Using Wearable Sensors to Enhance Communication in the U.S. Army

Author(s):

Megan Weaver, Kayna Anderson, Ethan Hicks, Christian Kantz, Stephen Cain, Russell Cohen Hoffing, and Andy Tennant

Abstract:

Efficient communication in the military is essential for the success of wars and safety of Soldiers. Many current means of communication present in the military are outdated (e.g., radio), require Soldiers to turn from their line of sight to receive the hand signals, or require sufficient waiting periods to receive the message. The group devised an innovative approach to address this challenge through a nonverbal hand communication system. The device consists of three main components: an inertial measurement unit (IMU) glove for tracking and transmitting hand signals, heads-up display glasses for receiving incoming messages, and a commander’s app for sending and receiving signals. The objective was to focus on the proof of concept that IMU gloves can generate and process hand signals for transmission. Using Rokoko smart gloves, we successfully tracked static hand signals from numbers zero to nine. The files were exported, converted, and then inputted in MATLAB for data analysis. Each hand signal comprised 88 unique quaternion points spanned across the hand and onto the arm, recorded at 30 fps. Five individuals (three female, two male) participated in recording the hand signals from zero to nine. Quaternion data for each number and individual were initialized and inputted into a machine-learning algorithm. The algorithm successfully identified numbers based on finger angle orientations and displayed the designated signal. The machine model achieved 99.9% classification accuracy with one false positive and negative. These findings demonstrate the success of the proof of concept and hold promise for a nonverbal communication device.