Logo
ScienceStack
Table of Contents
Fetching ...
Sign In

SoGraB: A Visual Method for Soft Grasping Benchmarking and Evaluation

Benjamin G. Greenland, Josh Pinskier, Xing Wang, Daniel Nguyen, Ge Shi, Tirthankar Bandyopadhyay, Jen Jen Chung, David Howard

TL;DR

The paper tackles the absence of standardized evaluation for soft grippers and deformation-sensitive grasp quality. It introduces SoGraB, a visual benchmarking method that quantifies object deformation during grasping using the Density-Aware Chamfer Distance ($DCD$) between pre- and post-grasp point clouds, supplemented by grasp success and holding time into a single score. The methodology is demonstrated with Fin-Ray soft fingers of varying stiffness across a diverse object set, producing a 900-grasp baseline dataset that reveals clear regimes where soft gripping offers advantages and where it does not. SoGraB enables objective, deformation-aware comparisons of soft-gripper designs without requiring specialized instrumentation, and provides a scalable platform for future research and dataset expansion. The approach advances practical soft-gripper development by linking deformation, safety, and grasp quality in a repeatable benchmark.

Abstract

Recent years have seen soft robotic grippers gain increasing attention due to their ability to robustly grasp soft and fragile objects. However, a commonly available standardised evaluation protocol has not yet been developed to assess the performance of varying soft robotic gripper designs. This work introduces a novel protocol, the Soft Grasping Benchmarking and Evaluation (SoGraB) method, to evaluate grasping quality, which quantifies object deformation by using the Density-Aware Chamfer Distance (DCD) between point clouds of soft objects before and after grasping. We validated our protocol in extensive experiments, which involved ranking three Fin-Ray gripper designs with a subset of the EGAD object dataset. The protocol appropriately ranked grippers based on object deformation information, validating the method's ability to select soft grippers for complex grasping tasks and benchmark them for comparison against future designs.

SoGraB: A Visual Method for Soft Grasping Benchmarking and Evaluation

TL;DR

The paper tackles the absence of standardized evaluation for soft grippers and deformation-sensitive grasp quality. It introduces SoGraB, a visual benchmarking method that quantifies object deformation during grasping using the Density-Aware Chamfer Distance () between pre- and post-grasp point clouds, supplemented by grasp success and holding time into a single score. The methodology is demonstrated with Fin-Ray soft fingers of varying stiffness across a diverse object set, producing a 900-grasp baseline dataset that reveals clear regimes where soft gripping offers advantages and where it does not. SoGraB enables objective, deformation-aware comparisons of soft-gripper designs without requiring specialized instrumentation, and provides a scalable platform for future research and dataset expansion. The approach advances practical soft-gripper development by linking deformation, safety, and grasp quality in a repeatable benchmark.

Abstract

Recent years have seen soft robotic grippers gain increasing attention due to their ability to robustly grasp soft and fragile objects. However, a commonly available standardised evaluation protocol has not yet been developed to assess the performance of varying soft robotic gripper designs. This work introduces a novel protocol, the Soft Grasping Benchmarking and Evaluation (SoGraB) method, to evaluate grasping quality, which quantifies object deformation by using the Density-Aware Chamfer Distance (DCD) between point clouds of soft objects before and after grasping. We validated our protocol in extensive experiments, which involved ranking three Fin-Ray gripper designs with a subset of the EGAD object dataset. The protocol appropriately ranked grippers based on object deformation information, validating the method's ability to select soft grippers for complex grasping tasks and benchmark them for comparison against future designs.

Paper Structure

This paper contains 15 sections, 2 equations, 7 figures.

Table of Contents

  1. Introduction
  2. Related Work
  3. Grasp Evaluation
  4. Evaluation Object Dataset
  5. SoGraB Methodology
  6. Grasp Quality Assessment
  7. Point Cloud Alignment
  8. Experimental Grasp Evaluation Implementation
  9. Experimentally Evaluated Grippers
  10. Evaluation Object Selection
  11. Experimental Results
  12. Relatively Stiff Objects
  13. Relatively Soft
  14. Just Right
  15. Discussion and Conclusion

Figures (7)

  • Figure 1: The SoGraB method: (Left) A soft, Shore 40A object grasped by a soft Fin-Ray gripper. (Right) The extracted point clouds showing initial state (black) and deformed state during grasping (red). By comparing the two states as well as the grasp success, we produce a grasp quality benchmark.
  • Figure 2: Demonstration of ICP point cloud alignment (Shore 40A EGAD B1 object). (a) Initial position of point clouds, transformed into the same coordinates in the end-effector frame. (b) After ICP alignment.
  • Figure 3: (a) Fin-Ray soft fingers with 4, 6 and 8 ribs. (b) Experimental grasp evaluation platform. Note: Grippers are rotated 90° in yaw prior to grasping
  • Figure 4: Evaluation objects used in this study: B1-F5 are selected objects from the EGAD dataset. O1-O3 are custom evaluation objects designed to be symmetric and highly deformable.
  • Figure 5: Heatmaps of the complete grasp evaluation dataset, showing the mean and standard deviations of each set of grasps.
  • ...and 2 more figures