Automatic Fingerpad Customization for Precise and Stable Grasping of 3D-Print Parts

TL;DR

A fast, end-to-end approach to customize rigid gripper fingerpads that could achieve precise and stable grasping for different objects at multiple grasp points and experimentally demonstrates the validity of the approach by synthesizing fingerpads that, once mounted on a physical robot gripper, are able to grasp different objects at multiple grasp points.

Abstract

The rise in additive manufacturing comes with unique opportunities and challenges. Massive part customization and rapid design changes are made possible with additive manufacturing, however, manufacturing industries that desire the implementation of robotics automation to improve production efficiency could face challenges in the gripper design and grasp planning due to highly complex geometrical shapes resulting from massive part customization. Yet, current gripper design for such objects are often manual and rely on ad-hoc design intuition. This would be limiting as such grippers would lack the ability to grasp different objects or grasp points, which is important for practical implementations. Hence, we introduce a fast, end-to-end approach to customize rigid gripper fingerpads that could achieve precise and stable grasping for different objects at multiple grasp points. Our approach relies on two key components: (i) a method based on set Boolean operations, e.g. intersections, subtractions, and unions to extract object features and synthesize gripper surfaces that conform to different local shapes to form caging grasps; (ii) a method to evaluate the grasp quality of synthesized grippers. We experimentally demonstrate the validity of our approach by synthesizing fingerpads that, once mounted on a physical robot gripper, are able to grasp different objects at multiple grasp points, all with tightly constrained grasps.

Figures (9)

• Figure 1: Fingerpad Customization with Set Operators (FCSO): A single pair of fingerpads is capable of tightly grasping different objects at multiple poses per object. The figure shows a pair of fingerpads that have been designed by FCSO to conform optimally and simultaneously to the geometries of four grasped surfaces (2 objects $\times$ 2 poses per object) to form caging grasps. Physical grasping experiments are available at https://youtu.be/M68YagfUF1g
• Figure 2: Proposed pipeline for FCSO. It accepts the STL files of the objects, user-defined parameters from a configuration file, and the flat finger model of a gripper, to automatically return the best grasp surfaces and the best gripper design.
• Figure 3: Grasp sampling by a sliding pair of rectangular samples $(S)$ along the lateral axis of an object, with a stride equivalent to $L$. Each sample pair has the same color code.
• Figure 4: Fingerpad customization (without filter) based on the number of geometries $(N)$, while illustrating a three-step procedure on a pair of fingerpads. (a) Independent Boolean intersections $(I_n)$ resulting from the intersection of every valid rectangular sample $(S)$ and $G_n$, which is the $n^{th}$ geometry of the mesh bounded by the $S$. The samples are obtained from the grasp sampler (Section \ref{['ssec:graspsampler']}); (b) Boolean union of $N$ intersections $(M_N)$; (c) Boolean subtraction of $S$ and $M_N$ to obtain fingerpad $(P)$ that has a shape which conforms to the mesh at all $G_n$.
• Figure 5: Comparing effects of the filter with good and bad geometries. (a) Without filter: Undesirable $P$, in yellow, obtained in the presence of a single bad geometry. This shows the need of a filter to differentiate between geometries; (b) With filter: Visible improved performance. Illustrating three possible cases discussed in Section \ref{['ssec:withfilter']}, with $d_1>0, d_2>0, d_3=d_4=0$. In Example C, $d_B=min(d_1,d_2)*K$, whereas in Example D, $d_B=d_2*K$.