Arranging Test Tubes in Racks Using Combined Task and Motion Planning

结合强化学习与运动规划的机器人试管重排

Robotic Test Tube Rearrangement Using Combined Reinforcement Learning and Motion Planning

本文提出任务级强化学习与运动规划闭环框架，结合D3QN分布式学习和A*后处理，解决多类别架内试管重排问题，并通过仿真与真实实验验证其有效性和鲁棒性。

Soft Pneumatic Actuator Capable of Generating Various Bending and Extension Motions Inspired by an Elephant Trunk

Inspired by the dexterous handling ability of an elephant's trunk, we propose a pneumatic actuator that generates diverse bending and extension motions in a flexible arm. The actuator consists of two flexible tubes. Each flexible tube is restrained by a single string with variable length and tilt angle. Even if a single tube can perform only three simple types of motions (bending, extension, and helical), a variety of complex bending patterns can be created by arranging a pair of tubes in parallel and making the restraint variable. This performance takes advantage of the effect of the superpos…

Automatically Prepare Training Data for YOLO Using Robotic In-Hand Observation and Synthesis

Deep learning methods have recently exhibited impressive performance in object detection. However, such methods needed much training data to achieve high recognition accuracy, which was time-consuming and required considerable manual work like labeling images. In this paper, we automatically prepare training data using robots. Considering the low efficiency and high energy consumption in robot motion, we proposed combining robotic in-hand observation and data synthesis to enlarge the limited data set collected by the robot. We first used a robot with a depth sensor to collect images of objects…