Touch driven controller and tactile features for physical interactions - Université Clermont Auvergne Access content directly
Journal Articles Robotics and Autonomous Systems Year : 2020

Touch driven controller and tactile features for physical interactions

Abstract

We propose an approach that considers controlling contact between a robot and the environment during physical interactions. Current physical interaction control approaches are limited in terms of the range of tasks that can be performed. To allow robots to perform more tasks, we derive tactile features representing deformations of the mechanically compliant sensing surface of a tactile sensor and incorporate these features to a robot controller, akin to a visual servo, via touch-and task-dependent tactile feature mapping matrices. As a first contribution, we derive tactile features to localize a contact coordinate frame between an object and an array of pressure sensing elements, with a mechanically compliant surface, attached onto a robot arm end-effector interacting with the object. As a second contribution, we propose tactile projection matrices to design a tactile servoing controller that combines these tactile features with a Cartesian impedance controller of the robot arm. These matrices convert the proposed tactile features to balance not only normal forces but also torques about the sensor's axes. It allows the end-effector to steer the contact frame in a desired manner by regulating errors in the tactile features to address several common issues in robotics: exploration and co-manipulation.
Fichier principal
Vignette du fichier
Kappassov_RAS_TouchDrivenControl_FINAL_Preprint (1).pdf (2 Mo) Télécharger le fichier
Origin : Files produced by the author(s)
Loading...

Dates and versions

hal-02335723 , version 1 (28-10-2019)

Identifiers

Cite

Zhanat Kappassov, Juan Antonio Corrales Ramon, Véronique Perdereau. Touch driven controller and tactile features for physical interactions. Robotics and Autonomous Systems, 2020, 123, pp.103332. ⟨10.1016/j.robot.2019.103332⟩. ⟨hal-02335723⟩
108 View
408 Download

Altmetric

Share

Gmail Facebook X LinkedIn More