Grasp Quality Research Articles

Integration of Deep Q-Learning with a Grasp Quality Network for Robot Grasping in Cluttered Environments

During the movement of a robotic arm, collisions can easily occur if the arm directly grasps at multiple tightly stacked objects, thereby leading to grasp failures or machine damage. Grasp success can be improved through the rearrangement or movement of objects to clear space for grasping. This paper presents a high-performance deep Q-learning framework that can help robotic arms to learn synchronized push and grasp tasks. In this framework, a grasp quality network is used for precisely identifying stable grasp positions on objects to expedite model convergence and solve the problem of sparse rewards caused during training because of grasp failures. Furthermore, a novel reward function is proposed for effectively evaluating whether a pushing action is effective. The proposed framework achieved grasp success rates of 92% and 89% in simulations and real-world experiments, respectively. Furthermore, only 200 training steps were required to achieve a grasp success rate of 80%, which indicates the suitability of the proposed framework for rapid deployment in industrial settings.

Automated Grasp Planning and Finger Design Space Search Using Multiple Grasp Quality Measures

As the industry shifts to automated manufacturing and the assembly of parts in smaller batches, there is a clear need for an efficient design of grippers. This paper presents a method for automated grasp planning and finger design for multiple parts using four grasp quality measures that capture the following important requirements for grasping: (i) uniform contact force distribution; (ii) better gravity wrench resistance; (iii) robustness against gripper positioning error; and (iv) ability to resist larger external wrench on the object. We introduce the fingertip score to quantify the grasp performance of a fingertip design over all the objects. The method takes the CAD model of the objects as the input and outputs the optimal grasp location and the best finger design. We use the method for a three-point grasp with a parallel jaw gripper. We validate our method on two sets of objects. Results show how each grasp quality measure behaves on different objects and the variation in the fingertip score with finger design. Finally, we test the effectiveness of the optimal finger design experimentally. The three-point grasp is suitable for grasping objects larger than is possible with shape-matching fingertips.

Self-Supervised Antipodal Grasp Learning With Fine-Grained Grasp Quality Feedback in Clutter

It is a challenging goal in robotics to make a robot grasp like a human being in a cluttered environment. Self-supervised grasp learning is one of the most promising approaches to human-like robotic grasp. However, due to inadequate feedback on grasp quality, almost existing self-supervised grasp learning methods are coarse-grained. This paper presents a fine-grained antipodal grasp learning (FAGL) method with augmented learning feedback. First, an indicator called antipodal degree of a grasp (ADG) is designed as a non-increasing monotonous function for the fine-grained evaluation of grasp quality according to the disturbance incurred by a grasp to the surroundings in the image space. Next, we design a restorative sampling strategy to collect antipodal grasp samples and propose a refined affordance network to generate grasp affordance maps for FAGL's decision of grasp policies. Finally, in grasping actual metal workpieces, FAGL outperforms its peers in grasp success rate and ADG in cluttered scenarios by reducing the grasp effects on the surroundings.

Learning Suction Graspability Considering Grasp Quality and Robot Reachability for Bin-Picking.

Deep learning has been widely used for inferring robust grasps. Although human-labeled RGB-D datasets were initially used to learn grasp configurations, preparation of this kind of large dataset is expensive. To address this problem, images were generated by a physical simulator, and a physically inspired model (e.g., a contact model between a suction vacuum cup and object) was used as a grasp quality evaluation metric to annotate the synthesized images. However, this kind of contact model is complicated and requires parameter identification by experiments to ensure real world performance. In addition, previous studies have not considered manipulator reachability such as when a grasp configuration with high grasp quality is unable to reach the target due to collisions or the physical limitations of the robot. In this study, we propose an intuitive geometric analytic-based grasp quality evaluation metric. We further incorporate a reachability evaluation metric. We annotate the pixel-wise grasp quality and reachability by the proposed evaluation metric on synthesized images in a simulator to train an auto-encoder–decoder called suction graspability U-Net++ (SG-U-Net++). Experiment results show that our intuitive grasp quality evaluation metric is competitive with a physically-inspired metric. Learning the reachability helps to reduce motion planning computation time by removing obviously unreachable candidates. The system achieves an overall picking speed of 560 PPH (pieces per hour).

Humans Can Visually Judge Grasp Quality and Refine Their Judgments Through Visual and Haptic Feedback.

How humans visually select where to grasp objects is determined by the physical object properties (e.g., size, shape, weight), the degrees of freedom of the arm and hand, as well as the task to be performed. We recently demonstrated that human grasps are near-optimal with respect to a weighted combination of different cost functions that make grasps uncomfortable, unstable, or impossible, e.g., due to unnatural grasp apertures or large torques. Here, we ask whether humans can consciously access these rules. We test if humans can explicitly judge grasp quality derived from rules regarding grasp size, orientation, torque, and visibility. More specifically, we test if grasp quality can be inferred (i) by using visual cues and motor imagery alone, (ii) from watching grasps executed by others, and (iii) through performing grasps, i.e., receiving visual, proprioceptive and haptic feedback. Stimuli were novel objects made of 10 cubes of brass and wood (side length 2.5 cm) in various configurations. On each object, one near-optimal and one sub-optimal grasp were selected based on one cost function (e.g., torque), while the other constraints (grasp size, orientation, and visibility) were kept approximately constant or counterbalanced. Participants were visually cued to the location of the selected grasps on each object and verbally reported which of the two grasps was best. Across three experiments, participants were required to either (i) passively view the static objects and imagine executing the two competing grasps, (ii) passively view videos of other participants grasping the objects, or (iii) actively grasp the objects themselves. Our results show that, for a majority of tested objects, participants could already judge grasp optimality from simply viewing the objects and imagining to grasp them, but were significantly better in the video and grasping session. These findings suggest that humans can determine grasp quality even without performing the grasp—perhaps through motor imagery—and can further refine their understanding of how to correctly grasp an object through sensorimotor feedback but also by passively viewing others grasp objects.

Predicting grasp success in the real world - A study of quality metrics and human assessment

Grasp quality metrics aim at quantifying different aspects of a grasp configuration between a specific robot hand and object. They produce a numerical value that allows to rank grasp configurations and optimize based on them. Grasp quality metrics are a key part of most analytical grasp-planning approaches. Additionally, they are often used to generate ground-truth labels for synthetically generated grasp exemplars required for learning-based approaches. Recent studies have highlighted the limitations of grasp quality metrics when used to predict the outcome of a grasp execution on a real robot. In this paper, we systematically study how well seven commonly-used grasp quality metrics perform in the real world. To this end, we generated two datasets of grasp candidates in simulation, each one for a different robotic system. The quality of these synthetic grasp candidates is quantified by the aforementioned metrics. For validation, we developed an experimental procedure to accurately replicate grasp candidates on two real robotic systems and to evaluate the performance of each grasp. Given the resulting datasets, we trained different classifiers to predict grasp success using only grasp quality metrics as input. Our results show that combinations of quality metrics can achieve up to a 85% classification accuracy for real grasps.

Fast finger design automation for industrial robots

Finger design automation is highly demanded from robot industries to fulfill the requirements of the agile market. Nevertheless, literature lacks a promising approach to automate the design process of reliable fingers for industrial robots. Hence, this work proposes the generic optimized finger design (GOFD) method which automates the design process of single- and multi-function finger grippers. The proposed method includes an optimization algorithm to minimize the design process time. The method is utilized to generate fingers for several groups of objects. Results show that the GOFD method outperforms existing methods and is able to reduce the design time by an average of 16,600 s. While the proposed method substantially reduces the design process time of fingers, the quality of grasps is comparable to the traditional exhaustive search method. The grasp quality of GOFD deviates only 0.47% from the absolute best grasp known from the exhaustive search method in average. The designed fingers are lastly manufactured and experimentally verified.

Modeling the Effects of Contact Sensor Resolution on Grasp Success

This letter presents a quantitative method for analyzing the effect of sensor resolution on grasp stability prediction. Resolution limits for contact sensors are expressed as a range of contact locations and contact surface normals that cannot be disambiguated by the sensor. Grasp quality is assessed at the limits of this range to determine whether the uncertainties caused by sensor resolution lead to uncertainty in grasp outcome prediction. The analysis also enables calculation of the specific contact locations on an object where the tactile sensors are trustworthy and where the object is reliably graspable. Our approach lays the foundation for quantitative evaluation in design tradeoffs in sensor choices and sensor layout, as well as finger shapes and materials.

Manipulation of Unknown Objects to Improve the Grasp Quality Using Tactile Information.

This work presents a novel and simple approach in the area of manipulation of unknown objects considering both geometric and mechanical constraints of the robotic hand. Starting with an initial blind grasp, our method improves the grasp quality through manipulation considering the three common goals of the manipulation process: improving the hand configuration, the grasp quality and the object positioning, and, at the same time, prevents the object from falling. Tactile feedback is used to obtain local information of the contacts between the fingertips and the object, and no additional exteroceptive feedback sources are considered in the approach. The main novelty of this work lies in the fact that the grasp optimization is performed on-line as a reactive procedure using the tactile and kinematic information obtained during the manipulation. Experimental results are shown to illustrate the efficiency of the approach.

Application of Hybrid Nelder-Mead Bat Algorithm to Improve the Grasp Quality during the Automated Robotic Grasping

Robots are introduced in the manufacturing industry to automate the manufacturing process to meet the high productivity in the manufacturing industry as well as for manufacturing high quality and low-cost product for the customer. Robots are playing a crucial role in handling the object in pick and place operation, palletizing, welding, handling of materials and other operations in automated manufacturing industries. Robot-mounted with the gripper as the end effector need to identify the best position on the object for positioning the tip of the gripper for a successful grasp. The key to a successful and a stable grasp in automation is to obtain the most appropriate grasp planning and grasp synthesis for the desired task. In this work, a hybrid Meta heuristics method has been developed by combing an efficient local search method known as Nelder Mead method with the Bat algorithm and the new proposed hybrid algorithm is known as the hybrid Nelder-Mead Bat algorithm (HNMBA). The proposed algorithm was introduced for solving the automated grasp synthesis and grasp planning problem in the robotic grasping in an automated manufacturing line. The fitness function for the considered robotic grasp synthesis problem has been formulated using the largest ball grasp quality criteria. The proposed methodology is validated for the tessellated 3D objects, and the obtained results are compared with the previously proposed methods.

Методы определения качества захвата объекта

Методы определения качества захвата объекта

On Grasp Quality Measures: Grasp Robustness and Contact Force Distribution in Underactuated and Compliant Robotic Hands

The availability of grasp quality measures is fundamental for grasp planning and control, and also to drive designers in the definition and optimization of robotic hands. This work investigates on grasp robustness and quality indexes that can be applied to power grasps with underactuated and compliant hands. When dealing with such types of hands, there is the need of an evaluation method that takes into account the forces that can be actually controlled by the hand, depending on its actuation system. In this paper, we study the potential contact robustness and the potential grasp robustness (PCR, PGR) indexes. They both consider main grasp properties: contact points, friction coefficient, etc., but also hand degrees of freedom and consequently, the directions of controllable contact forces. The PCR comes directly from the classical grasp theory and can be easily evaluated, but often leads to too conservative solutions, particularly when the grasp has many contacts. The PGR is more complex and computationally heavier, but gives a more realistic, even if still conservative, estimation of the overall grasp robustness, also in power grasps. We evaluated the indexes for various simulated grasps, performed with underactuated and compliant hands, and we analyzed their variations with respect to the main grasp parameters.

Grasp quality measures: review and performance.

The correct grasp of objects is a key aspect for the right fulfillment of a given task. Obtaining a good grasp requires algorithms to automatically determine proper contact points on the object as well as proper hand configurations, especially when dexterous manipulation is desired, and the quantification of a good grasp requires the definition of suitable grasp quality measures. This article reviews the quality measures proposed in the literature to evaluate grasp quality. The quality measures are classified into two groups according to the main aspect they evaluate: location of contact points on the object and hand configuration. The approaches that combine different measures from the two previous groups to obtain a global quality measure are also reviewed, as well as some measures related to human hand studies and grasp performance. Several examples are presented to illustrate and compare the performance of the reviewed measures.

Grasping deformable planar objects: Squeeze, stick/slip analysis, and energy-based optimalities

Robotic grasping of deformable objects is difficult and under-researched, not simply due to the high computational cost of modeling. More fundamentally, several issues arise with the deformation of an object being grasped: a changing wrench space, growing finger contact areas, and pointwise varying contact modes inside these areas. Consequently, contact constraints needed for deformable modeling are hardly established at the beginning of the grasping operation. This paper presents a grasping strategy that squeezes the object with two fingers under specified displacements rather than forces. A ‘stable’ squeeze minimizes the potential energy for the same amount of squeezing, while a ‘pure’ squeeze ensures that the object undergoes no rigid body motion as it deforms. Assuming linear elasticity, a finite element analysis guarantees equilibrium and the uniqueness of deformation during a squeeze action. An event-driven algorithm tracks the contact regions as well as the modes of contact in their interiors under Coulomb friction, which in turn serve as the needed constraints for deformation update. Grasp quality is characterized as the amount of work performed by the grasping fingers in resisting a known push by some adversary finger. Simulation and multiple experiments have been conducted to validate the results over solid and ring-like 2D objects.

Computational Optimization and Experimental Evaluation of Grasp Quality for Tendon-Driven Hands Subject to Design Constraints

The chief tasks of robotic and prosthetic hands are grasping and manipulating objects, and size and weight constraints are very influential in their design. In this study, we use computational modeling to both predict and optimize the grasp quality of a reconfigurable, tendon-driven hand. Our computational results show that grasp quality, measured by the radius of the largest ball in wrench space, could be improved up to 259% by simply making some pulleys smaller and redistributing the maximal tensions of the tendons. We experimentally evaluated several optimized and unoptimized designs, which had either 4, 5, or 6 tendons and found that the theoretical calculations are effective at predicting grasp quality, with an average friction loss in this system of around 30%. We conclude that this optimization can be a very useful design tool and that using biologically inspired asymmetry and parameter adjustments can be used to maximize performance.

Physically Based Grasp Quality Evaluation Under Pose Uncertainty

Although there has been great progress in robot grasp planning, automatically generated grasp sets using a quality metric are not as robust as human-generated grasp sets when applied to real problems. Most previous research on grasp quality metrics has focused on measuring the quality of established grasp contacts after grasping, but it is difficult to reproduce the same planned final grasp configuration with a real robot hand, which makes the quality evaluation less useful in practice. In this study, we focus more on the grasping process, which usually involves changes in contact and object location, and explore the efficacy of using dynamic simulation in estimating the likely success or failure of a grasp in the real environment. Among many factors that can possibly affect the result of grasping, we particularly investigated the effect of considering object dynamics and pose uncertainty on the performance in estimating the actual grasp success rates measured from experiments. We observed that considering both dynamics and uncertainty improved the performance significantly, and when applied to automatic grasp set generation, this method generated more stable and natural grasp sets compared with a commonly used method based on kinematic simulation and force-closure analysis.

An Efficient Algorithm for a Grasp Quality Measure

This paper presents an efficient algorithm to compute the minimum of the largest wrenches that a grasp can resist over all wrench directions with limited contact forces, which equals the minimum distance from the origin of the wrench space to the boundary of a grasp wrench set. This value has been used as an important grasp quality measure in optimal grasp planning for over two decades, but there has been no efficient way to compute it until now. The proposed algorithm starts with a polytope containing the origin in the grasp wrench set and iteratively grows it such that the minimum distance from the origin to the boundary of the polytope quickly converges to the aforementioned value. The superior efficiency and accuracy of this algorithm over the previous methods have been verified through theoretical and numerical comparisons.

Evaluation of Human Prehension Using Grasp Quality Measures

One of the main features of the human hand is its grasping ability. Robot grasping has been studied for years and different quality measures have been proposed to evaluate the stability and manipulability of grasps. Although the human hand is obviously more complex than robot hands, the methods used in robotics might be adopted to study the human grasp. The purpose of this work is to propose a set of measures that allow the evaluation of different aspects of the human grasp. The most common robotic grasp quality measures have been adapted to the evaluation of the human hand and a new quality measure – the fatigue index – is proposed in order to incorporate the biomechanical aspect into the evaluation. The minimum set of indices that allows the evaluation of the different aspects of the grasp is obtained from the analysis of a human prehension experiment.

A Novel Synthesis of Computational Approaches Enables Optimization of Grasp Quality of Tendon-Driven Hands

We propose a complete methodology to find the full set of feasible grasp wrenches and the corresponding wrench-direction-independent grasp quality for a tendon-driven hand with arbitrary design parameters. Monte Carlo simulations on two representative designs combined with multiple linear regression identified the parameters with the greatest potential to increase this grasp metric. This synthesis of computational approaches now enables the systematic design, evaluation, and optimization of tendon-driven hands.

Discrete fuzzy grasp affordance for robotic manipulators

Grasp affordance determines the object-hand relative configurations which lead to successful grasps. Generation and representation of grasp affordances can increase achieved grasp quality and be integrated in path planning algorithms facilitating increased efficiency. Grasp quality is determined by various measures and may have a major impact on task success. Fuzzy grasp affordance can be defined based on a fuzzy grasp quality grade and enhance the previously Boolean notion of grasp affordance. Fuzzy grasp affordances can be represented using a discrete manifold. This facilitates integration of data from various sources and representation optimization using evolutionary algorithms. A method for construction of a discrete fuzzy grasp affordance manifold is presented and demonstrated for apple selective harvesting. The affordance constructed is based on learning from human demonstration. It includes quality grade determination, manifold structure determination, cell quantization, and smoothing. An algorithm for adaptation of the computed manifold to different manipulators and grippers is developed and implemented for two different end effectors. Additionally a method for online integration of the developed affordance is presented.