aiqbench analysis
This research addresses the critical sim-to-real gap in robotics by leveraging proprioceptive consistency to train policies that generalize directly to hardware without fine-tuning. The proposed PACE encoder effectively filters out static domain-specific noise like lighting variations while preserving essential motion cues for contact-rich assembly tasks. Achieving a 93.3% success rate with minimal performance degradation, this work advances the practical deployment of reinforcement learning in complex physical environments.
Zero-Shot Sim-to-Real Contact-Rich Assembly via Proprioception-Anchored Cross-Modal Pretraining
arXiv:2609.07534v1 Announce Type: new Abstract: Contact-rich assembly remains challenging because it requires submillimeter spatial accuracy and reliable interpretation of forces during sustained contact. Although simulation-based reinforcement learning offers a scalable training paradigm, discrepancies in visual observations, contact dynamics, and force/torque (F/T) measurements often limit policy transfer. We observe that proprioception is comparatively consistent across domains because calibrated joint positions and consistently computed joint velocities align closely between simulation and hardware. Based on this observation, we present PACE (Proprioception-Anchored Cross-Modal Encoder), which supervises temporal visual and F/T representations by predicting proprioceptive state transitions. Static domain-specific factors, including lighting, texture, and sensor bias, contain little information about joint motion; the proposed objective therefore encourages the encoder to suppress these factors while retaining task-relevant motion cues. Policies trained on frozen PACE features are deployed on hardware without real-world fine-tuning or object-pose tracking. Across four contact-rich assembly tasks, PACE attains an average real-world success rate of 93.3\% and only a 2.7-percentage-point sim-to-real drop, meanwhile remaining robust to perturbations that substantially degrade pose-based and learned-fusion baselines.
Abstract:Contact-rich assembly remains challenging because it requires submillimeter spatial accuracy and reliable interpretation of forces during sustained contact. Although simulation-based reinforcement learning offers a scalable training paradigm, discrepancies in visual observations, contact dynamics, and force/torque (F/T) measurements often limit policy transfer. We observe that proprioception is comparatively consistent across domains because calibrated joint positions and consistently computed joint velocities align closely between simulation and hardware. Based on this observation, we present PACE (Proprioception-Anchored Cross-Modal Encoder), which supervises temporal visual and F/T representations by predicting proprioceptive state transitions. Static domain-specific factors, including lighting, texture, and sensor bias, contain little information about joint motion; the proposed objective therefore encourages the encoder to suppress these factors while retaining task-relevant motion cues. Policies trained on frozen PACE features are deployed on hardware without real-world fine-tuning or object-pose tracking. Across four contact-rich assembly tasks, PACE attains an average real-world success rate of 93.3\% and only a 2.7-percentage-point sim-to-real drop, meanwhile remaining robust to perturbations that substantially degrade pose-based and learned-fusion baselines.
Robotics (cs.RO); Artificial Intelligence (cs.AI)
arXiv:2609.07534 [cs.RO]
(or arXiv:2609.07534v1 [cs.RO] for this version)
https://doi.org/10.48550/arXiv.2609.07534 Focus to learn more arXiv-issued DOI via DataCite (pending registration)
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 · 262 words
Related coverage
Contact-Rich Robotic Manipulation in Construction via Zero-Shot Learning: A Diffusion Policy-Guided Adaptive Control
arXiv · Aug 25, 2026
TAM: Torque Adaptation Module for Robust Motion Transfer in Manipulation
arXiv · Sep 9, 2026
TAM: Torque Adaptation Module for Robust Motion Transfer in Manipulation
arXiv · Jun 6, 2026
Play2Perfect: What Matters in Dexterous Play Pretraining for Precise Assembly?
arXiv · Sep 9, 2026