Speech-Based Robotic Control for Dismounted Soldiers: Evaluation of Visual Display Options
Abstract
Twenty-one enlisted Soldiers supported this user-based evaluation of robot controller options: speech control versus a manual wrist-worn mouse, crossed with two types of visual display options grid versus object segmentation (OS). Performance on the maneuver course showed slower times for the speech controller, across both levels of workload. The slower times for the speech control can be attributed to the percent of times that the speech control had to be repeated. In addition, the two visual display options were evaluated using a search lane that was cluttered with objects. In this event, the grid display was significantly faster. The OS display was much slower when used with the speech controller. Driving errors were higher in the OS-speech controller condition. Soldier feedback with regard to user experience showed positive regard for the concept of speech control, particularly for its contribution to hands-free operation and maintenance of situation awareness. Improvements for speech control were suggested with regard to robot responsiveness, robot maneuvering, and staying within course limits. The OS display also needs improvement with regard to object labeling updates. As it is currently configured, the labels of the objects can change while the robot is moving, particularly in a cluttered environment.
Document Details
- Document Type
- Technical Report
- Publication Date
- May 01, 2014
- Accession Number
- ADA601525
Entities
People
- Christian B. Carstens
- Linda R. Elliott
- Rodger A. Pettitt
Organizations
- United States Army Research Laboratory