Robotics and Intelligent Learning Lab

We develop intelligent, human-centered wearable robotic systems and humanoid robots that learn from and adapt to dynamic environments. Our research integrates robotics, artificial intelligence, biomechanics and control to advance assistive technologies and embodied intelligence.

The Robotics and Intelligent Learning Lab is dedicated to advancing the human-robot interaction through intelligent control, biomechanical simulation and reinforcement learning. Our research spans a range of domains, including wearable exoskeletons, neuromechanical modeling, low-gravity human locomotion and AI-driven rehabilitation technologies. The lab emphasizes interdisciplinary collaboration and real-world validation of algorithms in both simulation and hardware systems. Current projects include the development of full lower-limb exoskeleton, hip exoskeleton and upper-limb exoskeletons for mobility assistance, physics-based human-exoskeleton co-simulation frameworks, and learning-based controllers for dynamic and uncertain environments.

The Robotics and Intelligent Learning Lab also has been instrumented with an eight-degrees-of-freedom (8-DOF) portable full lower limb exoskeleton, robotic hip exoskeleton, a four-DOF upper limb exoskeleton, a heavy-duty treadmill, mobile insole sensors, state-of-the-art 3D printers and a low-gravity experimental platform.

Portable, Full Lower Limb Exoskeleton and Suspension-Based, Low-Gravity Experimental Platform

The eight-degrees-of-freedom (8-DOF) portable full lower-limb exoskeleton (Fig. 1) developed by Dr. Luo's research team provides active torque assistance at the hip, knee and ankle joints. It incorporates high-efficiency Cycloidal drive motors with an optimized gear ratio, enabling torque outputs up to 80 Nm per joint. This design balances compact form factor with high-torque capability, essential for supporting body weight during stance phases or enhancing joint movement during rehabilitation and load-bearing tasks.

This suspension-based, low-gravity simulator (Fig. 2) uses barbell plates to generate the pull-up force, a pulley cable for force transmission, a load cell sensor to measure the force and a harness that applies the force to the subject’s center of mass. The system also includes a treadmill to control the subject’s walking or running speed under different low-gravity conditions. The weight of the barbell plates is customizable, allowing them to apply a pull-up force that counteracts the subject’s body weight and reduces the load on their center of mass.

Wearable Robot Systems

Exoskeletons: Portable robotic hip exosuit (Fig. 3a) (24Nm,  2.3kg); Four-DOF upper limb exoskeleton. (Fig. 3b)

Portable Biomechanics Analysis Equipment

• Wearable motion sensors: A set of customized wireless IMU systems includes eight measurement nodes and 20 Individual IMU measurement nodes.
• Portable Muscle sensors: Noraxon Ultium Electromyography (EMG) Measurement Units; Ultium EMG sensors sample up to 4,000 times per second, synchronize in real-time and demonstrate low baseline noise (<1 μV RMS) with minimal native artifacts.
• Torque and Force Measurement Sensors: Two loadcells (0~50Nm, Futek, Inc.), four customized loadcells (0~50Nm, 0~100Nm) and one six-axis loadcell (Sunrise, Inc.)

Miscellaneous

• Education platform: Advanced biomedical exoskeleton education kits, 20 Arduino Mega2560 R3, Arduino control board-based elbow exoskeleton kits are designed for teaching kits, which can provide elbow joint extension/flexion assistance.