||Dr. Zhu Xiaoxiao (Specially Appointed Associate Professor, Global Alliance Laboratory, University of Electro-Communications, and Researcher, Institute of Robotics, Shanghai Jiaotong University, China)
||How to build a robot control system with robustness, scalability, intelligence…? How to start playing with a robot if you are not an expert of robotics but do have some good ideas？In this report, the Robot Operating System (ROS) is introduced, which is a good answer to these questions. ROS is an open-source, meta-operating system that provides the services required for operating a robot system, including hardware abstraction, low-level device control, implementation of commonly-used functionality, message-passing between processes, and package management.After the introduction of ROS, an intelligent 7-DoF robot arm system is then discussed in detail. This robot arm is designed for a Brain Machine Interface (BMI) system. Thanks to the coupled tendon-driven mechanism, the arm is 0.8m long and weighs 2.5 kg, but can lift a 1 kg object at a speed of 1 m/s. Besides that, the robot has a RGB-D sensor for computer vision task and four multi-point proximity sensors for obstacle avoiding. Several ROS modules of hardware driver, controller, object recognition and arm path planning are developed for constructing the control system of the robot. The robot arm controlled by ROS can naturally simulate the movement of a human arm, which is desired by BMI applications that execute human motion intention decoded from brain signals with robots. The strategies to integrate the robot arm system with the BMI system will also be discussed.