Share Email Print
cover

Proceedings Paper

An agent-based approach to decluttering the interfaces of multi-UAV command and control systems
Author(s): Stephen O'Hara; Nathan Dwyer
Format Member Price Non-Member Price
PDF $14.40 $18.00

Paper Abstract

Information overload and cluttered user interfaces can lead to decreased situational awareness and lowered performance of human operators. Irrelevant data increases searching times for tasks requiring the identification of threats, causing delayed decisions. Cognitive burden on the user increases as displays become more cluttered, which results in increased operator stress leading to poor decision-making ability. To address this issue, we have designed an intelligent agentbased system for the automatic de-cluttering of a representative net-centric interface designed for controlling multiple unmanned aerial vehicles (UAVs) by a single operator. Our concept is called ARID, for Agent-based Reduction of Information Density. The ARID hypothesis is that intelligent agents can improve operator performance by deemphasizing those aspects of a display that can be inferred as less-important to the mission goals. ARID agents receive information about the world via data feeds provided by various net-centric sources. Each agent has an understanding of the user interface symbols that are used to represent various entities, terrain features, and zones. The agent also is provided with a mission goal which is used for inferring the relevance of a given symbol to the success of the mission goal. First level facts, such as spatial relationships, are calculated by supporting agents and assigned a BDU (belief/disbelief/uncertainty) value. A dynamic set of rules provides an inference mechanism by which an agent can infer new facts from the given assertions. We have developed a Subjective Logic-based Evidential Reasoning Network that explicitly deals with belief and uncertainty in the knowledge base, and is used to derive a relevancy belief for every UI symbol in the map display. Subjective Logic is used to combine values when different sources provide different results for the same symbol. User Interface agents apply the results of the relevancy beliefs and transform the display to minimize the apparent clutter caused by less relevant elements. Two transformations, transparency and grouping, are used in the current implementation.

Paper Details

Date Published: 2 May 2007
PDF: 11 pages
Proc. SPIE 6561, Unmanned Systems Technology IX, 65610W (2 May 2007); doi: 10.1117/12.718388
Show Author Affiliations
Stephen O'Hara, 21st Century Systems, Inc. (United States)
Nathan Dwyer, 21st Century Systems, Inc. (United States)


Published in SPIE Proceedings Vol. 6561:
Unmanned Systems Technology IX
Grant R. Gerhart; Douglas W. Gage; Charles M. Shoemaker, Editor(s)

© SPIE. Terms of Use
Back to Top