Accurate Estimation of Joint Reaction of Users in Wearable Robots Using Multiple Wireless 3-DOF Interaction Force Sensor System

Prof. Hyung-Soon Park’s group at KAIST has developed a soft, wireless 3-degree-of-freedom (DOF) force measurement system designed to be embedded within a wearable robot to allow real-time measurement of physical human-robot interaction. This system addresses a critical challenge in wearable robotics: accurate estimation of joint reaction forces and moments. This is essential for preventing possible issues to ensure long-term user safety, a concern of particular importance as these wearable devices are increasingly adopted by vulnerable populations such as the elderly. Conventional methods for analyzing the effects of wearable robots have offered limited performance, as they typically measure only one-directional normal interaction forces while ignoring crucial shear components. The team’s new system overcomes this by using its soft sensors to measure 3-DOF interaction forces and then incorporating the obtained data into a “modified inverse dynamics” model to more accurately and comprehensively estimate reaction forces and moments on the user’s joint.

This system is composed of compact, soft 3-DOF sensor units, with each unit containing a magnet and Hall sensors, and a wireless data acquisition module. Designed to replace the original contact parts of a wearable robot, the sensors measure the full 3-DOF interaction forces, capturing both normal and sheer components. The data are then incorporated into a novel “modified inverse dynamics” model to more accurately calculate the forces and moments acting exclusively on the user’s joints. Experimental results confirmed that the conventional analysis, which uses only the ground reaction forces, significantly underestimates the biomechanical loads on the user’s lower limb joints during assisted movements.

“This technology addresses a critical safety blind spot. By understanding how wearable robots could inadvertently harm users, especially the elderly in their daily lives, we can now engineer next-generation systems that are truly safe for long-term use,” said Hyung-Soon Park, the corresponding author and a director of KI for Human Augmentation Convergence at KAIST. “This development is significant because it overcomes the limitations of conventional estimation methods. It provides a foundational tool that will be crucial for assessing and improving the safety of future wearable systems,” said the first author, Seongyun Cho.

This study, titled “A soft 3-DOF interaction force measurement system for estimating the biomechanical effects of a soft wearable robot on the human joint” was published in Wearable Technologies, Volume 6: e32, 2025. This work was supported by T-Robotics Co., Ltd.