A systematic comparison of two state-of-the-art motion-imitation reinforcement learning (MIRL) pipelines, one built on SCONE/HyFyDy and one built on MuJoCo/MyoSim, concludes that the more advanced physiological realism of HyFyDy currently makes it more suitable for musculoskeletal modeling.
Abstract
In this work, we conduct a systematic comparison of two state-of-the-art motion-imitation reinforcement learning (MIRL) pipelines, one built on SCONE/HyFyDy and one built on MuJoCo/MyoSim. HyFyDy emphasizes physiological realism through detailed musculotendon modeling, while MuJoCo prioritizes computational efficiency and scalable policy learning. While recent work has demonstrated that both pipelines reproduce human kinematics with high fidelity, it remains unclear if they accurately capture the underlying neuromuscular behavior that produced the movement. This limitation is particularly important for robotic assistive-device design and control, where outcome measures such as muscle activation patterns and metabolic cost are often used as optimization targets. To conduct a systematic comparison, our work compares both pipelines using a common set of human motion-capture and electromyography (EMG) measurements. The results find that while both pipelines produce similar kinematics with relative accuracy, the muscle activations from HyFyDy are more aligned with the experimental EMG, as supported by the average pooled (RMSE, r) values for muscle activations from HyFyDy and MuJoCo: (0.164, 0.4) and (0.344, 0.11), respectively. While we conclude that the more advanced physiological realism of HyFyDy currently makes it more suitable for musculoskeletal modeling, both require further development to bring physiological realism to GPU-parallelizable simulation environments and advance robotic assistive device design.
Predictive simulation of human movement is a promising tool for studying ``what-if''scenarios in human movement and its underlying motor control, yet its realism is often limited. To address this gap, we incorporate a learned state prior that is trained on a large-scale dataset of human gait kinematics and external for...
Markus Gambietz, Zhi-Hao Zhao, Theodoros Balougias et al.· 0 citations
Reinforcement learning (RL) has achieved notable success in locomotion of robotic systems, yet remains challenged by high-dimensional and overactuated musculoskeletal systems. In this work, we propose a synergy-guided RL framework that incorporates biologically inspired priors extracted from pre-trained locomotion poli...
Daily locomotion encompasses diverse activities and frequent transitions between them, yet most exoskeleton controllers are designed for a single activity or a narrow set of related movements. Changes in activity therefore typically require explicit mode switching and separately tuned or retrained controllers. Simulati...
Yi-Fei Yuan, Jakob H. Wolf, Ghaith J. Androwis et al.· 0 citations
The results show that work-aware policy optimization can suppress physically inefficient motion without requiring an explicit differentiable dynamics model.
Toshiki Otani, Hiromu Taketsugu, N. Ukita· 0 citations
Accurate intention-aware joint angle trajectory prediction from surface electromyography (sEMG) is central to responsive lower-limb exoskeleton reference trajectory generation, particularly under subject and protocol variability. This paper proposes NCE-Net, a novel neuromuscular context encoding architecture for real-...
M. Belal, N. Elhendawi, A. Aljarah et al.· Computers in Biology and Med...· 0 citations
CHOREO, a framework for training-free composition of heterogeneous humanoid skills, demonstrates that executable trajectories provide a scalable interface for accumulating and composing pretrained humanoid capabilities.
Zi-Yi Sun, Jing-Wen Chen, Yu-Xin Wang et al.· arXiv.org· 1 citation· ⚡1
Related blog posts
MIT News · Artificial Intelligence· news.mit.eduNov 9, 2026