Non-Fixed Contact Manipulation Control Framework for Deformable Objects With Active Contact Adjustment

Huang, Jason; Cai, Yuanpei; Chu, Xiangyu; Taylor, Russell H.; Au, Kwok Wai Samuel

doi:10.1109/lra.2021.3062302

Cited by 11 publications

(5 citation statements)

References 26 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Sundaresan et al [17] presented a grasping point selection method by defining a disentangling hierarchy over cable crossings for robotic untangling of cable. Huang et al [18] proposed an active adjustment algorithm for soft tissue manipulation with non-fixed contact. The contact area can be adjusted by sliding the end effector, but the grasping point is fixed unless opening the grasper.…”

Section: Introductionmentioning

confidence: 99%

Incorporating Prior Knowledge into Reinforcement Learning for Soft Tissue Manipulation with Autonomous Grasping Point Selection

He¹,

Zhang²,

Yang³

et al. 2022

Preprint

View full text Add to dashboard Cite

Previous soft tissue manipulation studies assumed that the grasping point was known and the target deformation can be achieved. During the operation, the constraints are supposed to be constant, and there is no obstacles around the soft tissue. To go beyond these assumptions, a deep reinforcement learning framework with prior knowledge is proposed for soft tissue manipulation under unknown constraints, such as the force applied by fascia. The prior knowledge is represented through an intuitive manipulation strategy. As an action of the agent, a regulator factor is used to coordinate the intuitive approach and the deliberate network. A reward function is designed to balance the exploration and exploitation for large deformation. Successful simulation results verify that the proposed framework can manipulate the soft tissue while avoiding obstacles and adding new position constraints. Compared with the soft actor-critic (SAC) algorithm, the proposed framework can accelerate the training procedure and improve the generalization.

show abstract

Section: Introductionmentioning

confidence: 99%

Incorporating Prior Knowledge into Reinforcement Learning for Soft Tissue Manipulation with Autonomous Grasping Point Selection

He¹,

Zhang²,

Yang³

et al. 2022

Preprint

View full text Add to dashboard Cite

show abstract

“…In particular, actively deformable objects, such as animals, living organs (heart, blood vessels), pose significant challenges on the design of compliant grippers with large operating ranges and adjustable jaw clearance [6]. Compliant grippers equipped with force and displacement sensors are usually utilized to minimize the possibility of damaging the objects, resulting in higher cost and system complexity [7]- [9].…”

Section: Introductionmentioning

confidence: 99%

Design and Test of a Curved-Beam Based Compliant Gripper for Manipulations of Actively Deformable Objects

2022

View full text Add to dashboard Cite

Sizes and stiffness variations of actively deformable objects pose significant challenges on the design of compliant constant-force gripper. This paper presents a curved-beam based constant force compliant gripper which is composed of the constant force module, the bistable module, the preloading module and the linear guide. A curved-beam constant force mechanism is designed to generate constant force output, the non-constant force motion range of which is further eliminated via curved-based bistable mechanism and preloading module. After a formulation to find the optimal gripper configuration, the design is verified through comparison with simulation results. Finally, a prototype of the proposed gripper is tested to demonstrate its grasping capacity.

show abstract

“…Controlling the shape of deformable linear objects (DLOs) with robot manipulators has a wide range of industrial applications, such as cable routing [1], wire-harness assembly in manufacturing [2], or manipulation of endoscopes in robotic surgeries [3]. These shape control tasks pose a significant challenge due to the inherent mismatch between the finite constraints that can be imposed on DLOs by manipulators and the infinite degrees of freedom DLOs possess [4]. In tackling this challenge and striving to achieve complex shapes, additional constraints are required, typically provided by contacts from environmental fixtures [1], [5]- [9].…”

Section: Introductionmentioning

confidence: 99%

“…RELATED WORK The utilization of environmental fixtures to control the shape of DLOs was first introduced by Zhu et al [5]. Since then, there has been a widespread adoption of this approach in DLO manipulation [1], [4], [6]- [9]. A crucial step in this process is estimating the contact status of the DLO constrained by fixtures.…”

Section: Introductionmentioning

confidence: 99%

“…Finally, DLO manipulation primarily involves contact with only a small segment of the DLO, which makes focusing on the existence and position of contact points more suitable than creating a contact distribution. Therefore, previous works in manipulation of DLOs tend to adopt more intuitive and computationally efficient criteria to estimate the contact states, either by detecting the contact establishment and resulting shape changes from top images [4]- [8], or by detecting the contact position on the DLO from force measurements [1], [9].…”

Section: Introductionmentioning

confidence: 99%

See 1 more Smart Citation

Improved results on state estimation for switched continuous‐time linear systems

Chen

Wang

2021

Asian Journal of Control

View full text Add to dashboard Cite

This paper investigates the state estimation problem for switched linear systems involving additive disturbance and measurement noise, whose modes need not to be cooperative. Under the assumption that there exist nonsingular matrices that transform the observer error system matrices into Metzler ones, by using multiple Lyapunov functions and average dwell time method, hybrid state observers, which can produce a framer covering all possible state trajectories, are designed in two cases, that is, all the subsystems are stable and some of the subsystems may be unstable. The constructed interval observer for the considered systems are hybrid dynamical systems and the initialized states of the upper and lower observers at each switching instants are coupled, since the transformation matrices for different subsystems are different. Compared with the robust observer design for switched systems with arbitrary switching sequences in the literature, the difficulty consists in proposing new observer design conditions related to the state reset at the switching instants for switched systems under average dwell time scheme. Finally, two examples are given to show the efficiency of the proposed approach.

show abstract

Non-Fixed Contact Manipulation Control Framework for Deformable Objects With Active Contact Adjustment

Cited by 11 publications

References 26 publications

Incorporating Prior Knowledge into Reinforcement Learning for Soft Tissue Manipulation with Autonomous Grasping Point Selection

Incorporating Prior Knowledge into Reinforcement Learning for Soft Tissue Manipulation with Autonomous Grasping Point Selection

Design and Test of a Curved-Beam Based Compliant Gripper for Manipulations of Actively Deformable Objects

Improved results on state estimation for switched continuous‐time linear systems

Contact Info

Product

Resources

About