Underactuated Robotic Hands for Grasping in Warehouses

Abstract

The automation of order picking in warehouses requires adaptive robotic hands that are cheap, robust, easy to control, and capable to reliably grasp a large range of products. The state-of-the-art graspers or robotic hands cannot fulfil this need. The goal of this research is to design and evaluate an adaptive and simple hand by applying the concept of underactuation (i.e. having fewer actuators than independently moving fingers). The innovation is that this hand mechanically decides to firmly envelope large objects, and to grasp small objects between the tips of the fingers. Experiments with the new hand show that the full range of object sizes that appear in the workstation of a reference warehouse can be grasped with a minimum control effort and electronic parts (i.e. one motor and no sensors). It is concluded that this new, underactuated hand is applicable for automated order picking.