Current AchievementsWORK AT EPFL
Tactile Correction for Policy Refinement and Reuse
WORK AT IIT
An embedded skin for humanoid robots
The skin is made up of triangular, flexible printed circuit boards which act as sensors, and it covers much of iCub's body. Each bendy triangle is 3 centimetres to a side and contains 12 capacitive copper contacts (pictured). A layer of silicone rubber acts as a spacer between those boards and an outer layer of Lycra that carries a metal contact above each copper contact. The Lycra layer and flexible circuits constitute the two sides of the skin's pressure-sensing capacitors. This arrangement allows 12 "tactile pixels" - or taxels - to be sensed per triangle. This taxel resolution is enough to recognise patterns such as a hand grasping the robot's arm. The skin can detect a touch as light as 1 gram across each taxel. It is also peppered with semiconductor-based temperature sensors.
Later, IIT plans to add a layer of a piezoelectric polymer called PVDF to the skin. While the capacitance sensors measure absolute pressure, the voltage produced by PVDF as a result of its deformation when touched can be used to measure the rate of change of pressure. So if the robot runs its fingertip along a surface, the vibrations generated by friction give it clues about what that surface is made of. Such sensitivity might help it establish the level of grip needed to pick up, say, a slippery porcelain plate.
http://www.newscientist.com/article/mg20627566.800-robots-with-skin-enter-our-touchyfeely-world.html
A Prototype Fingertip with High Spatial Resolution Pressure Sensing for the Robot iCub
This work is related to the design and realization of a fingertip which includes a capacitive pressure sensor with 12 sensitive zones. It is naturally shaped and its size is small enough so that it can be mounted on the fingers of the humanoid robot iCub. It also embeds the electronic device. it is 14.5 mm long and 13 mm wide (see Figure 7). The fingertip is made of compliant and deformable silicone patches whose capacitance varies when pressure is applied at the surface. The capacitors surround the inner core of the fingertip, which is mounted on a small printed circuit board.
Covering iCub, KASPAR and NAO with the Artificial Skin (IIT- UNIGE)
This work is related to the covering of three different robot iCub (EPFL), KASPAR and NAO, with Artificial skin. In particular for iCub we covered the palms, the lower arms, while for KASPAR we have covered the hands, the arms, the chest and the cheeks with 68 triangles that correspond to 816 contact points and we used 10 MTB boards. For NAO we have covering hands and upper arms.
iCub ( EPFL)
 |  |
|
 |  |  |
 |  |  |
 |  |  |
 |
KASPAR
 |  |  |  |
NAO
 |  |
WORK AT UNIGE:
This work deals with the problem of calibrating a large number of tactile elements (i.e., taxels) organized in a skin sensor system after fixing them to a robot body part. This problem has not received much attention in literature because of the lack of large-scale skin sensor systems. The proposed approach is based on a controlled compliance motion with respect to external objects whose pose is known, which
allows a robot to determine the location of its own taxels. The major contribution of this work is the formulation of the skin calibration problem as a maximum-likelihood mapping problem in a 6D space, where both the position and the orientation of each taxel are recovered. An effective calibration process is envisaged that, given a compliance control law that assures prolonged contact maintainance between a given body part and an external object, returns a maximum-likelihood estimate of detected taxel poses. Simulations validate the approach.
Work at UH:
Work at UH has commenced in two lines, Work Package 3 (WP3) activities relevant to skin and tactile based behaviour and WP5 activities relevant to skin based robot assisted play.
Under (WP3), we are developing new machine learning technologies for use with artificial skin sensors. One such technology is an advance in the self-organized generation of sensoritopic maps from subjective skin data only. It provides a process which allows a model of the robot's sensory world to emerge from the sensory stimulation emerging in the robot's interaction with the world. The new algorithm significantly improves on previous methods, and in simulations can reconstruct the shape of the robot's skin based purely on the internal incoming tactile data from the sensors without using external data about shape and structure of the sensors.
Early reconstruction images for the sensoritopic map. Above: actual sensor layout. Below: reconstructed layout.
WP5 highlights, in the context of the ROBOSKIN project, the important role that can be played by assistive technology enabled with tactile feedback capabilities. The UH team studies the possible use of the technology advancement of skin-based social interaction models, in the problem domain of autism therapy.
To help us built appropriate social tactile play scenarios, during the first project year a series of experimental investigations took place with children with autism interacting playfully with the robot KASPAR. KASPAR is a child-sized robot that was developed previously at the University of Hertfordshire which acts as a platform for Human-Robot-Interaction studies. KASPAR was equipped with temporary tactile sensors to try to capture the temporal and spatial characteristics of any tactile interactions that may occur. Interaction studies and sensor data collection are still ongoing. To illustrate the data, the figure presents recordings by left and right hand sensors when a child was instructed to touch both robot hands.
