A self-adhesive, elastic fabric, nanocomposite skin-strain sensor (called Motion Tape) has been developed, tested in controlled laboratory environments, and validated through human subject studies. This study aimed to interpret Motion Tape data using deep learning methods to directly predict functional movement parameters (e.g., joint angles and limb positions) and verifying the results using optical motion capture. The approach was to obtain human participant Motion Tape testing data and training the datasets using ground truth values acquired from the optical motion capture system. Predictions of muscle engagement, strain, and range-of-movement of major joints were investigated to validate the proposed methods.
|