systems designed to allow for sudden changes in mission objectives, unexpected changes in environment, and possible irrational actions by adversaries. In the area of c) situation awareness technologies, there is a need for a human-system interface that (c.1) faithfully simulates the content of a human operator's working memory buffer and its update thus modeling the operator's dynamic awareness of inputs, constraints, goals, and problems, (c.2) optimizes information delivery, routing, refreshing, retrieval, and clearance to/from the human operator's awareness while utilizing the latter's long- term store for expert knowledge, memory and skills for robust decision making, (c.3) achieves symbiosis between human and machine systems in delegating and coordinating responsibilities for command and control decisions. In sum, new empirical and theoretical research is needed that provides a deeper understanding of the cognitive requirements for command and control by a decision maker with enhanced capability for situation awareness, allows for greater degree of uncertainty in terms of reasoning systems, produces greater robustness and adaptability in planning algorithms in dealing with unexpected interruptions and rapidly changing objectives, generates greater flexibility in terms of assumptions about adversarial agents, and gives clearer guidance for dealing with the complexities encountered in network-centric decision tasks.
AFOSR/RTC (703) 696-9545
DSN 426-9545; FAX (703) 696-7360
back to top
Science of Information, Computation and Fusion
Program Description: The U.S. Air Force collects vast amounts of data through various modes at various times in order to extract and derive needed "information" from these large and heterogeneous (mixed types) data sets. Some data, such as those collected from magnetometers, register limited information content which is more identifiable at the sensor level but beyond human's sensory reception. Other types of data, such as video cameras or text reports, possess more semantic information that is closer to human's cognition and understanding. Nevertheless, these are instances of disparate data which encapsulate different types of "information" pertained to, perhaps, the same event(s) captured by different modalities through sensing and collection.
In order to understand and interpret information contained in various data sources, it is necessary to extract relevant pieces of information from these datasets and to make inferences based on prior knowledge. The discovery of relevant pieces of information is primarily a data-driven process that is correlational in nature and, hence, offers point solutions. This bottom-up processing direction needs conceptually-driven reasoning to integrate or fuse the previously extracted snippets of information by leveraging domain knowledge. Furthermore, the top-down process can offer causal explanation or causal inference, generate new hypotheses, verify or test hypotheses in light of observed datasets. Between the data-driven and conceptually-driven ends, there may reside different levels of abstraction in which information is partially extracted and aggregated based on the nature of applications.
Basic Research Objectives: With the rationale and guiding principles outlined in the above paragraph, this program seeks fundamental research that potentially leads to scientific advancements in informatics and computation which can support processing and making sense of disparate information sources. After all, information processing can formally and fundamentally be described as computing and reasoning on various data structures. Successes in addressing the research sub-areas stated below would give the U.S. Air Force new capabilities to: (1) shift emphasis from sensing to information; (2) understand the underpinning of autonomy; (3) relieve human's cognitive overload in dealing with the data deluge problem; (4) enhance human-machine interface in information processing.
To accomplish the research objectives, this program focuses on, but is not limited to, new techniques in mathematics, computer science, statistics and logic which have potentials to: (1) cope with various disparate and complex data types; (2) construct expressive data structures for reasoning and computation; (3) bridge correlational with causal discovery; (4) determine solutions or obstructions to the local-to-global data-fusion problem; (5) mechanize reasoning and computing in the same computational environment; (6) yield provably efficient procedures to enable or facilitate data analytics; (7) deal with high-dimensional and massive datasets with provably guaranteed performance.
Dr. Tristan Nguyen, AFOSR/RTC (703) 696-7796
DSN 426-7796; FAX (703) 696-7360
back to top
Sensing, Surveillance and Navigation
Program Description: This research activity is concerned with the systematic analysis and interpretation of variable quantities that represent critical working knowledge and understanding of the changing Battlespace. "Signals Communication" is a sub-area referring to the conveyance of information physically through a channel. Surveillance images are of special importance in targeting, damage assessment and resource location. Signals are either naturally or deliberately transmitted, propagated as electromagnetic waves or other media, and recaptured at the receiving sensor. Modern radar, infrared, and electro-optical sensing systems produce large quantities of raw signaling that exhibit hidden correlations, are distorted by noise, but still retain features tied to their particular physical origin. Statistical research that treats spatial and temporal dependencies in such data is necessary to exploit its usable information.
Basic Research Objectives: An outstanding need in the treatment of signals is to develop resilient algorithms for data representation in fewer bits (compression), image reconstruction/enhancement, and spectral/frequency estimation in the presence of external corrupting factors. These factors can involve deliberate interference, noise, ground clutter, and multi-path effects. This AFOSR program searches for application of sophisticated mathematical methods, including time-frequency analysis and generalizations of the Fourier and wavelet transforms, that deal effectively with the degradation of signaling transmission across a channel. These methods hold promise in the detection and recognition of characteristic transient features, the synthesis of hard-to-intercept communications links, and the achievement of faithful compression and fast reconstruction for audio, video, and multi-spectral data. New combinations of known methods of asset location and navigation are being sought, based on analysis and high-performance computation that bring a force-multiplier effect to command/control capabilities. Continued upgrade and reliance on Global Positioning System makes it critical to achieve GPS-quality positioning in situations where GPS by itself is not sufficient. Ongoing research in Inertial and non-Inertial navigation methods (including optical flow and use of signals of opportunity) will bring location precision and reliability to a superlative level. Continuous improvement in its repertoire of signal processing and statistical tools will enable the U.S. Air Force to maintain its lead in Battlespace awareness through navigation and surveillance. Communications are what hold together the networked Infosphere and cost-effective systems innovations that enable phenomenal air power projection.
Dr. Tristan Nguyen, AFOSR/RTC (703) 696-7796
DSN 426-7796; FAX (703) 696-7360
back to top
Systems and Software
Program Description: The program is seeking proposals with ideas engendering transformational research in systems software to address the growing scale of software supporting U.S. Air Force systems and platforms and meet future U.S. Air Force needs in the air, space, and cyber domains. To this end, the program is seeking to foster new basic research that addresses the design, creation, and employment of software-intensive systems. Broad areas of interest include disciplinary and mostly multidisciplinary research on: 1) software methods to support distributed, heterogeneous platforms, as well as need for capabilities for autonomic systems, resilient autonomy, adaptive software systems, and verification for software systems; 2) new, multi-level and multi-modal approaches as well as representations, abstractions, and composition of models and measurements into comprehensive software frameworks to represent and manage the diverse interactions among the software, the systems on which the software resides, and the dynamic environments in which these systems operate, and in particular as such capabilities apply to support; 3) Human-in-the-loop interacting with, and supported by, such software systems, and autonomous reasoning and learning.
Basic Research Objectives: In the area of distributed computational platforms and their environments, transformative opportunities derived from exploiting the next generation of multicore-based systems and new paradigms of complex applications' computational approaches motivate fundamental research in programming environments, application development, and compiler/runtime support. Of particular interest are directions and efforts on new compiler-embedded-in-the-runtime approaches ("runtime-compiler") to support adaptive and optimized application mapping, runtime, and composition. Environments motivating us to address such research consist of distributed and heterogeneous computational platforms ranging from the high-end to small clusters, as well as emerging unified computational environments which dynamically integrate high-end systems with real-time data-acquisition and control systems (including those spanning emerging peta- and exascale-range platforms on the high-end side, and networks of heterogeneous sensors and networks of controllers, on the data-acquisition and control side, all of which will be multicore-based). These classes of platforms will exhibit multilevel heterogeneity in terms of their processor interconnects, memory-levels, and latencies. They will entail environments where the resources available to and required by the executing applications will vary during execution. New insights are also sought into the human's role and interactions with heterogeneous software: we seek new theories for modeling and developing systems that have human and machine components. It is important to consider integrated modeling approaches that jointly address the hardware, software, and human components of large-scale systems. Realization of these mixed-component (human-machine) systems may also require new approaches to how computation itself is modeled or even an entirely new understanding of computation.
Dr. Kathleen Kaplan AFOSR/RTC (703) 696-7312
DSN 426-7312; FAX (703) 696-7360
back to top
Trust and Influence
Program Description: The Trust and Influence program is aimed to develop a basic research portfolio that will provide the empirical foundation for the science of reliance and contemporary influence. This R&D portfolio specializes in basic research focused on: 1) Empirical science of trust in both interpersonal (i.e. , cross-cultural domains) and in complex human-machine/robot interactions, 2) Science of influence effects including the psychological and behavioral impact of novel technology on the battlefield (e.g. , new non-lethal weapons, robots in combat), 3) Understanding the cognitive and social avenues of influence based on cultural, social, or technological means. The resulting portfolio directly enhances the U.S. Air Force's impact on policies and operations related to national security by investing in the discovery of the foundational concepts of effective influence, deterrence, trust-building, trust calibration with technological systems, counter-terrorism and paths to violent radicalization. The AFOSR trust and influence R&D portfolio specifically invests in multi-disciplinary approaches ranging from psychology to computer science. Research designs that incorporate field research or laboratory research are encouraged to apply as are conceptual studies aimed at developing transformative novel theories.
Basic Research Objectives: This program encourages collaboration between psychologists, anthropologists, sociologists, linguists, behavioral, cognitive, and neuro-econometric scientists as well as computational researchers in disciplines such as computer science. Example topics include: (1) Empirical science to reveal the antecedents of trust in cross-cultural interactions (2) Empirical studies to examine the malleable elements of trust calibration during complex human-machine/human-robot interactions (3) Empirical studies to identify the cognitive mechanisms associated with persuasion and social influence in a digitized world --"socio-digital influence" (4) Conceptual, empirical, or modeling studies to examine the psychological/behavioral impact of new weapon technologies such as non-lethal weapons and robotic platforms.
Dr. Benjamin Knott, 711 HPW/RHCPA
back to top