Contact-Anchored Policies: Contact Conditioning Creates Strong Robot Utility Models
This work addresses the challenge of improving robot manipulation robustness and generalization for robotics applications, representing a novel approach rather than an incremental improvement.
The paper tackles the problem of language being too abstract for robust robot manipulation by introducing Contact-Anchored Policies (CAP), which condition on physical contact points and use modular utility models, resulting in generalization to novel environments and embodiments with only 23 hours of data and outperforming state-of-the-art VLAs by 56% in zero-shot evaluations.
The prevalent paradigm in robot learning attempts to generalize across environments, embodiments, and tasks with language prompts at runtime. A fundamental tension limits this approach: language is often too abstract to guide the concrete physical understanding required for robust manipulation. In this work, we introduce Contact-Anchored Policies (CAP), which replace language conditioning with points of physical contact in space. Simultaneously, we structure CAP as a library of modular utility models rather than a monolithic generalist policy. This factorization allows us to implement a real-to-sim iteration cycle: we build EgoGym, a lightweight simulation benchmark, to rapidly identify failure modes and refine our models and datasets prior to real-world deployment. We show that by conditioning on contact and iterating via simulation, CAP generalizes to novel environments and embodiments out of the box on three fundamental manipulation skills while using only 23 hours of demonstration data, and outperforms large, state-of-the-art VLAs in zero-shot evaluations by 56%. All model checkpoints, codebase, hardware, simulation, and datasets will be open-sourced. Project page: https://cap-policy.github.io/