Agile Information Exchange in Autonomous Air Systems
Abstract
When the quality of service (QoS) of an unmanned air vehicle's (UAV) communications link is poor, and the UAV is tracking a moving target, a trade-off exists between the fidelity and timeliness of information provided from the vehicle to an operator. The authors' 2011 ICCRTS paper showed that an optimal representation scheme for transmitting information from UAV to operator can be found by using information theory. This paper extends that work by introducing an adaptive, autonomous UAV command and control system that autonomously changes the fidelity of information communicated to an operator in response to variances in communications QoS. Results and analysis of hardware-in-the-loop experiments using a UAV that is tracking a moving car are presented. The paper also examines the impact of information decay and network performance on both human tele-operation and on-board autonomous control, comparing the relative performance of tele-operation and autonomy as a function of entropic drag, which is a measurement, in information theory "bits," of the rate at which information is lost due to unpredictable changes in the environment. In the experiments described in this paper, entropic drag is produced by the movement of the vehicle being tracked.
Document Details
- Document Type
- Technical Report
- Publication Date
- Jun 01, 2013
- Accession Number
- ADA588227
Entities
People
- Blair Chisholm
- David Scheidt
- Kevin Schultz
Organizations
- Johns Hopkins University