Search papers, labs, and topics across Lattice.
This paper develops a data-driven framework for estimating contact forces in robotic surface swabbing using a compliant tool, addressing the challenges posed by nonlinear viscoelastic hysteresis that complicates traditional force measurements. By employing a compact LSTM architecture and a parameter-isolated few-shot adaptation strategy, the method achieves significant improvements in robustness against domain shifts, reducing zero-shot estimation error by up to 63%. The findings underscore the effectiveness of separating deformation-history dynamics from domain-specific conditions, enabling reliable force estimation in diverse and non-stationary environments.
Achieving a 63% reduction in zero-shot estimation error for robotic force sensing could revolutionize how robots interact with complex surfaces in real-world applications.
Robotic surface swabbing requires sustained interaction between a compliant tool and heterogeneous environments, where accurate estimation of tip-level contact force is critical for consistent sampling performance. However, deformable tool dynamics introduce nonlinear viscoelastic hysteresis that decouples wrist-mounted force measurements from true contact forces, while tool-integrated sensors are impractical for deployment due to sterility and disposability constraints. This paper presents a data-driven framework for contact force estimation in Deformable Tool Manipulation (DTM) that leverages proprioceptive sensing without requiring explicit physical models or permanent embedded sensing hardware at the tool tip. A recurrent architecture is first identified through a comparative evaluation of temporal models, where a compact LSTM achieves the lowest estimation error and sub-millisecond inference latency. To address generalization across unseen surfaces and tool compliance conditions, we introduce a parameter-isolated few-shot adaptation strategy that augments a frozen recurrent backbone with low-dimensional context embeddings using feature-wise linear modulation (FiLM). Experiments on a UR5e platform across nine tool-surface interaction regimes demonstrate that the proposed approach significantly improves robustness under domain shift, reducing zero-shot estimation error by up to 63\% while preserving baseline performance without catastrophic forgetting. These results show that separating shared deformation-history dynamics from domain-specific conditioning enables reliable force estimation for DTM in non-stationary environments.
