Dexterous Gripper Research Articles

Physics-aware iterative learning and prediction of saliency map for bimanual grasp planning

Learning the skill of human bimanual grasping can extend the capabilities of robotic systems when grasping large or heavy objects. However, it requires a much larger search space for grasp points than single-hand grasping and numerous bimanual grasping annotations for network learning, making both data-driven or analytical grasping methods inefficient and insufficient. We propose a framework for bimanual grasp saliency learning that aims to predict the contact points for bimanual grasping based on existing human single-handed grasping data. We learn saliency corresponding vectors through minimal bimanual contact annotations that establishes correspondences between grasp positions of both hands, capable of eliminating the need for training a large-scale bimanual grasp dataset. The existing single-handed grasp saliency value serves as the initial value for bimanual grasp saliency, and we learn a saliency adjusted score that adds the initial value to obtain the final bimanual grasp saliency value, capable of predicting preferred bimanual grasp positions from single-handed grasp saliency. We also introduce a physics-balance loss function and a physics-aware refinement module that enables physical grasp balance, capable of enhancing the generalization of unknown objects. Comprehensive experiments in simulation and comparisons on dexterous grippers have demonstrated that our method can achieve balanced bimanual grasping effectively.

A Variable Stiffness Gripper with Reconfigurable Finger Joint for Versatile Manipulations.

A reconfigurable dexterous gripper is designed which can switch states, including rigidity and flexibility, for different application scenarios. Moreover, the stiffness of the fingers in the flexible state can also be tuned for different objects. Three fingers are connected to the revolute joints of the palm, and each finger has a reshape mechanism with a slider moving up and down to lock or release the fingertip joint. When the slider moves upward, the gripper works in the rigid state and the fingers are actuated by the servos. When the slider moves downward, the gripper works in the flexible state that the fingertip is supported by a spring, and the fingertip joint is rotated by an embedded motor with two group cables for tuning stiffness. This novel design provides the gripper with the advantages of high precision and strong load capacity of rigid grippers and shape adaptability and safety of soft grippers. The reconfigurable mechanism allows the gripper great versatility for grasping and manipulation, which facilitates the planning and execution of the motion of objects with different shapes and stiffness. We discuss the stiffness-tunable mechanism with different states, analyze the kinematic characteristics, and test the manipulator performance to investigate the application in rigid-flexible collaborative works. Experimental results show the practicability of this gripper under different requirements and the rationality of this proposed concept.

On the perception and handling of deformable objects – A robotic cell for white goods industry

• Design and implementation of a robotic system for the manipulation and assembly of linear non-rigid components. • Multi-sensor system for flexible part cognition and safe human-robot interaction. • Dexterous gripper for non-rigid material manipulation. • Control framework for process coordination and safe human -robot collaboration. • Validation and testing in a white goods industry case study. Advancements in mechatronics, perception and computational methods allowed robotics to focus on non-rigid part manipulation, which is characterized by high complexity due to the unpredictable and compliant behavior of flexible materials. This paper presents the automation of a manufacturing process involving linear non-rigid components through the implementation of a robotic system that addresses challenges related to flexible material perception and handling. The corresponding technological approach consists of five steps namely safe material supply, detection, grasping, cross section recognition, and assembly. The proposed robotic cell comprises: (1) a multi-sensor system for flexible part recognition and safe human-robot interaction (2) a collaborative robot equipped with a dexterous gripper for flexible material manipulation, and (3) a control framework for process coordination. The robotic system has been tested on a case study originating from the white goods industry, where rubber gaskets are manipulated and assembled on household appliances. Experimental results proved that the implemented robotic system can tackle the challenges of flexible material manipulation with great consistency in the aspects of robustness and repeatability.

Robotic Multifinger Grasping State Recognition Based on Adaptive Multikernel Dictionary Learning

Traditional grasping analysis of mechanical dexterous grippers tends to flatten a multi-finger tactile series into one dimension, which ignores the force coupling between fingers and their different grasping force characteristics. To overcome this problem, this work proposes a novel Adaptive Multi-kernel Dictionary Learning (AMDL) method. First, in order to capture the nonlinear feature similarity of different tactile samples, multiple basic kernel functions are used to map all the training samples into Hilbert space, and the corresponding kernel matrix of each basic kernel is computed respectively. Then, an adaptive kernel weight calculation method is developed to learn the adaptive kernel of each basic kernel. A composite kernel, which is the linear combination of multiple basic kernels by using the learned adaptive weights, is constructed to calculate the multi-dimensional kernel matrix. Finally, this work utilizes the proposed AMDL to fuse grasping tactile information of multiple fingers to further consider the force coupling among them, during which the sparse pattern of the coding vector of each finger’s tactile data is restricted to be consistent. The proposed algorithm is compared with other state-of-the-art algorithms in terms of F1 score on the public BioTac SP tactile dataset and our collected tactile dataset. Its grasping state recognition result shows its validity and feasibility.

Multifunctional mandibles of ants: Variation in gripping behavior facilitated by specific microstructures and kinematics

The elongated mandibles of certain ant species are dexterous grippers that can output a wide range of forces as needed for various tasks. Our combined experimental and theoretical research reveals the multifunctionality of the mandibles of Harpegnathos venator that is facilitated by specific microstructures and characteristic kinematics. First, we found that H. venator can pull off a spider’s (Heteropoda venatoria) leg by closing its long mandibles. We observed that the ant usually clamps the spider’s leg using the distal or middle part of its mandibles. In contrast, the ant can grip its egg with the proximal parts of its mandibles without causing damage. Our results showed that the spider’s legs are always fractured at the coxa-trochanteral joint. Second, we found that the force required to fracture the spider’s leg can be up to 500 times the ant’s body weight. On the other hand, the maximum force can be controlled to less than 2×10- 6N while gripping an egg. By combining microstructure imaging, kinematic tracking and mathematical modeling, we uncovered that the sharp teeth and dense bristles on the internal side of the mandibles determine the high adhesion force, while the concave teeth and biaxial rotation of the mandibles facilitate gentle gripping. We validated our findings by constructing an artificial mandible pair. This work expands the knowledge of the physiological multifunctionality in ant mandibles, and provides novel ways to reveal the multifunctionality in insect appendages by applying the tools of mechanical analysis and related experimental devices.

A soft pneumatic dexterous gripper with convertible grasping modes

This paper proposes a novel soft pneumatic dexterous gripper that consists of four soft fingers and a movable sucker. Compared with rigid dexterous hands and other soft grippers, the soft pneumatic dexterous gripper has four convertible grasping modes, which can grasp a wide range of objects due to the proposed conversion mechanism. Most parts of the dexterous gripper are 3D-printed with resin in order to make it lighter. The soft finger is made of silicone rubber with the pleated channel, making the grasping process more flexible and safer. The curvature performance of single finger is predicted through finite element analysis (FEA) before making a prototype. The simplified kinematic model and tip workspace of the gripper are obtained to reveal its grasping range. Then, a mathematical model of the soft bending finger is proposed in order to explain the relationship between input air pressure and the finger bending angle. Finally, the single finger bending and the gripper grasping experiments results validate the accuracy of the model, consistency of finite element analysis, and rationality of the design.

Dexterous Gripper for In-Hand Manipulation with Embedded Object Localization Algorithm

Since the last decade, thanks to the spreading of the concept of Industry 4.0 and Smart Factory, more and more companies have started to investigate the robotic field looking for reliable solutions aiming at improving the efficiency of assembly lines.Promising technologies are connected to the speeding up of production stages like fast algorithms for object localization, as well as dexterous grippers for manipulation and assembly. Nowadays, most of the solutions for pick and place tasks involve the use of robotic grippers for grasping objects, while robotic manipulators are responsible for their accurate placements. Focusing on the grippers, although their simple structure can be appreciated, it greatly reduces their in-hand manipulation abilities, making unfeasible the twists of grasped objects and their release in a desired pose. As consequence, the efficiency of the pick and place operation is reduced since several adjustments of the robotic arm are required to accomplish the task.In this paper, a novel dexterous gripper coupled with a vision system algorithm for object localization and pose estimation are presented, and their performances in manipulating different objects are discussed. The designed gripper has a symmetrical structure with two finger modules, each one consisting in a couple of linear actuators arranged mutually orthogonal, so the translations in two axis, namely y and z directions, are allowed. As terminal part of each finger there is a revolute joint to whom is attached a fingertip modelled according to the shape of the target objects and easily replaceable. The embedded vision system algorithm adopted estimates position and orientation of the objects on a flat surface, and it coordinates the gripper placement to grasp them. The case study of the handling of a Spanish fan is presented and discussed in details.

Development of Efficient Strategy for Square Peg-in-Hole Assembly Task

In this paper, we present a dexterous gripper with an angle measurement system and efficient strategy for estimating a hole position from reaction force measurements in a square peg-in-hole task. The contact conditions between a square peg and hole have been analyzed to show that the redundancy in reaction force-moment relation around hole originated from contact conditions. In order to remove such redundancy in contact conditions, an angular error between a square peg and hole is first corrected with the help of the angle measurement system. Then, the positional error between them is identified by applying two clustering algorithms (K means algorithm and Gaussian Mixture Model (GMM) algorithm) to the reaction moment data obtained by the XY search. For grasping and manipulating a peg precisely, a 4-DOF dexterous gripper is designed, which consists of two fingers equipped with 6-axis force/torque sensor at fingertip. The extensive experiments using the proposed gripper with the angle measurement system demonstrate that during the square peg-in-hole task, the proposed strategy ensures the tact time of around 5 sec with the estimation accuracy of less than 0.35 mm and the deviation of less than 0.44 mm, respectively.

Current Control and Force Control in the Drives of Robot Grippers

Currently, the task of improving the robot grippers and their force control algorithms is urgent, thanks to the development of such areas of robotics as medical and rehabilitation robotics, prosthetics, collaborative robotics and other areas where there is the task of limiting or minimizing the loads acting on the object of manipulation. The force control in the electrically driven dexterous gripper is ultimately provided by the current control. Therefore, the purpose of this article is to analyze the operation of various current regulators in the context of their application in dexterous gripper and manipulators with a force control. The article considers the following current controllers: proportional-integral, adaptive, hysteretic and hysteretic with adjusting loop. Among these controllers, the hysteresis controller with adjusting loop is promising for the force control gripper and manipulator, since it provides switching frequency stabilization and current ripples reduction in a motor driver while maintaining fast response and sufficient robustness. For the hysteresis current controller with adjusting loop, reduced and linearized mathematical models of the adjusting loop are proposed. Moreover, based on the linearized model, methodic is constructed for the adjusting loop synthesis, which assumes the using of the classical control theory. Based on the synthesized current controllers, the analysis of the two-fingered gripper control system carried out, taking into account such factors as the force control accuracy, switching frequency stability and current ripples in the motor driver. Comparison results of current controllers can be used for the rational choice of a current controller providing force control, and the proposed synthesis methodic of the current controller with adjusting loop can to provide the controller generation with given control quality parameters.

In-Hand Grasping Pose Estimation Using Particle Filters in Combination with Haptic Rendering Models

Specialized grippers used in the industry are often restricted to specific tasks and objects. However, with the development of dexterous grippers, such as humanoid hands, in-hand pose estimation becomes crucial for successful manipulations, since objects will change their pose during and after the grasping process. In this paper, we present a gripping system and describe a new pose estimation algorithm based on tactile sensory information in combination with haptic rendering models (HRMs). We use a 3-finger manipulator equipped with tactile force sensing elements. A particle filter processes the tactile measurements from these sensor elements to estimate the grasp pose of an object. The algorithm evaluates hypotheses of grasp poses by comparing tactile measurements and expected tactile information from CAD-based haptic renderings, where distance values between the sensor and 3D-model are converted to forces. Our approach compares the force distribution instead of absolute forces or distance values of each taxel. The haptic rendering models of the objects allow us to estimate the pose of soft deformable objects. In comparison to mesh-based approaches, our algorithm reduces the calculation complexity and recognizes ambiguous and geometrically impossible solutions.

Active Soft End Effectors for Efficient Grasping and Safe Handling

The end effector is a major part of a robot system and it defines the task the robot can perform. However, typically, a gripper is suited to grasping only a single or relatively small number of different objects. Dexterous grippers offer greater grasping ability but they are often very expensive, difficult to control and are insufficiently robust for industrial operation. This paper explores the principles of soft robotics and the design of low-cost grippers able to grasp a broad range of objects without the need for complex control schemes. Two different soft end effectors have been designed and built and their physical structure, characteristics, and operational performances have been analyzed. The soft grippers deform and conform to the object being grasped, meaning they are simple to control and minimal grasp planning is required. The soft nature of the grippers also makes them better suited to handling fragile and delicate objects than a traditional rigid gripper.

Design of a Variable Stiffness Soft Dexterous Gripper.

This article presents the design of a variable stiffness, soft, three-fingered dexterous gripper. The gripper uses two designs of McKibben muscles. Extensor muscles that increase in length when pressurized are used to form the fingers of the gripper. Contractor muscles that decrease in length when pressurized are then used to apply forces to the fingers through tendons, which cause flexion and extension of the fingers. The two types of muscles are arranged to act antagonistically and this means that by raising the pressure in all of the pneumatic muscles, the stiffness of the system can be increased without a resulting change in finger position. The article presents the design of the gripper, some basic kinematics to describe its function, and then experimental results demonstrating the ability to adjust the bending stiffness of the gripper's fingers. It has been demonstrated that the fingers' bending stiffness can be increased by more than 150%. The article concludes by demonstrating that the fingers can be closed loop position controlled and are able to track step and sinusoidal inputs.

Matlab Toolbox for Kinematic Analysis and Simulation of Dexterous Robotic Grippers

Robot simulators have supported the science of robotics for a long period and have been an indispensable tool for understanding, planning, designing and programming of educational and industrial robots. These soft tools allows the users to program and test their applications without using the real hardware or even building it since such tools allow the analysis of behaviours and performance beforehand. Some of these soft tools have already been designed and developed for professional as well as educational and research purposes by researchers. However, the majority of these tools are specifically designed for serial manipulators and simulation tools for robotic grasping is rather limited e.g Graspit!, OpenGRASP, SynGrasp, and hence there is a need to develop a general purpose software package for performing different analyses on dexterous robotic hands.In this research work, a new MATLAB based dexterous robotic gripper toolbox is developed with GUI based simulating environment for Kinematic analysis and additional features with inbuilt library of various dexterous grippers of Robot. This toolbox has capabilities like Hand modeling, Grasp definition, Grasp modeling, Grasp analysis and Graphics to support representation of various commercially available existing dexterous grippers and has provision and flexibility to accommodate any newly developed hand. The library of toolbox has different parameters like Denavit - Hartenberg parameters, finger size, and joint angles etc. which are used for kinematics of various dexterous grippers. Form closure analysis is a special feature that has been included for understanding the actual positioning and orientation of fingers of dexterous grippers according to the object position. The library of proposed toolbox has detailed definition of objects to make form closure by various models of dexterous grippers.

Kinematic Analysis of a Dexterous Hand

Handling of objects with irregular shapes and that of flexible/soft objects by ordinary robot grippers is difficult. Multi fingered gripper may be a solution to such handling tasks. However, dexterous grippers will be the appropriate solution to such problems. Although it is possible to develop robotic hands which can be very closely mapped to human hands, it is sometimes not to be done due to control, manufacturing and economic reasons. The present work aims at designing and developing a dexterous robotic hand for manipulation of objects.

Meso-Scale Piezoelectric Gripper with High Dexterity

Design, fabrication and testing of a piezoelectrically actuated meso-scale dexterous gripper are reported. The gripper utilizes a piezoelectric ringmorph, which is excited by means of quarterly separated electrodes yielding 3 degree of freedom (3-dof) motion. Utilizing a specially designed arm, the 3-dof ringmorph actuation is transformed into a gripper motion bearing high dexterity. The combination of in- and out-of-plane motion unique to the gripper generates the herein explained sliding grip release, which improves the pick-and-place performance by avoiding sticking of the substrate. The 27×20×7 mm3 gripper can generate 300 mg gripping force while it has 110 µm out-of-plane and 30 µm in-plane no-load maximum stroke at ±100 V DC. The characterization of the device, including hysteresis and transient response are implemented as a function of loading condition and change in field intensity. The experimental results and sample pick-and-place tests have shown prominent force, stroke and frequency bandwidth of the gripper with a novel releasing method.

Spatial grasp synthesis for complex objects using model‐based simulation

PurposeSelection of an effective grasp of a complex object using a multifingered gripper is a challenging problem because of the many possible grasp positions that are typically available.Design/methodology/approachGiven the geometrical description of the particular object feature to be grasped, all feasible grasps are performed in offline simulation using a geometrically accurate model of the desired gripper. The six‐dimensional convex hull for each grasp is computed and archived. This convex hull indicates the span of forces and torques that the grasp can resist. When a grasp is needed the force/torque due to the total object weight is estimated and the best grasp is selected. The selected grasp has minimum peak contact force consistent with equilibrium.FindingsExperimental trials with several complex object show the method is capable of producing grasps which can support the object and resist external force/torque.Research limitations/implicationsAn accurate geometrical description of the feature to be grasped must be known in advance. This would typically be a cylindrical or prismatic portion of the object.Practical implicationsThere are many environments in which a dexterous multifingered gripper must be used due to the variety of objects which must be grasped. The results indicate that effective grasps can be selected for complex objects from a database of simulated grasps.Originality/valueThe primary contribution of this paper is the use of a database of simulated grasps on simple graspable features to synthesize grasps on complex objects.

Vision-based action control system of a multi-finger mechanical gripper

A study about the action control of a dexterous mechanical gripper based on stereo-vision system was proposed. The vision-based system was used to replace the data-glove for gesture measurement. The stereo vision theory was applied to calculate the 3D information of the hand gesture. The information was used to generate the grasping action parameters of a 3-finger dexterous mechanical gripper. Combined with a force feedback device, a closed control loop could be constructed. The test for the precision of the algorithms and action control simulation result were shown in the paper.

A dexterous robotic gripper for autonomous grasping

In advanced robotics applications in unstructured environments (e.g. those foreseen in space) some degree of dexterity and autonomy is necessary in order to safely and successfully execute the required tasks. With this respect, besides the kinematic configuration, important aspects to be considered in the design of robotic end‐effectors are the sensorial equipment and proper control strategies. In this paper, an activity for designing and experimenting a gripper for this operation in unstructured environments is reported, and laboratory results are presented and discussed.

Dexterous Grippers: Putting Nonholonomy to Work for Fine Manipulation

In this paper, we describe the realization and control of robotic end-effectors that are designed to achieve high operational versatility with limited constructive complexity. The design of such end-effectors, which can be regarded either as low-complexity robot hands or as highly versatile robot grippers, is based on the intentional exploitation of nonholonomic effects that occur in rolling. While the potential usefulness of manipulation by rolling has been theoretically established in the literature, several problems in the practical implementation of the concept remained open. In particular, manipulation of parts of complex, and a priori unknown, shape is considered in this paper. Experimental low-complexity grippers that realize dexterous manipulation by rolling are also described.

Dexterous Grippers: Putting Nonholonomy to Work for Fine Manipulation

Dexterous Grippers: Putting Nonholonomy to Work for Fine Manipulation