Reactive and pre-planned encirclement for whole arm grasping

Abstract

In this paper two novel methods of encircling objects for the purpose of grasping them are demonstrated and compared. Firstly, a method that does not require in-depth models or sophisticated external sensors is described. Such a method does have its disadvantages such as requiring tuning on the part of the operator to work correctly. Therefore, the paper also describes a new method that plans where the manipulator should contact the object such that the surface area of contact is maximised. It has been shown that it is possible for the pre-planned method to gain a grasp that is significantly better than the reactive method and overcomes the shortcomings of a particular robotic manipulator configuration.