Implementation of spiking tactile system on a 7-DoF manipulator and development of manipulation algorithms for
stable grasping of objects of different shape and softness. Tactile data will be integrated in the control loop of the manipulator
to obtain optimal grasps taking the object shape, softness and deformability into account. Different types of end-effectors will
be considered, i.e. from parallel grippers to under-actuated multi-fingered hands.
a new set of force controllers, i.e. impedance and admittance controllers, will be obtained allowing
different robots, from mobile manipulators to humanoid robots, performing grasping of a large variety of objects. The
performance of the new grasping algorithms will be compared to the performance obtained using grasping algorithms based on
open loop approaches and those relying on data provided by a wrist-mounted force/torque sensor. The goal is to show that the
new algorithms based on the tactile system provide a much higher ratio of successful grasps considering a set of objects and
different configurations. The demonstrator will feature a service robot being able to grasp a variety of objects involved in
typical home daily tasks of elderly care scenarios.
implementation of spiking tactile skin for robots
to learn predictive methods
to learn manipulation of switches
Enrolments (in Doctoral degree/s)
S. Terreri, C. Bartolozzi, L. Natale, T. Prescott, R. Haschke