| 2005 Technology
Symposium > Decision Support
Decision Support
This area focuses on cognitive-centered decision support applications
and new methods and tools for developing effective systems that support
decision-making. Emphasis is placed on decision-making in dynamically
changing real-time environments (occurring in a day or less). Research
in human decision-making to enable the development of better support systems
for the military or other sponsors is covered in this area. Also covered
is the demonstration of decision aids that advance the state of the art.
2020 Vision for Future NAS Operations
Christopher DeSenti, Principal Investigator
Location(s): Washington
Problem
Many diverse activities related to future vision concepts are going on in the aviation community. Individual concepts are being proposed and explored, however no one has yet taken an in-depth look at system-level approaches and implications, and demonstrated the viability of a comprehensive, integrated set of concepts.
Objectives
The project seeks to understand the range of viable aviation futures and develop a common understanding of the concepts across the aviation community. We intend to describe the principles, initiatives, and integrated set of concepts that move us along a beneficial path considering the viable range of futures. We must convince stakeholders that an integrated set of future concepts is more valuable than the sum of its parts.
Activities
Work will focus on refining concepts to collaboratively identify issues and communicate ideas, and on scrutinizing underlying assumptions. Leveraging MITRE's modeling capabilities will provide an assessment of the concepts. The resulting refinements will feed the building of interactive prototypes in MITRE's air traffic management laboratory. These will be used to visualize and communicate the concepts and to help move toward consensus among diverse stakeholders.
Impact
The project is intended to help build consensus on the right set of aviation initiatives for the future. This work will support industry-wide efforts to define and understand future NAS concepts of operations, and help to determine appropriate areas of focus for research and transition planning for long-term NAS concepts of operation.
^TOP
Advanced Soldier Sensor Information System Technololgy
Lisa Harper, Principal Investigator
Location(s):
^TOP
Air Traffic Planning in an Uncertain World
Craig Wanke, Principal Investigator
Location(s): Washington
Problem
Traffic flow management (TFM) decisions are based on imperfect predictions of air traffic demand; however, prediction uncertainty is not explicitly factored into the decision-making process. In the current Enhanced Traffic Management System, demand predictions are presented as "truth," even though traffic managers are aware of the uncertainty around them. This can result in unnecessary actions and conservative decisions.
Objectives
We seek to explore ways to manage the impact of uncertain demand predictions on TFM decision making. Last year, we quantified the uncertainty present in the demand predictions used in current TFM operations, and developed visualization techniques for uncertain demand. This year, we are developing practical computational approaches to automation-assisted TFM problem solving in the presence of both uncertain demand and uncertain airspace capacity.
Activities
Our approach uses a reward function to measure outcomes, and develops solutions to optimize that function in the presence of uncertain input data. We will develop reward functions that capture the characteristics of TFM decisions and create algorithms to find good solutions with respect to the reward function. We are exploring ways to characterize airspace capacity and the uncertainties in predicting capacity when severe weather is present.
Impact
The results of this work will be useful in improving present-day TFM decision making, improving the utility of TFM decision support tools currently being developed in the MITRE FAA work program, and developing a comprehensive vision for future TFM decision support systems. The airspace capacity research also has implications for development of near-term TFM procedures and for airspace design.
^TOP
Analysis Support to Predictive Battlespace Awareness
Steve Frey, Principal Investigator
Location(s): Washington and Bedford
Problem
Operational modeling and analysis was performed in prior research on the subject of predictive battlespace awareness (PBA). Results indicated a wide range of development opportunities. The most critical need is for analytic tools to suppport tracking information and predicting future enemy actions. This is a complex type of analysis and not easily supportable by a single "tool" or system.
Objectives
This project seeks to establish a PBA analysis framework, determine the "ground truth" for predictive analysis capabilities through experimentation, and contribute to furthering the state of the art for the various components of the framework. Results from experimentation will be compiled into recommendations on "way ahead" planning and investment strategies for achieving PBA.
Activities
First, we will identify the activities and data sets applicable to predictive analysis as depicted in the PBA operational process and executable models. Next, we will identify capabilities and components that would support those operations together in a technical framework. Finally, we would investigate and experiment with applicable COTS, GOTS, and R&D efforts and evaluate them against this framework.
Impact
The research will result in a structured approach to developing predictive analysis systems and an investment strategy. This activity is relevant to the USAF's ongoing PBA definition. In addition, we intend to have a continuing experimental facility and knowledge base to support further research and experimentation opportunities in the areas of integrated intelligence and operations.
^TOP
Applying User Models to Improve Team Decision Making
Brad Goodman, Principal Investigator
Location(s): Washington and Bedford
Problem
People who do not always work together must sometimes meet to solve important and possibly time-critical problems. MITREs sponsors are rapidly accepting computer-supported collaborative work environments as a medium to bring people together to discuss such problems. The participants share information, insight, and advice. The team decision-making structure afforded by collaborative environments, however, does not necessarily encourage productive participation.
Objectives
This project will develop models of users of collaborative environments to promote common situation awareness and understanding that can lead to successful collaborative decision making. The user models help monitor the quality of group interaction in pursuit of a solution to a mutual problem. These user models move with users and persist from session to session and group to group.
Activities
We will develop an intelligent steward agent to guide team members through collaborative decision making. We will analyze meeting corpora from the research community to determine characteristics of decision-making dialogues, and will conduct controlled studies to evaluate collaborative activities and highlight collaborative behavior. We will track topics and stages of decision making and formulate user models as meetings proceed.
Impact
Many MITRE sponsors use collaborative technology to bring experts together to make decisions. The development of criteria to help in the selection of team members and of techniques to guide teams towards consensus can lead to more cohesive teams, improved decisions, and enhanced knowledge sharing. The proposed research will provide a more effective foundation for sharing expertise in collaborative environments.
^TOP
Evolution of a Combined Missile Defense/Missile Warning Picture
Alberto Andrade, Principal Investigator
Location(s): Washington
Problem
Missile defense (MD) and missile warning (MW) mission decision makers must manually assimilate information from multiple systems, with different data sources, in order to construct an accurate missile event picture. Developing a common missile event picture is complicated by the differing perspectives of the two missions, and requires integration at the communications, sensor, mission algorithm, and display levels.
Objectives
We will assess the impact of a common missile event picture on mission decision making. We will explore the near-term migration to a combined picture by integrating existing MD/MW information using correlation and by integrating additional information into the picture via net-centric techniques. We will identify long-term changes required to bring clarity and accuracy to the missile event picture.
Activities
We will reuse MD and MW displays and prototype additional screens. We will integrate information from the independent missions by reusing and prototyping correlation algorithms, use additional data sources that enhance the picture, and specify constraints to avoid ambiguity. We will identify net-centric techniques for delivering this new data and survey the operational community to obtain input on our solutions.
Impact
A common missile event picture produces military utility by reducing ambiguity and improving coordination of offensive/defensive operations. Common displays are a starting point, but greater benefits are achieved when integration occurs at earlier stages. The research will demonstrate the benefit of a common MD/MW operational picture that is achieved through rapid, net-centric integration of external information sources.
^TOP
Explanation-Based Decision Making
Brant Cheikes, Principal Investigator
Location(s): Washington and Bedford
Problem
Research in human judgment and decision-making shows that novices and domain experts alike make a variety of predictable mistakes in forming assessments, estimates, and predictions. Failures in judgment can be enormously costly in both wealth and lives. Yet systematic methods to help minimize such failures or reduce their impact are few, and inconsistently taught or applied across the government.
Objectives
This project will evaluate a new framework for the design of cognitive support tools, based on the claim that explanations are central to how people reason and decide. Driving this research is a vision of the role and purpose of cognitive support tools for analytic judgment, namely, that such tools should enable analysts to create external representations of decision problems.
Activities
We will conduct controlled experiments and observational studies to evaluate explanation-based decision making (EBDM) as a guiding theory for cognitive support tools. Three themes will structure our research: (1) validating EBDM in sponsor-relevant domains; (2) graphical approaches to externalizing explanations; and (3) support for effective peer review of analytic work.
Impact
This research will deepen MITRE's understanding of the core cognitive processes that drive decision making, and enable us to articulate technical requirements for cognitive support tools for analysis and decision making. This work will position us to advise the government knowledgeably on current and future technology investments in this area.
^TOP
Improving Fire Support Coordination with Blue Force Tracking
Jim Dimarogonas, Principal Investigator
Location(s): Washington
^TOP
Improving Human/UAV Interaction for Intelligence, Surveillance, and Reconnaissance
Jill Drury, Principal Investigator
Location(s): Washington and Bedford
Problem
Due to the increasing importance of UAV operations and the ramp-up of the Distributed Common Ground System (DCGS), we urgently need to support efficient and flexible UAV operations. UAV operations are currently manpower intensive and confined to a single location. Moreover, they are often non-intuitive and plagued by "mishaps."
Objectives
Our research will determine ways to provide better situation awareness (SA) of both UAV team members' activities and the ISR environment and identify opportunities for generalizing our understanding of improved SA to other C2 domains. We will provide guidance for whether/how UAV teams could be distributed, and develop and evaluate interaction designs that support small teams operating multiple UAVs simultaneously.
Activities
Our technical approach combines observation, interviews, analysis, prototyping, and experimental techniques. We will study both live and simulated UAV operations with both large and small UAV platforms. Our analysis results will inform new interaction designs that will be validated via experimentation. We are partnering with Dr. Michael Goodrich of Brigham Young University, a recognized expert in human-UAV interaction.
Impact
This work will result in UAV interaction that better facilitates SA acquisition and team coordination, a smaller number of operators placed in harms way, input into the debate about distributed UAV operations, and fewer UAVs lost to mishaps. A broader impact could arise from insights into providing better SA for domains beyond UAV operations.
^TOP
Improving Time-Sensitive Team Decision-Making
Lindsley Boiney, Principal Investigator
Location(s): Washington and Bedford
Problem
Given the progress made on system interoperability and machine-to-machine data transmission, C2 operator teams can shift more of their attention from routine data manipulation to making judgment calls and critical decisions. Unfortunately, systems often incompletely support team information sharing, as evidenced in real-world operations by teams struggling to make decisions quickly, accurately, and confidently.
Objectives
Many time-sensitive decisions must be made collaboratively. We will apply cognitive techniques to better understand the team decision-making process. We will explore effective ways to augment and leverage human capabilities via information technologies and improved processes. Key goals are to reduce operators cognitive load, confusion, and stress, thereby improving collaboration, shared situation awareness, and the timeliness and accuracy of decisions.
Activities
We will observe and interview operator teams in several time-sensitive domains to identify key collaborative processes and challenges. We will develop a model of time-sensitive collaboration, generate recommendations for system and/or team process modifications, and validate the impact of proposed modifications on performance.
Impact
The need for collaboration, the amount of information, and the level of complexity are increasing. New systems abound, yet human decision-making is often the limiting factor in performance. Research exploring dynamic team decision-making is critical to improving C2. It can reduce not only timelines, but operator stress, cognitive load, confusion, and errors. It can enable more adaptive responses to unanticipated events.
^TOP
Integrating Judgment and Technical Data Analysis
Paul Lehner, Principal Investigator
Location(s): Washington
Problem
When applied in the real world, technical analysis and interpretation of sensor data involve considerable subjective judgment. Unfortunately, both research in psychology and experience tell us that expert judgments of this type are consistently subject to well-known biases and errors.
Objectives
This project will develop and test a structured analytic method, called Causal Judgment Analysis (CJA), to support judgment in technical data analysis.
Activities
CJA was initially developed to support counter deception analysis. On this project CJA will be adapted to problems in technical data analysis. We will test the method by examining its performance on problems where ground truth is known and by assessing its impact on analytic thought on difficult current analysis problems.
Impact
If successful, this project will provide a rigorous analytic method for stepping through difficult technical analysis problems. This will lead to more accurate interpretation of sensor returns, and to better understanding of the subjective judgments on which the interpretation depends.
^TOP
Lightweight Collaborative Whiteboard +
Peter Firey, Principal Investigator
Location(s): Washington
Problem
Distributed teams conduct planning, coordination, and operations via shared files (e.g., PowerPoint), disjoint "common" operational pictures, text chat, audio and VTCs. New approaches to C2 applications are needed to reduce time, increase precision and improve accuracy across a spectrum of visual and textual information which can be rapidly configured for unanticipated disparate information and retrospective as well as current perspectives.
Objectives
Develop collaborative geospatial common operational picture application (CGCOP) to support distributed decision support for rapid situation awareness (SA) battlefield geometry and intelligence and coordination and handoff across shift changes, new team members, and communities. Investigate and understand design patterns for SOA Web Service-based services that support role based access control with discretionary access controlled information using NCES ws-security approaches. Have a scalable solution that meets GCCS criteria for number of users and dynamic information flow.
Activities
Apply cognitive systems engineering and methodology to improve collaborative SA. Develop and investigate temporal services in SA through the web browser-based Common Operational Picture. Investigate NCES security approaches and the ability to realize fine grained object level access control within the architecture. Assess applicability tactical wireless. Develop approaches to realize scalable services.
Impact
This work will guide the next generation (network centric enterprise services) and transformational C2 architectures and design implementation patterns through technology insertion and transfer. And the work will provide DoD/IC new concepts for C2, decision support and analysis through collaboration enabled mission applications. It may improve interoperability for DOD collaboration by showing examples of loosely coupled collaboration services. Furthermore, it will influence commercial vendors to provide collaborative SVG/Web service enabled capabilities.
^TOP
Logical Expansion of Arrival and Departures to Enhance RNP (LEADER)
Satish Mohleji, Principal Investigator
Location(s): Washington
Problem
Flight Management System-equipped aircraft can navigate precisely from point to point within their Required Navigation Performance (RNP) limits, but controllers still use ad hoc path changes to separate such aircraft from unequipped ones. This makes it difficult to predict schedules. Airlines need route repeatability to improve schedule predictability, while controllers require procedures involving less workload to smoothly merge and separate aircraft with varying equipage.
Objectives
This research will address (1) how to develop terminal route designs for departures and arrivals to ensure that full airport capacity and efficiency benefits are achieved; (2) how to deal with future unscheduled air traffic demand; (3) how to improve schedule predictability at different stages of flight; and (4) how to conduct "what if" impact analyses for new terminal procedures.
Activities
This research will develop a human-in-the-loop simulation prototype capability to include: (1) design of RNP routes with appropriate altitude and speed requirements at specified waypoints based on a wide range of aircraft navigation capabilities; (2) development of planning algorithms for arrival/departure traffic flow; (3) demonstration of speed control processes to deal with flight deviations without moving aircraft off planned routes; and (4) demonstration of the benefits of automation planning.
Impact
This project will demonstrate a prototype terminal decision support system for current and future RNP-based flight planning and execution from gate to gate.
^TOP
Mental Models in Naturalistic Decision Making
Kevin Burns, Principal Investigator
Location(s): Washington and Bedford
Problem
System engineering efforts are often based on informal and inadequate models of how, psychologically speaking, people make decisions and what, computationally speaking, people need to make better decisions. MITRE and its sponsors need computational models of cognitive processes to advance the design of information systems in command and control applications.
Objectives
This research project will develop computational models of the cognitive processes that underlie human decision making in prototypical command and control tasks, such as "risk assessment," "resource management," and "rational engagement." The computational models will use a "bounded Bayesian" approach that treats human decision makers as rational (Bayesian) but limited (bounded) by natural constraints.
Activities
The research methods will include human experiments and agent simulations in a micro world that replicates cognitive challenges of the real world. The micro world is a card game called Poker TRACS, played with a special deck of two-sided cards to simulate the clue/truth (back/front) structure of incomplete information in practical problems of diagnosis and decision making.
Impact
The research products will be computational models of human decision making and conceptual systems for automating and improving human decision making. These models and systems will be validated and evaluated in the lab (micro world) to establish a credible basis for improving practical systems in the field (real world).
^TOP
MITRE Crisis Response System
Doug Phair, Principal Investigator
Location(s): Bedford
Problem
The MITRE Crisis Response System was designed to replace emergency contact calling trees with a multimodal employee contact architecture. We have implemented a prototype system that consists of a centralized notification framework using existing enterprise email, Web site, instant messaging, and voice over IP (VOIP) infrastructure. This system allows for the rapid, multimodal distribution of critical information and the real-time assimilation of specific responses. Response and audit data are collected and stored for analysis, and can be reviewed using a variety of methods in real time. Crisis responders can access multiple views of the contact data in real time, including management and organization reports. Beyond the prototype, we are exploring the issues and costs for scaling our experimental system to an enterprise.
^TOP
Modeling Complex Adaptive Behavior
Daniel Venese, Principal Investigator
Location(s): Washington
Problem
Border screening is a highly constrained problem with many unknowns regarding the threat and opposition tactics. There is a large volume of traffic with relatively few that are suspect out of the total population--that is, the needle in the haystack problem.
Objectives
Deploy prototype decision support application (DSA) for use by Customs and Border Patrol (CBP) Arizona Customs Management center. Lessen ability of smugglers to predict occurrence of special operations. Ensure operations are planned in conformance with management objectives', e.g., optimize counterdrug seizures, optimize counter terrorism screening.
Activities
Develop a DSA to assist in planning special operations using a knowledge base of past performance. The DSA will incorporate a behavioral model of smuggler decision making. A prototype application will be deployed for experimental use. The DSA will be evaluated during development through use of a simulation model of the land border developed in FY04.
Impact
Improve understanding of smuggler behavior for incorporation into other targeting systems and analysis tools. Improve acceptance and understanding of decision support technology in homeland security community.
^TOP
Nonlinear Visualization Techniques
David deMoulpied, Principal Investigator
Location(s): Washington and Bedford
Problem
Warfighters and other situationally oriented information consumers constantly face situations that require immediate identification of essential information within a large information space. Recent advances in hardware and network data access exacerbate this information overload problem. While data processing and data access continue to progress rapidly, real-time visualization techniques have not evolved to keep pace with this massive influx of data.
Objectives
This project will investigate nonlinear techniques for visualization of graphical data, with specific emphasis on semantic lensing strategies. Evaluation of, and experimentation with, new visualization techniques will facilitate understanding of their effectiveness in aiding decision making under high-stress and high-workload conditions.
Activities
Research into a selection of nonlinear visualization techniques and their contextual effectiveness will be used to define an experimentation plan. This experiment will enable evaluation of semantic lensing techniques and their effectiveness compared to traditional visualization techniques for a variety of domains, including track-based air defense, homeland security, combat support, and netted sensors.
Impact
This effort will advance nascent research in nonlinear visualization techniques for the military domain by determining whether using these techniques enables faster decision making by reducing the need to change foci in a high-stress and high-workload environment. We will determine whether patterns and relationships within data can more easily be seen through nonlinear visualization techniques, thereby enabling better informed decision-making.
^TOP
Perceptive Assistive Agents in Team Spaces
Lisa Harper, Principal Investigator
Location(s): Washington
Problem
The use of next generation interface technologies to facilitate human-human collaboration and human-computer interaction has tremendous potential for enhancing team interaction. However, new devices and technologies cannot be inserted successfully unless they are adapted to the dynamics of the group's structure and interactions. Furthermore, the insertion of any particular technology may have unanticipated and unintended consequences. We have designed an Experimental Team Room (ETR) that is a replica of MITRE's standard conferencing environment. However, this facility has been modified to observe and record user interactions with room devices and research products.
Objectives
We focus on the ability of assistive agents to monitor, access, and manipulate elements of the physical context. Our hypothesis is that we can enhance the interactions of teams and performance in both local and remote groups by using perceptive personal agents and team space agents that mediate interactions between humans and the collaboration environment.
Activities
We are leveraging the ETR to develop a sensor-based environment and a virtual human (Emma) that can access, monitor and change the physical environment. Users can communicate with Emma, the Electronic MITRE Meeting Assistant, in a variety of ways such as speech, gesture or via a graphical interface. We are endowing Emma with the ability to recognize individuals via speech and face recognition as well as the ability to track their location in space. We believe that a perceptive agent that is more aware of the physical context of a meeting will be more flexible and responsive to user needs than current meeting room interfaces.
Impact
By leveraging our own operational mission and corporate expertise, we have a tremendous opportunity to study real user populations over an extended period of time in such a way to inform usability issues critical to related C3I environments and to Section 508 of the Rehabilitation Act.
^TOP
Turbofraud: Counter-Deception Support for Forensic Accounting
Frank Stech, Principal Investigator
Location(s): Washington
Problem
Denial and deception aim to disrupt an adversary's ability to "observe, orient, and decide," and to induce inaccurate impressions about friendly capabilities or intentions, causing the adversary to apply intelligence collection assets inappropriately, or fail to employ capabilities to best advantage. Although the need for counter deception (CD) is recognized, proposed solutions make little or no use of the psychology of deception and decision making.
Objectives
We will develop a decision framework based on existing research on the psychology of deception and integrate the framework with belief modeling tools to create a CD decision support system for intelligence analysts. Our hypothesis is that the psychology of decision making and deception can be combined with existing belief management and planning technology to produce a CD decision support system.
Activities
In the Modeling phase we will construct a psychological framework of deception based on a deception taxonomy and deception cognitive model. In the Development phase we will develop tools for generating deception hypotheses and assessing the evidence of deceptions. The result will be a computational system that helps analysts to recognize potential deception moves, evaluate evidence, identify probable deceptions, and de-bias estimates. The Assessment phase will test the hypothesis through experiments with intelligence analysts.
Impact
Research in CD will position MITRE to assist in several intelligence community initiatives and to develop systems to address several of our sponsor's identified "hard problems." It will also augment the Information Operations Planning Tool Advanced Concept Technology Demonstration (ACTD) with deception planning aids.
^TOP
|
