Manipulation Skill Acquisition for Robotic Assembly Based on Multi-Modal Information Description

Abstract

Automatic assembly of elastic components is difficult because of the potential deformation of parts during the assembly process. Consequently, robots cannot adapt their manipulation to dynamic changes. Designing systems that learn assembly skills can help in alleviating the uncertain factor for industrialgrade assembly robots. This study proposes a skill acquisition method based on multi-modal information description to realize the assembly of systems with elastic components. This multi-modal information includes two-dimensional images, poses, forces/torques, and robot joint parameters. In this method, robots acquire searching, location determination, and pose adjustment skills using these multi-modal information parameters. As a result, robots can reach the assembly target by analyzing two-dimensional images with no position constraint. While acquiring pose adjustment skills, the reward function with depth and assembly steps is used to improve the learning efficiency. The deep deterministic policy gradient (DDPG) algorithm is applied for acquiring skills. Experiments using a KUKA iiwa robot demonstrated the effectiveness and conciseness of our method. Our results indicate that the robot acquired searching, location determination, and pose adjustment skills that allowed it to successfully complete elastic assembly. INDEX TERMS Industrial robots, acquisition of manipulation skills, deep reinforcement learning, multi-modal information description. I. INTRODUCTION A. MOTIVATION Robots are used for more efficient production assembly. Research on automation of assembly systems has focused on intelligent, automatic, and flexible robots. During an actual assembly process, force measurement data are not available owing to elastic deformation between workpieces. Elastic deformation can be neglected when dealing with assembly of rigid parts. However, in the case of weakly rigid components, such deformation causes a series of uncertain changes, such as position, posture, and depth changes. There are methods based on force-control and compliance control that are already facing the problem of assembly of elastic The associate editor coordinating the review of this article and approving it for publication was Okyay Kaynak.

Dynamics Learning With Object-Centric Interaction Networks for Robot Manipulation

“…Many approaches on the prediction of physical interactions have been applied from video prediction [15] to robotic manipulation tasks [16]. Some methods build the system dynamics by explicitly modeling the state transitions, leveraging ground-truth poses [17], [18] or known physical properties [19]. However, access to the ground-truth states and physical properties may be difficult for most real-world applications.…”

Multiple peg-in-hole compliant assembly based on a learning-accelerated deep deterministic policy gradient strategy

“…However, the knowledge base greatly affects the learning effect of the agent. When the knowledge is insufficient, the learning effect will also decrease (Li et al, 2020). Kim et al (2020) proposed an imitation learning strategy and combined this strategy with the DDPG algorithm to complete the assembly task.…”

“…However, it is difficult to complete precise assembly tasks only by RL. In Li et al (2020), the DDPG agent directly outputs the displacement command of the joint space of the robot. This method has a good performance in non-contact tasks, such as robotic painting or welding (Lynch and Park, 2017).…”

A Skill Programming Method Based on Assembly Motion Primitive for Modular Assembly System

“…Because different types of sensors have their own limitations, the modular assembly system integrating multiple sensors can meet the requirements of complex assembly tasks. In addition to trajectory teaching, the modular assembly system also has functions such as image processing [2], force feedback control [3], IO control, multimodular parallel or serial control, and collision detection [4]. And the modular assembly system has the characteristics of multi-information fusion, complex coordination, and diverse control methods [5].…”