aiqbench analysis
This research addresses the transfer gap in legged robotics by using simplified rigid-body models to pretrain policies before progressively adapting them to full-body dynamics. By employing a homotopy-inspired continuation strategy that gradually redistributes mass and inertia, the method achieves faster convergence and superior stability compared to standard baselines. The approach successfully enables complex dynamic behaviors, such as flips and wall-assisted maneuvers, on real quadrupedal robots, offering a practical pathway to reduce reliance on extensive reward tuning.
Dynamic Policy Learning for Legged Robot with Simplified Model Pretraining and Model-Homotopy-Inspired Transfer
arXiv:2512.24698v2 Announce Type: replace Abstract: Generating dynamic motions for legged robots remains a challenging problem. While reinforcement learning has achieved notable success in various legged locomotion tasks, producing highly dynamic behaviors often requires extensive reward tuning or high-quality demonstrations. Leveraging reduced-order models can help mitigate these challenges. However, the model discrepancy poses a significant challenge when transferring policies to full-body dynamics environments. In this work, we introduce a continuation-based learning framework that combines simplified model pretraining and model-homotopy-inspired transfer to efficiently generate and refine complex dynamic behaviors. First, we pretrain the policy using a single rigid body model to capture core motion patterns in a simplified environment. Next, we employ a continuation strategy to progressively transfer the policy to the full-body environment, minimizing performance loss. To define the continuation path, we introduce a parametric transition path from the single rigid body model to the full-body model by gradually redistributing mass and inertia between the trunk and legs. The proposed method achieves faster convergence and demonstrates superior stability during the transfer process compared to baseline methods. Our framework is validated on a range of dynamic tasks, including flips and wall-assisted maneuvers, and is successfully deployed on a real quadrupedal robot.
Abstract:Generating dynamic motions for legged robots remains a challenging problem. While reinforcement learning has achieved notable success in various legged locomotion tasks, producing highly dynamic behaviors often requires extensive reward tuning or high-quality demonstrations. Leveraging reduced-order models can help mitigate these challenges. However, the model discrepancy poses a significant challenge when transferring policies to full-body dynamics environments. In this work, we introduce a continuation-based learning framework that combines simplified model pretraining and model-homotopy-inspired transfer to efficiently generate and refine complex dynamic behaviors. First, we pretrain the policy using a single rigid body model to capture core motion patterns in a simplified environment. Next, we employ a continuation strategy to progressively transfer the policy to the full-body environment, minimizing performance loss. To define the continuation path, we introduce a parametric transition path from the single rigid body model to the full-body model by gradually redistributing mass and inertia between the trunk and legs. The proposed method achieves faster convergence and demonstrates superior stability during the transfer process compared to baseline methods. Our framework is validated on a range of dynamic tasks, including flips and wall-assisted maneuvers, and is successfully deployed on a real quadrupedal robot.
Robotics (cs.RO)
arXiv:2512.24698 [cs.RO]
(or arXiv:2512.24698v2 [cs.RO] for this version)
https://doi.org/10.48550/arXiv.2512.24698 Focus to learn more arXiv-issued DOI via DataCite
IEEE Robotics and Automation Letters, vol. 11, no. 7, pp. 8068-8075, July 2026
https://doi.org/10.1109/LRA.2026.3692082 Focus to learn more DOI(s) linking to related resources
arXivLabs is a framework that allows collaborators to develop and share new arXiv features directly on our website.
Both individuals and organizations that work with arXivLabs have embraced and accepted our values of openness, community, excellence, and user data privacy. arXiv is committed to these values and only works with partners that adhere to them.
Have an idea for a project that will add value for arXiv's community? Learn more about arXivLabs.
Full text under the publisher's open licence · 308 words