Abstract:Planning with learned dynamics models offers a promising approach toward versatile real-world manipulation, particularly in nonprehensile settings such as pushing or rolling, where accurate analytical models are difficult to obtain. However, collecting training data for learning-based methods can be costly and inefficient, as it often relies on randomly sampled interactions that are not necessarily the most informative. Furthermore, learned models tend to exhibit high uncertainty in underexplored regions of the skill space, undermining the reliability of long-horizon planning. To address these challenges, we propose ActivePusher, a novel framework that combines residual-physics modeling with uncertainty-based active learning, to focus data acquisition on the most informative skill parameters. Additionally, ActivePusher seamlessly integrates with model-based kinodynamic planners, leveraging uncertainty estimates to bias control sampling toward more reliable actions. We evaluate our approach in both simulation and real-world environments, and demonstrate that it consistently improves data efficiency and achieves higher planning success rates in comparison to baseline methods. The source code is available at this https URL.
| Comments: | Accepted by the 2026 IEEE International Conference on Robotics & Automation (ICRA 2026) |
| Subjects: | Robotics (cs.RO); Machine Learning (cs.LG) |
| Cite as: | arXiv:2506.04646 [cs.RO] |
| (or arXiv:2506.04646v4 [cs.RO] for this version) | |
| https://doi.org/10.48550/arXiv.2506.04646 arXiv-issued DOI via DataCite |
Submission history
From: Zhuoyun Zhong [view email]
[v1]
Thu, 5 Jun 2025 05:28:14 UTC (4,016 KB)
[v2]
Thu, 18 Sep 2025 14:45:55 UTC (850 KB)
[v3]
Sat, 7 Mar 2026 21:15:56 UTC (880 KB)
[v4]
Wed, 13 May 2026 20:24:13 UTC (880 KB)