Accession Number:

ADA595822

Title:

Multi-Modal Interaction for Robotic Mules

Descriptive Note:

Journal article

Corporate Author:

SOAR TECHNOLOGY INC ANN ARBOR MI

Report Date:

2014-02-26

Pagination or Media Count:

8.0

Abstract:

Today s soldier carries on average over 100lbs of gear, which takes its toll on the soldier and on the mission. To help mitigate this problem, the US Department of Defense is researching the use of robotic mules to move along with squads and help offload some of the excess weight carried by soldiers. The operator control units OCUs for these are typically portable computers with tele-operation or point-and-click interfaces. Instead, the DoD wants heads-up, hands-free methods of interaction that can fit seamlessly into the normal squad interaction patterns. This paper describes our research and prototyping in multi-modal interaction with robotic mules, focused on speech and gesture. We present an analysis of squad interactions to help determine the kind of technology useful for user input recognition. We describe an algorithm for gesture recognition using a 9-axis IMU, results of a formative evaluation, and a prototype multi-modal interface that can be used to command a robotic platform.

Subject Categories:

  • Cybernetics

Distribution Statement:

APPROVED FOR PUBLIC RELEASE