Reactive grasping using optical proximity sensors

Abstract

We propose a system for improving grasping using fingertip optical proximity sensors that allows us to perform online grasp adjustments to an initial grasp point without requiring premature object contact or regrasping strategies. We present novel optical proximity sensors that fit inside the fingertips of a Barrett Hand, and demonstrate their use alongside a probabilistic model for robustly combining sensor readings and a hierarchical reactive controller for improving grasps online. This system can be used to complement existing grasp planning algorithms, or be used in more interactive settings where a human indicates the location of objects. Finally, we perform a series of experiments using a Barrett hand equipped with our sensors to grasp a variety of common objects with mixed geometries and surface textures.

Sensors Applied in Healthcare Environments

“…Optical infrared sensors are introduced for pre-touch during final grasp adjustments. The method in [37] detects the orientation of an object surface using the IR sensors that fit inside the fingers. However, [37] can only adjust the fingers and be used for one dimension of the end-effector.…”

Coupled dynamical system based arm–hand grasping model for learning fast adaptation strategies

“…Although both the issues of reaching to a pre-grasp pose and formation of grasp around arbitrary objects are intensively studied, very few [22][23][24] have looked into combining the two so as to have a unified reach-grasp system.…”

Iterative learning of grasp adaptation through human corrections

“…The limitation here therefore is the requirement of precise sensing, which is not available for many robot platforms. Data gathered through reactive grasping strategies is used to estimate the position and orientation of a novel object [27], to systematically gather information about the object shape [28], [29] and to infer areas where the fingertip might safely be moved to gather more sensor data [30]. Other approaches go further and gather shape information with the intent to build an explicit object model [31].…”

“…A grasping force is applied to counteract the perturbing force that results from object manipulation by the robot [8], and force feedback control is used for stabilizing the grasps during explicit finger repositioning for object rotation and translation behaviors [3]. Other approaches pair upper-level controllers that target grasp points with lower-level reactive controllers that avoid collisions [30], [2].…”