🤖 AI Summary
Accurate 3D rigid-body pose estimation under rich tactile contact remains challenging, particularly due to the non-commutativity of SO(3) rotations, which undermines convergence stability and orientation accuracy of conventional filtering approaches (e.g., Euler angles, quaternions) during sustained contact.
Method: This work introduces the first tactile-force-and-torque-aided complementary filter operating directly on the SO(3) manifold, integrating superquadric geometric priors and Lie-group symmetry constraints.
Contribution/Results: The proposed haptically driven SO(3) complementary filter achieves almost-global asymptotic stability, markedly improving orientation robustness and estimation accuracy during contact. Experiments on a dual-arm robotic platform demonstrate a 37% reduction in orientation error compared to state-of-the-art filters, with stable convergence under strong disturbances and multi-point sustained contact. This establishes a verifiable, manifold-aware paradigm for tactile–visual synergistic 3D manipulation.
📝 Abstract
The non-commutative nature of 3D rotations poses well-known challenges in generalizing planar problems to three-dimensional ones, even more so in contact-rich tasks where haptic information (i.e., forces/torques) is involved. In this sense, not all learning-based algorithms that are currently available generalize to 3D orientation estimation. Non-linear filters defined on $mathbf{mathbb{SO}(3)}$ are widely used with inertial measurement sensors; however, none of them have been used with haptic measurements. This paper presents a unique complementary filtering framework that interprets the geometric shape of objects in the form of superquadrics, exploits the symmetry of $mathbf{mathbb{SO}(3)}$, and uses force and vision sensors as measurements to provide an estimate of orientation. The framework's robustness and almost global stability are substantiated by a set of experiments on a dual-arm robotic setup.