Accurate real‐time motion tracking is essential for applications such as human–robot interaction (HRI) and rehabilitation. However, traditional wearable systems often suffer from rigidity, discomfort, and complex wiring, which restrict natural movement and limit long‐term usability. Additionally, their accuracy in real‐time environments is compromised due to challenges in determining optimal sensor placement for neural network‐based motion estimation. To address these limitations, we have developed a fully soft sensing suit that integrates soft sensors, wires, and electrodes. Our approach employs an embedded optimization method for sensor placement, supported by developed programs that automate the entire process—from data preprocessing to verification—while comparing placements derived from wrapper and filter methods. Using motion capture data, we identified the optimal sensor placement and evaluated the motion tracking performance of the fabricated sensing suit. Experimental verification demonstrated an average root mean square error (RMSE) of 24.35 mm and an R 2 value of 0.8736, highlighting the system's potential for accurate and unobtrusive real‐time motion tracking in daily‐life scenarios.
Oh et al. (Thu,) studied this question.