Mobile Robot Capable of Crossing Floors for Library Management

Abstract

In order to reduce the labor and material of library management and contribute to the development of unmanned library, an intelligent mobile robot for library management is proposed. The robot consists of a mobile platform, a three-DOF rise-and-fall robotic arm and a multisource image recognition information fusion system. Besides recognizing data and position of the book, the robot is also able to conduct book grasp and shelving in order task. To adapt to the existing library environment and enlarge its move area, an efficient autonomous elevator button recognition system based on neural networks is designed, which enables the mobile robot span floors by elevator. Combined with visual recognition technology, the mobile robot has functions of autonomous navigation and obstacle avoidance. There is also a data monitoring center and graphical user interface corresponding to the mobile robot, which is developed to simplify the operation instructions. The effectiveness of the proposed research is examined by simulation and experiments. The innovation of the robot is that it can complete shelving book task on the premise of adapting present library to the most extent.