Current Work
MyoAssist
I am the Project Lead and active developer of MyoAssist, which is an open-source simulation platform that accelerates the development of assistive technologies like prosthetics and exoskeletons by modeling human-device interactions. Built within the MyoSuite framework, it provides researchers with ready-to-use simulation environments combining detailed musculoskeletal models with assistive devices, along with machine learning tools for developing advanced control strategies.
Assistive Device Optimization
I am working on developing a simulation-based approach to optimize wearable assistive devices through neuromechanical simulations in MyoAssist. Projects have included exoskeleton control optimization across various locomotion tasks and prosthetic ankle stiffness optimization.
Actuator Design
I am redesigning a compact nonlinear series elastic actuator for cable-driven exoskeletons and other systems. The actuator uses a CAM-based spring mechanism to maintain sensitive low-force cable tension while smoothly transitioning to high-force assistance with the goal of improving actuator packaging, wearability, and controllability.
Joint Moment Estimation
I am developing a task-agnostic ankle exoskeleton control framework that estimates biological ankle joint moment directly from onboard wearable sensors without real time ground reaction force data. The project uses a temporal convolutional neural network to map exoskeleton sensor data to ankle torque and stance/swing predictions, reducing the need for explicit gait-state classification or task-specific controller tuning.
Papers & Presentations
Journal Papers
Price, M., Robbins, C., Szemethy, S., Abdikadirova, B., Olson, G., Hoogkamer, W., & Huber, M. E. (2025). Adjustable Compliance Footwear Technology to Investigate Gait Adaptation. IEEE Robotics and Automation Letters, vol. 10, no. 12, pp. 13352–13358. doi:10.1109/LRA.2025.3629969
Conference Papers
Tan, C. K., Wang, C., Lyu, S., Hodossy, B. K., Schumacher, P., Wilson, E. B., Caggiano, V., Kumar, V., Farina, D., Gionfrida, L., Rouse, E. J., Durandau, G., & Song, S. (2025). Myoassist 0.1: Myosuite for Dexterity and Agility in Bionic Humans. In 2025 International Conference On Rehabilitation Robotics (ICORR) (pp. 437-442). IEEE. doi:10.1109/ICORR66766.2025.11063089
NOTE: Presented on behalf of Chun Kwang Tan @ ICORR 2025
Conference Abstracts
Robbins, C., Tan, C. K., & Song, S. (2025). Towards Simulation-Based Exoskeleton Control Design and Optimization. American Society of Biomechanics (ASB).
Robbins, C., Grimmitt, A., Price, M., & Hoogkamer, W. (2024). Joint Kinematics During Steep Uphill Walking. American Society of Biomechanics (ASB).
Presentations
Song, S., Robbins, C., & Son, H. (2025). MyoAssist: Simulating Human Neuromechanics and Assistive Devices in MyoSuite. ASB 2025 Tutorial Workshop.
A hands-on tutorial introducing MyoAssist, our open-source platform for neuromechanical simulation and assistive robotics development. The workshop covered practical implementation of exoskeleton environments, RL-based controllers, and optimization techniques for generating diverse gait behaviors.
August 13th, 2025, Pittsburgh, PA
American Society of Biomechanics
August 2025, Pittsburgh, PA
Rehabweek - ICORR
May 2025, Chicago, IL
American Society of Biomechanics
August 2024, Madison, WI
Previous Work
Wearable Robotic Shoes
I helped develop wearable robotic shoes for gait rehabilitation. I designed and prototyped custom pneumatic air pockets that can adjust the stiffness of the shoe sole during walking, allowing patients to experience gait perturbations in natural environments rather than being confined to laboratory settings. The system uses flexible components and embedded sensors to maintain a lightweight, portable design.
Steep Slopes and Footwear Bending Stiffness
I investigated whether carbon-plated running shoes provide biomechanical advantages during extreme uphill walking. I designed and carried out the full study, analyzing joint kinematics across three footwear types (minimalist, standard, and carbon-plated) on inclines up to 30°. Prior to the study, I helped build a custom steep-incline treadmill system capable of adjusting from 5-40° for testing at these extreme slopes.