Robotics: Science and Systems XXII

Semantic Contact Fields for Category-Level Generalizable Tool Manipulation

Kevin Yuchen Ma, Heng Zhang, Weisi Lin, Mike Zheng Shou, Yan Wu

Abstract:

Generalizing tool manipulation requires both semantic planning and precise physical control. Modern generalist robot policies, such as Vision-Language-Action (VLA) models, often lack the high-fidelity physical grounding required for contact-rich tool manipulation. Conversely, existing contact-aware policies that leverage tactile or haptic sensing are typically instance-specific and fail to generalize across diverse tool geometries. Bridging this gap requires learning unified contact representations from diverse data, yet a fundamental barrier remains: diverse real-world tactile data are prohibitive at scale, while direct zero-shot sim-to-real transfer is challenging due to the complex dynamics of nonlinear deformation of soft sensors.
    To address this, we propose Semantic-Contact Fields (SCFields), a unified 3D representation fusing visual semantics with dense contact estimates. We enable this via a two-stage Sim-to-Real Contact Learning Pipeline: first, we pre-train on a large simulation data set to learn general contact physics; second, we fine-tune on a small set of real data, pseudo-labeled via geometric heuristics and force optimization, to align sensor characteristics. This allows physical generalization to unseen tools. We leverage SCFields as the dense observation input for a diffusion policy to enable robust execution of contact-rich tool manipulation tasks. Experiments on scraping, crayon drawing, and peeling demonstrate robust category-level generalization, significantly outperforming vision-only and raw-tactile baselines.

Download:

Bibtex:

  
@INPROCEEDINGS{MaK-RSS-26, 
    AUTHOR    = {Kevin Yuchen Ma AND Heng Zhang AND Weisi Lin AND Mike Zheng Shou AND Yan Wu}, 
    TITLE     = {{Semantic Contact Fields for Category-Level Generalizable Tool Manipulation}}, 
    BOOKTITLE = {Proceedings of Robotics: Science and Systems}, 
    YEAR      = {2026}, 
    ADDRESS   = {Sydney, Australia}, 
    MONTH     = {July}, 
    DOI       = {10.15607/RSS.2026.XXII.004} 
}