Task and Context Sensitive Gripper Design Learning Using Dynamic Grasp Simulation

Abstract

In this work, we present a generic approach to optimize the design of a parametrized robot gripper including both selected gripper mechanism parameters, and parameters of the finger geometry. We suggest six gripper quality indices that indicate different aspects of the performance of a gripper given a CAD model of an object and a task description. These quality indices are then used to learn task-specific finger designs based on dynamic simulation. We demonstrate our gripper optimization on a parallel finger type gripper described by twelve parameters. We furthermore present a parametrization of the grasping task and context, which is essential as an input to the computation of gripper performance. We exemplify important aspects of the indices by looking at their performance on subsets of the parameter space by discussing the decoupling of parameters and show optimization results for two use cases for different task contexts. We provide a qualitative evaluation of the obtained results based on existing design guidelines and our engineering experience. In addition, we show that with our method we achieve superior alignment properties compared to a naive approach with a cutout based on the "inverse of an object". Furthermore, we provide an experimental evaluation of our proposed method by verifying the simulated grasp outcomes through a real-world experiment.

Compensating Pose Uncertainties through Appropriate Gripper Finger Cutouts

“…In (Wolniakowski et al, 2013), we have presented a general and flexible method that allows to compute an array of gripper Quality Indices based on statistical analysis of the simulated grasp outcomes. In (Wolniakowski et al, 2015) and (Wolniakowski et al, 2017), we have extended our method with additional alignment quality assessment, proposed the gripper finger parametrization, and shown an optimization procedure that yields gripper designs optimized for specific.…”

“…The gripper evaluation method based on dynamic simulation (in dynamic simulation, the object movements are calculated based on the object contacts, forces and torques, as opposed to kinematic simulation, which uses velocity and position trajectories) of grasps has been introduced and described in detail in our previous works (Wolniakowski et al, 2013(Wolniakowski et al, , 2015(Wolniakowski et al, , 2017. We will, however, present the key concepts here for the sake of completeness.…”

“…In this work, we first present the use case of an industrial ob-ject grasping scenario, to which we apply our simulation-based gripper optimization method introduced in (Wolniakowski et al, 2013(Wolniakowski et al, , 2015(Wolniakowski et al, , 2017. As compared to these works, we show in this paper the optimization process and the optimal finger geometry computed for the task context, and provide in-depth analysis of our gripper optimization method performance with regard to expected alignment of the objects.…”

The performance identification of gripper by isocline analysis

“…In this research, the For parameters of the optimization, there are size parameters of gripper and the geometrical dimensions of the objects. 9, [14][15][16] So first of all, this research analyzed the performance of the gripper by isocline image model analysis, followed by adjusting the geometry of grippers and objects, then analyzing the geometry of adjusted gripper and object again, and finally optimizing the performance of the grippers. Figure 1 shows the preparation of this research: first, select testing gripper; second, decide testing boundary; third, select testing object; forth, test and mark; fifth, identifying points or areas that are easy to catch; sixth, determine whether the boundary has been found; seventh, inward offset this boundary to find the optimal catching point or areas; and finally, adjust the gripper to the best, according to result.…”

Automated Fixture Design using an Imprint-based Design Approach and Optimisation in Simulation

“…The idea of using optimizations in simulations was inspired by the approach presented in [17] for gripper design, where a parameterized nger model is op-timized to get the best suited gripper for a specic task. Unlike in [17], the process is signicantly automated by introducing an automated parametrization method. This method, while applied here for xtures, was also generalized for gripper design [14].…”