A Transfer Learning for Estimation of Operation Time for 6-axis Robot Arms

We propose a transfer learning method to estimate the motion time of a 6 -axis robot arm obtained by 3D simulation using a machine learning method. The model is constructed from two perspectives: one without obstacles and one with obstacles. The motion planning method is Rapidly exploring Random Trees Star (RRT^*), where the initial and target postures of the robot are set randomly within its range of motion, and the resulting motion times are used for learning. The model in an environment without obstacles can be constructed with high speed and high accuracy. We develop a transfer learning method for a heterogeneous 6 -axis robot arm to verify the speed-up of model construction. For the machine learning model with obstacles, the problem setup was extended to adapt to obstacles based on the former model. Computational results show that both the accuracy and reliability of the model were improved by the proposed model.

I. Introduction

In recent years, the working population has been decreasing due to the falling birthrate and the aging population in Japan. It has a significant impact on the retail industry, such as supermarkets and convenience stores. The retail industry requires a flexible and multifunctional workforce to display and restock merchandise, clean stores, and operate cash registers, so it has been considered that those complex operations cannot be substituted without human workers. With the rapid development of robotic arms and AI technology, robots are increasingly replacing these tasks. By integrating the robot arm with AI technology, it is possible to add functions such as object recognition, automatic obstacle avoidance, object tracking, and path planning to the robot arm as well as autonomous control. Aisin Corporation has realized a next-generation supply system that automatically performs processes such as parts distribution, storage, and empty box collection at appropriate times by utilizing two handling intelligent robots and 31 AGVs [1].