This paper studies the end actuator motion state of an RRR industrial robot, and gives a bottom up decision-making scheme based on PID algorithm. Under the background of complex industrial control, any environmental factors will have a deep effect on the accuracy. The physical representation of the accuracy of the end effector is the electrical signal. Foscus on the signal difference from joint to joint, based on the ROS and gazebo, the simulation of the RRR mechanical arm model can theoretically and intuitively Reflect the differences. To minimize the error, proportional–integral–derivative controller will be sued in this model and then Adjust PID parameters of each controller. Assume that the external interference is a linear superposition of impulse signals. The impulse response which the model makes should return to normal as soon as possible. In this way, the purpose of anti-interference is achieved. At the same time, it focuses on solving the motion state of industrial manipulator. For general industrial robots, its actuator, a gripper, should have the function of reaching any position and taking any direction in a certain space area. In other words, if a reference frame o 'is fixed on the center of the hand claw, the robot can send o' to any direction at any position in the space area. The position and direction of the gripper are determined by the amount of motion from the base to the gripper pairs. In addition, with the rapid development of the Internet, image has become a common form of information. This paper will use image recognition technology to provide the required detection data and reference values for the PID controller. Thus, the workspace of the industrial robot can be reflected more completely.