Tutorial on Embodiment

3.3.3. Exercise

Theoretical Scheme: Grasping case studies*

In the two case studies presented, there is no need for the agent to ‘know' beforehand what the shape of the to-be-grasped object will be (which is normally the case in robotics, where the contact points are calculated before the grasping action: Molina-Vilaplana et al., 2007). In the first study, the shape adaptation is taken over by the morphology of the hand, the elasticity of the tendons, and the deformability of the finger tips, as the hand interacts with the shape of the object. In the second study, the physical properties of the granular material and how they change when air is evacuated play a key part.

How can we visualize this in the theoretical scheme?





* adapted from Hoffmann and Pfeifer, 2011


Molina-Vilaplana, J.; Feliu-Batlle, J. & Lopez-Coronado, J. (2007), 'A modular neural network architecture for step-wise learning of grasping tasks', Neural Networks 20, 631-645.
Hoffmann, M. & Pfeifer, R. (2011), The implications of embodiment for behavior and cognition: animal and robotic case studies, in W. Tschacher & C. Bergomi, ed., The Implications of Embodiment: Cognition and Communication, Exeter: Imprint Academic, pp. 31-58.