Geo-Spatial Navagation under Uncertainty (GNU)
Abstract
Geo-spatial Navigation in Uncertainty (GNU) Abstract A common task in the Navy is estimating the future state of some variable. For example, when planning the route for a ship, we might predict the weather at a certain places and times, perhaps to choose a route that avoids a storm. Decision-makers are almost always faced with some degree of uncertainty; there is a great deal of research on making decisions under uncertainty. However, there are a few important aspects of decision-making under uncertainty for which research is sparse: First, while decision makers may be accustomed to coping with current uncertainty, relatively little effort has been devoted to development of forecast decision strategies that deal with environments in which the amount of uncertainty is dynamic. There is evidence to suggest that humans are not proficient predicting the future state of uncertainty; but are overconfident in their ability to do so. Second, our ability to reason about the consequences of uncertainty is influenced by the tools we use to visualize and communicate about uncertainty. What is the best way to visualize future uncertainty? In a review of literature about decision-making under uncertainty, we found very limited prior work at the intersection of the topics of prediction, uncertainty, and visualization. The goal of our research is to understand the challenges and sources of biases in the task of predicting the future in the face of dynamic future uncertainty. This is important because if people overestimate the confidence of their prediction, they may not plan for the alternatives in case their predictions turn out to be wrong. And, to the extent that people are overconfident, they may not seek important information that could prove their prediction to be inaccurate. In our research, we wish to examine five major issues: 1. How do people understand the rate of change of uncertainty over time? What factors influence their behaviors? 2. How do people use their understanding of changing uncertainty to make judgments? 3. How do people use their knowledge of future uncertainty to seek more information to further reduce uncertainty? 4. How do people balance their subjective level of uncertainty in the future state with their need to take a timely action? 5. How can visualization techniques and training procedures improve predictions of future uncertainty? Our research will benefit the Navy in tangible ways. Improved understanding of the consequences of dynamic uncertainty has the potential to improve ship routing. The economic benefits of improved ship routing include reduced fuel costs, extended service life for capital equipment, and extended maintenance cycles. Improved routing may also reduce risks related to crew safety.
Document Details
- Document Type
- DoD Grant Award
- Publication Date
- Aug 08, 2016
- Source ID
- N000141512559
Entities
People
- Cap Smith
Organizations
- Colorado State University
- Office of Naval Research
- United States Navy