Reconfigurable Grasp Planning Pipeline with Grasp Synthesis and Selection Applied to Picking Operations in Aerospace Factories

Highlights

• Modular, re-configurable and flexible robot grasping planning pipeline;

• Introduction and review of multi-fingered grasping and simulated annealing theory;

• Real industrial application results with an Aerospace test-case.

Abstract

Several approaches with interesting results have been proposed over the years for robot grasp planning. However, the industry suffers from the lack of an intuitive and reliable system able to automatically estimate grasp poses while also allowing the integration of grasp information from the accumulated knowledge of the end user. In the presented paper it is proposed a non-object-agnostic grasping pipeline motivated by picking use cases from the aerospace industry. The planning system extends the functionality of the simulated annealing optimization algorithm for allowing its application within an industrial use case. Therefore, this paper addresses the first step of the design of a reconfigurable and modular grasping pipeline. The key idea is the creation of an intuitive and functional grasping framework for being used by factory floor operators according to the task demands. This software pipeline is capable of generating grasp solutions in an offline phase, and later on, in the robot operation phase, can choose the best grasp pose by taking into consideration a set of heuristics that try to achieve a successful grasp while also requiring the least effort for the robotic arm. The results are presented in a simulated and a real factory environment, relying on a mobile platform developed for intralogistic tasks. With this architecture, new state-of-art methodologies can be integrated in the future for growing the grasping pipeline and make it more robust and applicable to a wider range of use cases.

Introduction

Planning and performing a grasp movement is done effortlessly by humans, but for robots, this is a significant challenge. The current established industrial solutions are only capable of dealing with this problem in well-structured and controlled environments. Typically, these solutions resort to techniques that depend on the operators’ expertise, which manually programs the robotic system, or are based on inflexible, application oriented, software tools (e.g., drive through, lead through and offline programming), which do not convey with modern industry paradigms that ultimately seek for new autonomous and efficient techniques to enhance the flexibility of industrial robotic systems.

For decades the study of grasp techniques in complex scenarios has been explored by the scientific community, which led to the appearance of several analytical [1], [2], [3], [4], [5], [6] and data-driven [7] approaches aiming for the improvement of production lines, logistics processes, assembling operations, and bin-picking tasks. Despite these significant contributions, the complexity associated with designing a task-oriented analytical method or build a large dataset, required for the training of Machine Learning (ML) systems, limits the effective adoption of these technologies as an efficient, user-friendly, and applicable industrial solution.

In this context, this paper introduces a reconfigurable robot grasping software pipeline. It is based on a sequential architecture to autonomously compute the grasp solution for a robotic arm in an industrial application. This pipeline is built on top of Robot Operating System (ROS) and “GraspIt!” simulator [8], extending the applicability of Simulated Annealing (SA) [9] with a feasible application time.

With this work the authors goal is to deploy to both the industrial and scientific community a software tool capable of automatically generating robot grasp poses over a set of objects. Namely, this paper presents the backbone of the proposed modular and configurable pipeline, where methodologies and tools already consolidated in the scientific community will be further integrated. Furthermore, this tool will serve as the basis for future developments on the robot grasping topic.

Practical results are presented considering a real aerospace factory use case, targeting the execution of intralogistic operations by an omnidirectional robot equipped with a robotic arm, i.e., a mobile manipulator (Fig.  1).

Bearing these ideas in mind, this paper is structured as follows: Section 2 discusses the related work. Section 3 presents relevant background on robot based grasp topic. Section 4, presents the proposed grasp planning software backbone. Finally, in Section 5 the experimental results are presented and discussed, followed by the Conclusions and Future Work (Section 6).