An OpenStack-based Control Middleware For Cloud Robots

As the traditional robots being used more and more,there are some big problems we must face.They don't perform very well when the tasks,such as simultaneous localization and mapping,object recognition and capture,navigation and so on,being executed on them.Traditional robots have limited hardware resources,such as computing power and storage capacity.Today's era is the era of cloud computing,and OpenStack is the leading force in the development of cloud computing.Only the combination of the robot and the cloud computing can conform to the trend of the times,break through the bottleneck of the development of the traditional robot,make the robots development into a new era,the era of cloud robot.But how to make the traditional robots to use the virtual resources in the cloud on demand,make them become the cloud robots,the problem has become a new research hot spot in the world.To solve this problem,we propose a new method based on OpenStack cloud robot a control middleware,which is based on the State Key Laboratory of Synthetical Automation for Process Industries in Northeastern University.The main research work is summarized as follows:(1)The core components of the OpenStack are studed deeply.Through the analysis of the working process of the core components and their relations,I'm familiar with the working principle of OpenStack cloud,master the management method of OpenStack cloud platform(create a new image for the robot).For the design and implementation of control middleware has laid a solid foundation.(2)An in-depth analysis of the message oriented middleware RabbitMQ used in the OpenStack cloud.Master the message passing method,design and implement the message transmission module in the control middleware.Make sure that the robot request message is received,stored and transmitted reliably and efficiently.(3)The robot request message is encapsulated in the form of JSON,and the based on JSON message parser is designed and implemented.The message parser receives the robot request message from the message oriented middleware,and analyzes to obtain the parameters that are used in the virtual resource automatic configuration module.(4)The automatic configuration module of the virtual resources in the control middleware is implemented by the OpenStack Python SDK,which receives the parameters from the message parser.It calls the corresponding resource scheduling optimization algorithm,selects the optimal physical host,and automatically creates the virtual resources in the cloud for the robot.Using it,the robot can use the cloud resources on demand.(5)Ubuntu and robot operating system(ROS)are used to achieve the control module of control middleware.The virtual machine created according to the robot request message is integrated with Ubuntu and ROS,which can run different algorithms for robot control.(6)The OpenStack cloud platform is designed and deployed.Each module in the control middleware is verified by the Web RabbitMQ manager and the Nao robot.The correctness and feasibility of the control middleware based on OpenStack cloud robot are proved by experiments,which provide a new idea for the development of the cloud robot.