This thesis proposes an effective methodology for enhancing the perceptual capabilities and achieving interaction control of the iCub humanoid robot. The method is based on the integration of measurements from different sensors (force/torque, inertial and tactile sensors) distributed along the robot's kinematic chain. Humanoid robots require a substantial amount of sensor information to create their own representations of the surrounding environment. Tactile perception is of primary importance for the exploration process. Also in humans, the tactile system is completely functional at birth. In humanoid robotics, the measurements of forces and torques that the robot exchanges with its surroundings are essential for safe interaction with the environment and with humans. The approach proposed in this thesis can successfully enhance the perceptual capabilities of robots by exploiting only a limited number of both localized and distributed sensors, providing a feasible and convenient solution for achieving active compliance control of humanoid robots.
| ISBN: | 9783319375731 |
| Publication date: | 23rd August 2016 |
| Author: | Matteo Fumagalli |
| Publisher: | Springer an imprint of Springer International Publishing |
| Format: | Paperback |
| Pagination: | 105 pages |
| Series: | Springer Theses |
| Genres: |
Automatic control engineering Artificial intelligence |
This thesis proposes an effective methodology for enhancing the perceptual capabilities and achieving interaction control of the iCub humanoid robot. The method is based on the integration of measurements from different sensors (force/torque, inertial and tactile sensors) distributed along the robot's kinematic chain. Humanoid robots require a substantial amount of sensor information to create their own representations of the surrounding environment. Tactile perception is of primary importance for the exploration process. Also in humans, the tactile system is completely functional at birth. In humanoid robotics, the measurements of forces and torques that the robot exchanges with its surroundings are essential for safe interaction with the environment and with humans. The approach proposed in this thesis can successfully enhance the perceptual capabilities of robots by exploiting only a limited number of both localized and distributed sensors, providing a feasible and convenient solution for achieving active compliance control of humanoid robots.
Increasing Perceptual Skills of Robots Through Proximal Force/Torque Sensors features in the following genres: Automatic control engineering, Artificial intelligence
Increasing Perceptual Skills of Robots Through Proximal Force/Torque Sensors is available in Paperback, Hardback
Increasing Perceptual Skills of Robots Through Proximal Force/Torque Sensors was written by Matteo Fumagalli and published by Springer an imprint of Springer International Publishing
Increasing Perceptual Skills of Robots Through Proximal Force/Torque Sensors has 105 pages
Yes it is part of Springer Theses series