Authors:
Nicolas Gorges
;
Stefan Gaa
and
Heinz Wörn
Affiliation:
Institute for Process Control and Robotics, University of Karlsruhe, Germany
Keyword(s):
Tactile Sensing, Object Exploration, Robot Control, and Humanoid Robot.
Related
Ontology
Subjects/Areas/Topics:
Artificial Intelligence
;
Computer Vision, Visualization and Computer Graphics
;
Feature Extraction
;
Features Extraction
;
Formal Methods
;
Hierarchical Control
;
Image and Video Analysis
;
Image Processing
;
Informatics in Control, Automation and Robotics
;
Intelligent Control Systems and Optimization
;
Modeling, Simulation and Architectures
;
Planning and Scheduling
;
Robot Design, Development and Control
;
Robotics and Automation
;
Signal Processing, Sensors, Systems Modeling and Control
;
Simulation and Modeling
;
Symbolic Systems
Abstract:
This work deals with the reactive and autonomous exploration of objects with a humanoid robot using only tactile and kinesthetic sensor feedback. To coordinate the flow of the exploration, a novel hierarchical exploration system is introduced. The lowest level extracts contacts points and elementary features based on the direct contact with the object. It furthermore provides elementary movement primitives. The intermediate level consists of different controlling behaviors to generate exploration movements according to the sensor feedback. This level enables the robot to explore an object pointwisely or continously. The highest level evaluates the process of the exploration and determinates the reactive behavior of the underlying components. The evaluation scenario comprises the exploration of edges, which are arbitrarily located in space. The evaluation platform consists of a robot arm, a
force-torque sensor, and a tactile sensor matrix. The proposed approach is evaluated and the
different reactive behaviors as well as the used sensor modalities are compared.
(More)