State estimation plays a crucial role in wearable robotics, particularly for real-time balance monitoring and control. Whole-body state estimation is essential to ensure stable locomotion and safe interaction with users in dynamic motions. Control module (CM) is equipped with an independent 9-axis IMU sensor, enabling robust state estimation of the robot trunk to which time-varying complementary filter (TVCF) is implemented. However, due to the characteristics of wearable robots, which involve frame deformation, impacts, and dynamic motions, using only encoder-based forward kinematics for state estimation results in accumulated errors. Therefore, EXO Lab is researching and developing a robust whole-body state estimation method that utilizes IMU data from other segments, segment torque data calculated from measured actuation currents, and joint data to ensure stability over various motions.

Previously, wearable robots relied on crutches to maintain balance due to the difficulty of stabilizing the system. However, recent advancements have led to the development of wearable robots that support independent maintenance without the need for crutches. EXO Lab is also conducting research and development to support independent balance maintenance by utilizing embedded ground reaction force sensors and estimated whole-body state information. As control module (CM) is playing a role of cerebellum in human motion architecture, CM ensures the simple balancing of the wearable robot through calculating the trajectory of respective joint actuators that can match the real-time center of mass (CoM) and center of pressure (CoP).