Tactile-Proprioceptive Robotic Grasping

Abstract

Robotic grasping has been studied increasingly for a few decades. While progress has
\nbeen made in this field, robotic hands are still nowhere near the capability of human
\nhands. However, in the past few years, the increase in computational power and the
\navailability of commercial tactile sensors have made it easier to develop techniques that
\nexploit the feedback from the hand itself, the sense of touch. The focus of this thesis lies
\nin the use of this sense.
\nThe work described in this thesis focuses on robotic grasping from two different viewpoints:
\nrobotic systems and data-driven grasping. The robotic systems viewpoint describes
\na complete architecture for the act of grasping and, to a lesser extent, more
\ngeneral manipulation. Two central claims that the architecture was designed for are
\nhardware independence and the use of sensors during grasping. These properties enables
\nthe use of multiple different robotic platforms within the architecture.
\nSecondly, new data-driven methods are proposed that can be incorporated into the grasping
\nprocess. The first of these methods is a novel way of learning grasp stability from the
\ntactile and haptic feedback of the hand instead of analytically solving the stability from
\na set of known contacts between the hand and the object. By learning from the data
\ndirectly, there is no need to know the properties of the hand, such as kinematics, enabling
\nthe method to be utilized with complex hands. The second novel method, probabilistic
\ngrasping, combines the fields of tactile exploration and grasp planning. By employing
\nwell-known statistical methods and pre-existing knowledge of an object, object properties,
\nsuch as pose, can be inferred with related uncertainty. This uncertainty is utilized
\nby a grasp planning process which plans for stable grasps under the inferred uncertainty.