Compliant Residual DAgger

Improving Real-World Contact-Rich Manipulation with Human Corrections

We address key challenges in Dataset Aggregation (DAgger) for real-world contact-rich manipulation: how to collect informative human correction data and how to effectively update policies with this new data. We introduce Compliant Residual DAgger (CR-DAgger), which contains two novel components: 1) a Compliant Intervention Interface that leverages compliance control, allowing humans to provide gentle, accurate delta action corrections without interrupting the ongoing robot policy execution; and 2) a Compliant Residual Policy formulation that learns from human corrections while incorporating force feedback and force control. Our system significantly enhances performance on precise contact-rich manipulation tasks using minimal correction data, improving base policy success rates by over 50% on two challenging tasks (book flipping and belt assembly) while outperforming both retraining-from-scratch and finetuning approaches. Through extensive real-world experiments, we provide practical guidance for implementing effective DAgger in real-world robot learning tasks.

System Overview

To improve a robot manipulation policy, we propose a compliant intervention interface (a) for collecting human correction data, and use this data to update a compliant residual policy (b), and thoroughly study their effects by deploying the updated policy on two contact-rich manipulation tasks in the real world (c).

Compliant Intervention Interface

Findings & Results

Finding 1: Compliant Residual Policy improves base policy by a large margin

Finding 2: Residual policy allows additional useful modality to be added during correction

Finding 3: Smooth On-Policy Delta data makes training more stable

Finding 4: Retraining base policy is stable but learns correction behavior slowly

Finding 5: Finetuning base policy is unstable

Citation

	
@misc{xu2025compliantresidualdaggerimproving,
      title={Compliant Residual DAgger: Improving Real-World Contact-Rich Manipulation with Human Corrections}, 
      author={Xiaomeng Xu and Yifan Hou and Zeyi Liu and Shuran Song},
      year={2025},
      eprint={2506.16685},
      archivePrefix={arXiv},
      primaryClass={cs.RO},
      url={https://arxiv.org/abs/2506.16685}, 
}
						

Contact

Acknowledgement