Design And Implementation Of ROS Robot Cloud Native Platform

With the development of hardware capabilities and communication technology,the field of robotics has continued to develop.Robots have begun to show their skills in various fields.At the same time,the requirements for robotics computing capabilities are gradually increasing.I hope that they can be competent for more complex tasks and achieve more complex tasks.Computing tasks.An obvious idea began to appear,bringing the computing power of the cloud to the robot.This system mainly wants to realize a robot cloud application platform.By introducing cloud native technology into the field of ROS robots,the robot’s computing modules are packaged into applications and offloaded to the cloud,using Docker for image distribution,containerized packaging,and using Kubernetes.Cloud application management realizes cloud migration of robot computing power.The system is mainly divided into three parts: cloud platform,robot connector,and application connector.The cloud platform will be divided into user management,robot management,application management and other sub-modules.User management is mainly to isolate the user space of different users to ensure application isolation between different users.Robot management mainly involves monitoring,deploying robots,and issuing instructions to robot connectors.Application management includes the management of the application life cycle,that is,creation,destruction,etc.,as well as the monitoring of the application.The robot connector is responsible for docking the robot and receiving instructions issued by the cloud platform,including establishing a connection with the ROS robot,monitoring the position,power,and fault status of the robot.At the same time,the robot connector needs to forward the message of the ROS robot to the designated application and receiving application.The calculated results are fed back to the robot itself.The application connector is responsible for connecting the application and the cloud platform.The DNS discovery based on Kubernetes realizes the communication between the application and the robot.It is also the goal of this system to speed up application deployment and realize the rapid upload of robot computing modules to the cloud.Based on the cloud-native declarative API,rapid deployment and expansion of applications can be achieved.Compared with the traditional application deployment based on the robot itself,this method saves the robot’s computing power,and uses more efficient and faster cloud computing power to greatly expand the robot’s own capabilities,making it capable of more complex tasks.