Robots are equipped with an increasingly wide array of sensors in order to enable advanced sensorimotor capabilities. However, the efficient exploitation of the resulting data streams remains an open problem. We present a framework for learning when and where to attend in a sensorimotor stream in order to estimate specific task properties, such as the mass of an object. We also identify the qualitative similarity of this ability between simulation and robotic system. The framework is evaluated for a non-prehensile ”topple-and-slide” task, where the data from a set of sensorimotor streams are used to predict the task property, such as object mass, friction coefficient, and compliance of the block being manipulated. Given the collected data streams for situations where the block properties are known, the method combines the use of variance-based feature selection and partial least-squares estimation in order to build a robust predictive model for the block properties. This model can then be used to make accurate predictions during a new manipulation. We demonstrate results for both simulation and robotic system using up to 110 sensorimotor data streams, which include joint torques, wrist forces/torques, and tactile information. The results show that task properties such as object mass, friction coefficient and compliance can be estimated with good accuracy from the sensorimotor streams observed during a manipulation. ; Science, Faculty of ; Computer Science, Department of ; Graduate
Exploring structured predictions from sensorimotor data during non-prehensile manipulation using both simulations and robots
2013-01-01
Theses
Electronic Resource
English
DDC: | 629 |
Experiments in constrained prehensile manipulation: Distributed manipulation with ropes
Springer Verlag | 2008
|Non-Prehensile Aerial Manipulation using Model-Based Deep Reinforcement Learning
ArXiv | 2024
|Passive Non-Prehensile Manipulation on Helix Path Based on Mechanical Intelligence
BASE | 2014
|Passive Non-Prehensile Manipulation on Helix Path Based on Mechanical Intelligence
BASE | 2014
|