Today’s warfighters operate in a highly dynamic world. There are competing mission and task demands, data overload, and uncertainty make it a significant challenge to find useful data and integrate it into decision-relevant information. The phrase “too much data—not enough information” is a common refrain amongst tactical warfighters, mission planners and commanders.
The key element that differentiates data from information for a decision maker is context. Context is often poorly represented or completely absent in many current military communication and command and control systems. This lack of context requires decision makers to make continuous, internal reconstructions of their environment and the situation, reassessing what they know and what they don’t know, as they also attempt to comprehend and act on complex, dynamic data. Decisions made as a result of these factors are often poor or inappropriate. We have an opportunity to cognitive theory and computational modelling to create decision support tools for warfighters, and this is focus of the Command Decision Making Program.
The Command Decision Making (CDM) program is researching decision making processes to develop tools that enable warfighters to better manage information and make better decisions, faster. CDM is funding projects to: 1) Create a science of Context-Driven Decision Making (CDDM) and 2) Develop Proactive Decision Support (PDS) tools. At the center of this program is the idea that by being contextually aware of decision makers’ missions and tasks, PDS tools will anticipate the decision and information needs of decision makers. Decision makers will then more effectively address operational complexity and uncertainty and thereby make better decisions, faster.
PDS exploits predictive models that leverage the advances being made within the fields of cognitive science, decision theory, artificial intelligence, basic and applied mathematics, to create novel decision support tools.
Specific areas of interest from these fields for CDM include human behavior modeling and inferencing, team decision making, knowledge creation and transfer, and processes through which mental models are created and modified.
CDM’s aim is to demonstrate a paradigm shifting, scalable process for dynamically generating candidate decision events based on mission and task context that complement decision making in operational areas such as command & control resource allocation, autonomy control, cyber security, and others.
Research Challenges and Opportunities
The major research interests include:
- Cognitive models of individual and distributed decision making and collaboration
- Examining the role of context in applied decision making
- The study of Decision Making in Uncertainty
- Developing Proactive Decision Support tools for military command and control
- Model-based experiments and field trials with emphasis on decision processes in scalable Maritime Operation Centers
There is a need to develop proactive decision support agents that can:
- Represent and reason about the status, capabilities, information needs, and preferences of the CIC staff;
- Interpret and react competently to mission changes and shifts in command information needs;
- Display information (as needed) in a comprehensive & comprehendible visual Common Operating Picture (COP);
- Identify CIC staff proposals for changing an existing COA;
- Accurately predict the unfolding performance and tradeoffs COAs in the current (and hypothetical) battlespace situation;
- Search for and rank alternatives to the current COA relative to mission objectives (constrained by commander’s intent) that are expected to in-crease mission Measures of Effectiveness (MOEs);
- Recognize when and how to present alternatives COAs to CIC staff; and, importantly, provide justifications for them upon request.
- The recognized challenges include the need to make sense of uncertain information and changing conditions within the present and future battlespace using incomplete and imperfect information with which to weigh competing courses of action (COAs).
In response to these challenges, we seek to develop intelligent decision-support tools and data analytic, machine learning, cognitive, and artificial intelligence (AI) based technologies that accelerate decision-making capability/tempo and enable commanders and staffs to make exceptionally high quality decisions as fast as possible.