|
Multi-Perspective Decision Making in a Networked Environment
Navy STTR FY2011A - Topic N11A-T031 ONR - Mr. Steve Sullivan - [email protected] Opens: February 28, 2011 - Closes: March 30, 2011 6:00am EST N11A-T031 TITLE: Multi-Perspective Decision Making in a Networked Environment TECHNOLOGY AREAS: Information Systems, Human Systems ACQUISITION PROGRAM: Virginia Block IV and Ohio Replacement Program OBJECTIVE: Develop a decision aid (selected display and algorithm products) to dramatically enhance submarine decision making by allowing rich multi-perspective interaction between local control room operations and alternative operational command centers. DESCRIPTION: Across warfare and mission areas, and between and across echelons of command, the rate at which information is presented to decision-makers continues to outpace the rate at which Command and Control (C2) systems and human decision-makers have been able to effectively and prioritize, filter, and assimilate the information. In many cases the decision-maker is presented with too much information, while in other cases there is too little critical information. In many of these cases, the information is ambiguous, contradictory, and or contains uncertainties which are not well understood by the decision-maker. Consequently, the commander is increasingly challenged to make good decisions within critical operational timelines given the massive amount of information provided to him, and the increasing difficulty to determine the relevancy and uncertainty of that information. This mismatch between increasing amounts of information � and the difficulty in determining relevancy and uncertainty � and the need to make decisions within shortening timelines available also hampers the Navy�s desire to migrate to a C2 environment which is flexible and rapidly adaptable to unanticipated changes in C2 organizational structures and missions. Until the root causes for this mismatch are addressed, the decision-maker�s ability to meet critical operational decision timelines will deteriorate as new sensors and systems (such as multiple unmanned vehicles) and information sources are introduced into the operational environment, and manning levels are decreased. Submarine mission scope and complexity continues to grow. Although always a platform of choice for stealthy operations, since the end of the Cold War submarines have been increasingly operated in the littorals. There, submarines are expected to support asymmetric warfare operations and to become full-spectrum participants in net-centric warfare. Reflecting these trends, the submarine fleet is now asked to take on the challenge of incorporating unmanned aerial vehicles (UAVs) as a sensor in support of the Over-the-Horizon Targeting (OTH-T) missions, for strike and littoral surface warfare. This includes both deploying the UAVs while submerged, and ingesting their data to augment the tactical picture to plan and execute OTH-T missions. Robust network investment and advances in computing capabilities combined with the planned development of secure communications at speed and depth for submarines yield the potential to remotely perform (both on- and off-board) command functions currently located within the submarine�s control room. In preparation for this communication bandwidth capability, several technical challenges remain before a decision aid can be used. Specifically, integration of data and information from disparate modalities, mismatch between the temporal rates at which sensors produce the data, management and quantification of uncertainty inherited from all sensors and sources during processing, identification of relevant information and critical decision time-lines, integration of spatially disparate and heterogeneous information, and presentation of information in a manner that human decision-makers to choose from among several alternative COAs. Technological advances suggest ways of improving command effectiveness by escaping the "work station paradigm" (WSP) (Kranz et al., 2010). This traditional paradigm focuses information on single displays and passing information "up the chain" to a single decision-maker. Recent work in the area of teaming computer agents with human agents (Maarten et al., 2003) has developed a human-centered approach to human-agent interaction such as would be required by earth-bound controllers interacting with a remotely deployed vehicle, whether a UAV or Mars-bound explorer. Though this C2 enhancement has the potential to greatly enhance information gathering, analysis, and assimilation capabilities of enhancing team performance (Salas et al., 2007; Kozlowski & Ilgen, 2006) a comprehensive analysis for deployment of those capabilities that minimizes risk and enables comprehensive evaluation of tiered alternatives is needed, i.e., assuming future operational context for the UAV capability is fully networked, so elements of the UAV OTH-T process should be capable of modular offloading to networked elements. The output of this effort would provide a decision-maker multiple automatically generated valid courses-of-action (COAs) for a given mission that are tailored to the cognitive capabilities of the commander or command team. A robust methodology for defining mission-focused, decision-driven cognitive C2 information architectures based on local (control room-based) human control of some functions and decisions but remote participation in some set of operational decisions. Research is required for how decisions are to be divided among the operational (both on- and off-board) teams, how this decision structure changes with operational and environmental conditions, and what roles (both legacy and new) are required to execute in this environment. It is anticipated that scientific inquiry into the nature of decision making, optimal structures and protocols for collective and participatory decision making, and cognitive engagement depending upon decision making responsibility will be required. Especially important will be the avoidance or mitigation of potential "automation surprises" triggered when own ship operators are presented with status changes caused by automated alerts that are generated outside human involvement. The challenge is not only to develop a decision aid for understanding the information architectures that would lead to shared decision making and problem solving, but also to develop a system of mechanisms to allow dispersed teams of individuals to collaborate on operational decisions in ways that are both adaptive and resilient (Wreathall, 2006). Assuming a robust secure communication network is available at speed and depth, the final product may be selected display and algorithm products that adhere and implement elements of the cognitive information architecture, and which measurably demonstrate increased effectiveness in presenting the decision-maker with viable COAs in the time available. Achieving this programmatic vision for submarine decision making will identify (1) which decisions are best left to local command, and under what circumstances, (2) which decisions are best assigned to the off-board team because of analytical perspective or data access, and (3) a structure for assessing these decisions in a dynamic environment. Submarine force would implement this decision aid (selected display and algorithm products) in the command and control center to enhance C2 for on- and off-board communications. PHASE I: Provide a theoretical structure and work model, managing higher levels of risk in a controlled and deliberate manner, for developing operational decision making among local and remote team members. The concept should include strategies for deconstructing decisions, involving teams, and making criteria explicit. Develop a detailed method for evaluating the theory that includes both qualitative and quantitative measures. Enumerate a set of requirements for a proof of concept to be implemented in Phase II. PHASE II: Design, develop, and demonstrate decision aid prototype for the target application environment as investigated in Phase I. The implementation will test the predictions of the theory through measurements of the evaluation criteria developed in Phase I. Phase II will demonstrate the implementation in applications that illustrate the positive and negative consequences of multi-modal, multi-perspective decision making. A useful example would be to demonstrate critical decision making by a submarine command team in a non-routine situation using the proposed structure. Iterative research into cognitive models and workload will be required to assess the impact of the proposed structure. The Phase II effort may or may not be classified due to the content of the decision aid. PHASE III: This topic has many dual use applications, including the development of improved decision making in commercial as well as military environments. Additionally, there are potential total ownership cost implications of reduced recruiting and retention costs as a result of higher levels of employee engagement and control. Critical solutions would benefit industry, the government, and the military, and indeed, all large organizations that manage difficult, complex problems. REFERENCES: 2. Norman, D. (1998). The Design of Everyday Things. The MIT Press. 3. Maarten S., Bradshaw, J.M., Acquisti, A., Hoof, R., Jeffers, R., Uszok, A. (2003). Human-Agent Teamwork and Adjustable Autonomy in Practice. Proceedings of The 7th International Symposium on Artificial Intelligence, Robotics and Automation in Space (i-SAIRAS), Nara, Japan. 4. Salas, E., Rosen, M.A., Burke, C.S., Nicholson, D., and Howse, W.R. (2007). Markers for enhancing team cognition in complex environments: The power of team performance diagnosis. Aviation, Space, and Environmental Medicine, 78, 77-85. 5. Kozlowski S.W.J., and Ilgen, D.R. (2006). Enhancing the effectiveness of work groups and teams. Psychological Science in the Public Interest, 7, 77-124. 6. Wreathall, J. (2006). Properties of Resilient Organizations: An Initial View. In "Resilience Engineering�Concepts and Precepts" (E. Hollnagel, D.D. Woods, and N. G. Leveson, Eds.) Ashgate Press. KEYWORDS: decision modeling, cognitive models, operational command center Questions may also be submitted through DoD SBIR/STTR SITIS website.
|