Uploaded May 2018 | Updated September 2026, 2 weeks ago
ICRA 2018 Spotlight Video
Interactive Session Thu AM Pod B.5
Authors: Kollmitz, Marina; Büscher, Daniel; Schubert, Tobias; Burgard, Wolfram
Title: Whole-Body Sensory Concept for Compliant Mobile Robots
Abstract:
Most of the conventional approaches to mobile robot navigation avoid any kind of contact with the environment or with humans. As nowadays distance sensors typically have a limited - and often only two-dimensional - field of view, collisions with the environment or contacts with humans cannot be fully avoided in practical mobile robot applications. On the other hand, direct physical contact can be used for intuitive communication between a robot and humans. In this paper, we present a whole-body sensory concept based on a 6-DoF force-torque sensor to perceive physical interaction between the robot and humans. To distinguish between external contact and disturbance forces that result from the motion of the mobile platform or oscillations, we present a novel model-free filtering approach based on a neural network. In extensive experiments carried out with our robot Canny we demonstrate the effectiveness and advantages of the neural network approach, which clearly outperforms a classical model-based one.
ICRA 2018 Spotlight Video
Interactive Session Thu AM Pod B.5
Authors: Kollmitz, Marina; Büscher, Daniel; Schubert, Tobias; Burgard, Wolfram
Title: Whole-Body Sensory Concept for Compliant Mobile Robots
Abstract:
Most of the conventional approaches to mobile robot navigation avoid any kind of contact with the environment or with humans. As nowadays distance sensors typically have a limited - and often only two-dimensional - field of view, collisions with the environment or contacts with humans cannot be fully avoided in practical mobile robot applications. On the other hand, direct physical contact can be used for intuitive communication between a robot and humans. In this paper, we present a whole-body sensory concept based on a 6-DoF force-torque sensor to perceive physical interaction between the robot and humans. To distinguish between external contact and disturbance forces that result from the motion of the mobile platform or oscillations, we present a novel model-free filtering approach based on a neural network. In extensive experiments carried out with our robot Canny we demonstrate the effectiveness and advantages of the neural network approach, which clearly outperforms a classical model-based one.
