Experimental Evaluation of Precise Placement of the Hollow Object with Asymmetric Pivot Manipulation

JINSEONG PARK 1, JEONG-JUNG KIM 1 · 2025

Read Experimental Evaluation of Precise Placement of the Hollow Object with Asymmetric Pivot Manipulation by JINSEONG PARK 1, JEONG-JUNG KIM 1 (2025) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/arxiv-2504-05665.txt.

Introduction

Jinseong Park and co-authors show how a robot can pick up hollow objects by one finger slipped inside the hole. Humans grasp bottles and cups this way to place them precisely or feed them into machines. The method uses an asymmetric pivot motion: the arm tilts the object so one gripper finger enters the opening while the other presses the outside wall.

The team worked out the force conditions that make such a grasp hold, across objects of varied size and hole shape. It needs no special hardware, just a standard parallel-jaw gripper on a robot arm. Experiments achieved the grasp with a high success rate on bottles, cups, and ducts. Two factory-style tasks, aligning and feeding hollow cylinders, showed the practical payoffs of the technique.

Catalog record

CREATORJINSEONG PARK 1, JEONG-JUNG KIM 1
YEAR2025
COLLECTIONPAPERS
FORMATTXT
FILE SIZE44 KB
VERSION1.0
LICENSEhttp://creativecommons.org/licenses/by/4.0/
RIGHTSCC BY
SOURCEarXiv

Source: https://arxiv.org/abs/2504-05665

Browse the full library · About the Tower of Babel

Full text

The opening of the work, from the library's hosted text:

Experimental Evaluation of Precise Placement of the Hollow Object with Asymmetric Pivot Manipulation JINSEONG PARK 1, JEONG-JUNG KIM 1 Abstract In this paper, we present asymmetric pivot manipulation for picking up rigid hollow objects to achieve a hole grasp. The pivot motion, executed by a position-controlled robotic arm, enables the gripper to effectively grasp hollow objects placed horizontally such that one gripper finger is positioned inside the object’s hole, while the other contacts its outer surface along the length. Hole grasp is widely employed by humans to manipulate hollow objects, facilitating precise placement and enabling efficient subsequent operations, such as tightly packing objects into trays or accurately inserting them into narrow machine slots in manufacturing processes. Asymmetric pivoting for hole grasping is applicable to hollow objects of various sizes and hole shapes, including bottles, cups, and ducts. We investigate the variable parameters that satisfy the force balance conditions for successful grasping configurations. Our method can be implemented using a commercially available parallel-jaw gripper installed directly on a robot arm without modification. Experimental verification confirmed that hole grasp can be achieved using our proposed asymmetric pivot manipulation for various hollow objects, demonstrating a high success rate. Two use cases, namely aligning and feeding hollow cylindrical objects, were experimentally demonstrated on the testbed to clearly showcase the advantages of the hole grasp approach. † † history: Date of publication xxxx 00, 0000, date of current version xxxx 00, 0000. † † address: Department of Artificial Intelligence Machinery, Korea Institute of Machinery & Materials, Daejeon 34103, of Korea † † titlenote: This research was supported by the National Research Council of Science & Technology as part of the project entitled “Developed core technologies for the Robotic General Purpose Task Artificial Intelligence (RoGeTA) framework to enable a variety of everyday services” (NK254G) and the Industrial Strategic Technology Development Program (no. 20018745 and no. 00416440) funded by the Ministry of Trade, Industry & Energy (MOTIE, Korea). † † corresponding: Corresponding author: Jinseong Park (e-mail: [email protected]). I Introduction Figure 1: Entire manipulation process in the target scenario. Successful grasping of objects in desired poses is essential for precise robotic operations. This paper is concerned with the robotic grasping of objects that has various sizes and types of hole shapes, such as circle hole shapes (bushing, bottle, cup, or can), rectangular hole shapes (duct or mounting rail). The use of hole of the object can be more efficient for various tasks. Indeed, humans commonly use hole grasp for hollow object manipulation to get advantages in subsequent tasks such as transferring many objects in one hand, packing to tray with tight space or feeding to narrow machine slot as depicted in Fig. 1 . The target configuration of hole grasp is that one finger is positioned inside the object’s hole and the other contacts the outer surface, mimicking the way humans grasp hollow objects. The motivation for adopting a hole grasp approach arises from the need to enhance the feasibility of robotic automation in real-world manufacturing processes. In our previous research, we implemented the target scenario using a pushing primitive based on Cartesian force control [ 1 ] . To ensure precise placement into narrow machine slots, we utilized the slot’s flat surface as an external pusher, but feedback from machining process manager indicated that consistent impacts could compromise machine reliability and durability. In response to these concerns, we have embraced a hole grasp methodology that more closely mimics human motion. The most straightforward approach for achieving hole grasp is to lay the gripper on the ground to align with the object’s hole using a large range of arm motions and then grasp the hole directly; however, this method may require excessive free space for arm movements as illustrated in Fig. 13 (a). Sequential procedure with pinch grasping the lying object, placing it in the vertical direction and grasping it again with the hole can be another approach [ 2 ] . However, these conventional methods require regrasping, which increases tact time and is therefore undesirable for manufacturing automation [ 3 ] . The use of suction gripper is also not suitable for hole grasp since sufficient suction area cannot be guaranteed in the presence of the hole. On the other hand, in-hand manipulation with dexterity can be exploited. This method can be categorized based on whether it utilizes resources intrinsic to the hand (intrinsic dexterity) or extrinsic to the hand (extrinsic dexterity). From the perspective of intrinsic dexterity, in-hand manipulation with finely position- and force-controlled articulated hands has been developed to achieve the dexterity and flexibility of human hands [ 4 , 5 , 6 ] . Conversely, extrinsic dexterity with a simple gripper leverages external factors such as gravity, contact forces, and dynamic arm motions to perform push, sliding, and pivot primitives [ 7 ] . For practical applications, employing a simplified end effector can provide a more effective and appropriate solution than a gripper with complex mechanisms. Pivot manipulation has been actively explored through various approaches and perspectives. In essence, a passive end effector can be utilized to achieve pivot manipulation by leveraging gravity [ 8 ] . An open-loop robotic arm trajectory was found to rotate an object using a pinch grasp without considering form/force closure consideration [ 9 ] . Pivoting by prehensile pushing against the environment based on arm motion planning and a frictional contact model has been formulated [ 10 ] . A feedback control based pivoting grasp has also been studied. The pivoting of the object was achieved by using gravity and adaptive control of the gripping forces with vision/tactile [ 11 ] or force/tactile [ 12 ] feedback. Swing-up regrasping has been implemented by means of dynamic arm motions based on the energy control approach [ 13 ] . The non-prehensile pivoting utilizing contacts of environment was implemented with hybrid force-velocity control [ 17 ] . Sequential pivoting operation was presented to move an heavy object on the floor as if it walking [ 14 ] . Learning based pivoting grasp also has been implemented. A deep predictive model with tactile sensing was designed for the active slip control of an end effector in the presence of gravity or acceleration [ 15 ] . Q-learning based open loop pivoting was performed using inertial forces generated from the robot arm motion [ 16 ] . Implementation of pivot manipulation from a mechanical customization perspective has also been conducted. A mechanically improved parallel-jaw gripper has been developed to gain advantages in pivot-like grasping. A gripper consisting of one thumb and one finger based on epicyclic gear train was designed to achieve scooping without robot arm motion [ 18 ] . A gripper equipped with sheet embedded belt enabled passively pull-in operation during closing the gripper to flip the object [ 19 ] . A fingertip with pneumatic braking mechanism realized pivoting a grasped object with aligning in gravity direction [ 20 ] . However, most of pivoting manipulations have been realized by symmetric pinch grasping or a parallel-jaw gripper customization and, targeted in rigid and solid objects. The applications for thin objects grasping is more relevant to our purpose from the perspective of the asymmetric grasping configuration and the use of

Read the complete text