There is a limit to the optimal design of the robot platform due to a low power density and a large volume of the form factor of conventional actuators. Therefore, we are developing new actuators with a specific target use, to be used for cable-driven assistive robots. These include low-profile, high-bandwidth actuators, and actuators with multiple cable outputs that rely on lower number of motors.

We are developing a next-generation modular wearable robot, featuring flexible independent fabric-driven modular structures, real-time gait cycle recognition, and personalized optimization technologies. Due to the modular nature, This robot enables personalized assistance by configuring the target regions on the body. It provides tailored support according to the physical characteristics and various gait patterns of each wearer. By utilizing real-time sensor data to discern the terrain and support body parts, the robot provides optimal assistance strategies to the wearer.

Many motions in mobile manipulators are conducted quasi-static and non-periodically at low speed. Due to these motions’ features, about 90% of the required torque at actuators is generated by the gravitational force. In other words, reducing the gravitational torque can dramatically improve the efficiency of the actuator. Moreover, the payload on the end-effector is varied during the tasks, leading to the variation of the required gravitational torque even in the same mobile manipulator.

We are developing a variable gravity compensation module that can compensate for the gravitational torque and adjust the degree of compensation according to the payload variation. Furthermore, by combining the variable gravity compensation module with customized high-power and compact actuators, we present a novel approach to energy-efficient actuators in the robotic field.