Compliant Suction Gripper With Seamless Deployment and Retraction for Robust Picking Against Depth and Tilt Errors

Yuna Yoo1*, Jaemin Eom1*, MinJo Park1, Kyu-Jin Cho1
1Biorobotics Laboratory, Seoul National University
(* Yuna Yoo and Jaemin Eom contributed equally to this work.)

Abstract

Applying suction grippers in unstructured environments is a challenging task because of depth and tilt errors in vision systems, requiring additional costs in elaborate sensing and control. To reduce additional costs, suction grippers with compliant bodies or mechanisms have been proposed; however, their bulkiness and limited allowable error hinder their use in complex environments with large errors. Here, we propose a compact suction gripper that can pick objects over a wide range of distances and tilt angles without elaborate sensing and control. The spring-inserted gripper body deploys and conforms to distant and tilted objects until the suction cup completely seals with the object and retracts immediately after, while holding the object. This seamless deployment and retraction are enabled by connecting the gripper body and suction cup to the same vacuum source, which couples the vacuum picking and retraction of the gripper body. Experimental results validated that the proposed gripper can pick objects within 79 mm, which is 1.4 times the initial length, and can pick objects with tilt angles up to 60°. The feasibility of the gripper was verified by demonstrations, including picking objects of different heights from the same picking height and the bin picking of transparent objects.

Video

This video explains the research motivation, the mechanism design of the proposed gripper, and its applications.

Picking Distant and Tilted Objects

The proposed gripper has a deployable body, allowing it to pick distant and tilted objects. Additionally, with its pneumatic circuit design, once the suction cup makes contact and seals with the object, it can automatically pick up the object without the need for additional contact sensing. This design enhances picking robustness against vision sensing errors.

Picking Objects of Different Heights

The proposed gripper picks up objects of different heights from the same picking height. The gripper eleiminates the need for depth sensing to pick up different heights of objects.


Bin Picking of Transparent Objects

Perceiving transparent objects is a challenge in vision sensing. The proposed gripper can easily perform bin picking tasks of transparent dishes even in the presence of depth or tilt angle errors.

Depalletizing

By arranging multiple grippers in parallel, the grippers can pick up multiple objects at different heights at once. This feature enables depalletizing when the number of grippers and the number of the objects arranged on each layer do not match.


Warehouse Picking

Extending the length of the gripper enables warehous picking in occlude environments. It has 140 mm stroke and can penetrate a narrow, deep space that cannot be reached by a manipulator.

Dynamic Picking

The proposed gripper with 79 mm stroke was applied to catch a falling balloon at a speed of 1.02 m/s at a distance of 50.1 mm, demonstrating the possibility of dynamic grasping.


Motivation

Is there a way to achieve robust suction picking despite vision sensing errors?

Animated GIF

Depth and tilt errors in vision sensing often occur, especially when handling transparent objects. To address this issue, previous approaches have attempted to reduce sensing errors through learning-based methods or by incorporating additional sensing (e.g., proximity sensors). However, learning-based methods still have insufficient success rates, and using additional sensors increases cost and may hinder the speed of object picking.


Our Solution

Can we achieve robust picking against depth errors using only mechanical design, without additional sensors or learning?

By relying solely on mechanical design, we could potentially reduce costs and enable compatibility with existing suction-based actuation systems. To achieve this, we need to consider the following approaches:

  1. Designing a suction cup body that can seal objects effectively, even in the presence of depth and tilt errors.
    Deployable & Compliant Body

  2. Proposing an actuation system (pneumatic circuit design) that enables automatic picking without additional sensing or control once the suction cup seals the object.
    Pneumatic Circuit Design


Deployable, Compliant Gripper Body

Soft Deploying Mechanism for Suction Cups

Animated GIF

The proposed gripper comprised two concentric cylindrical chambers, where the outer and inner chambers are termed as the spring-inserted gripper body and spring-inserted air tube, respectively. The gripper body retracts and deploys when negative pressure is applied or released, and the air tube provides the airway to the suction cup.

The Gripper Body Conformable to Distant and Tilted Surfaces

Animated GIF

The gripper body was designed as a compressible spring enclosed in a cylindrical LDPE film for the following reasons: First, the gripper can be deployed at a high speed due to the restoring force of the spring. Second, the gripper can conform to tilted objects during deployment. Finally, the radially incompressible characteristics of the spring allow the gripper to retract effectively in the longitudinal direction.

The spring-inserted air tube which connects the suction cup and valve was built inside the gripper body with a compressible spring encased in a cylindrical LDPE film with the same structure as the gripper body. The radial incompressibility of the spring prevents airway clogging during the retraction.


Pneumatic Circuit Design

Connect Gripper Body and Suction Cup to the Same Vacuum Source.

Enabling seamless deployment, pick, and retraction without sensing and control by coupling the vacuum picking and retraction of the gripper body.

  • Pick up distant & tilted object without sensing and control.

BibTeX

@article{yoo2023compliant,
  title={Compliant suction gripper with seamless deployment and retraction for robust picking against depth and tilt errors},
  author={Yoo, Yuna and Eom, Jaemin and Park, MinJo and Cho, Kyu-Jin},
  journal={IEEE Robotics and Automation Letters},
  volume={8},
  number={3},
  pages={1311--1318},
  year={2023},
  publisher={IEEE}
}