A Smart Undershirt for Tracking Upper Body Motions: Task Classification and Angle Estimation

The use of interactive or smart textiles that have sensing material(s) incorporated into them supports an emerging technology for physical activity assessment called smart textile systems (STSs). STSs are an increasingly useful technology for researchers, athletes, patients, and others. In this paper, we developed and assessed a novel smart undershirt (SUS) that was designed to monitor low-back (thorax versus pelvis) and shoulder motions. The SUS consists of stretchable undershirt, electronic components, and an array of 11 body-worn sensors (BWSs) printed on the clothing. The BWSs are developed by coating electroactive polymers (i.e., polymerization) on the fabric and are wired using conductive threads. This shirt is the first smart garment for tracking both lower hack and shoulder motions using printed textile sensors. The 16 participants performed 10 upper body movements while wearing the SUS for the purpose of assessing the accuracy of task classification and angle estimation. Input from the SUS led to classification accuracy at the individual levels up to 94% and planar angle estimations with errors on the order of 1.3 degrees and 9.4 degrees for the low back and shoulder, respectively. The performance was poorer, though, at the group level. The SUS appears to be a promising alternative for the purpose of monitoring upper body motions and activities.