Search papers, labs, and topics across Lattice.
This paper introduces TWINS, a novel robotic system designed for manipulation tasks that require body-surface contact, addressing a significant gap in existing robotic systems that focus solely on end-effector manipulation. TWINS integrates a Wearable Dual-Arm Device worn by the operator with an Isomorphic Robot that mirrors the operator's joint configuration, allowing for synchronized data collection through distributed tactile sensors. Experimental results show that TWINS effectively enables the learning and execution of manipulation tasks guided by tactile observations from body-surface contact, marking a significant advancement in contact-rich manipulation learning.
TWINS revolutionizes manipulation learning by enabling robots to learn from body-surface contact, a previously overlooked aspect of human-robot interaction.
Recent advances in robot learning for manipulation have increased the importance of collecting real-world demonstration data. However, existing robotic systems primarily focus on end-effector manipulation, making it difficult to teach and execute manipulation tasks involving body-surface contact with the arms and chest. This paper presents TWINS (Tactile Wearable Isomorphic Arm Networked System), a robotic system for manipulation involving body-surface contact. TWINS consists of a Wearable Dual-Arm Device, which is worn by the operator, and an Isomorphic Robot with the same joint configuration and external dimensions. Distributed tactile sensors embedded in the chest and arms enable the measurement of body-surface contact synchronized with joint motion. Using the Wearable Dual-Arm Device, we collected demonstrations for four manipulation tasks involving body-surface contact. We then trained imitation learning policies using the collected demonstrations and deployed them on the Isomorphic Robot, enabling manipulation guided by body-surface tactile observations. Experimental results demonstrate that TWINS provides a unified robotic system for demonstration, learning, and execution of manipulation involving body-surface contact. https://mmurooka.github.io/twins-project-page/
