With the rapid development of the information age, the development of industrial robots is also advancing by leaps and bounds. In the scenes of automobile, medicine, aerospace, and public service, we have fully enjoyed the convenience brought by industrial robots. However, with the continuous development of industrial robot-related concepts and technologies, human-computer interaction and cooperation have become the development trend of industrial robot. In this paper, the human-machine cooperation and path optimization of industrial robot in a complex road environment are studied and analyzed. At the theoretical modeling level, firstly, the industrial robot is modeled and obstacle avoidance is analyzed based on the kinematics of industrial robot; thus, an efficient and concise collision detection model of industrial robot is proposed. At the algorithm level, in view of the complex road conditions faced by industrial robots, this paper will study and analyze the obstacle avoidance strategy of human-computer cooperation and real-time path optimization algorithm of industrial robots. Based on the virtual target point algorithm, this paper further improves the problem that the goal of the traditional path planning algorithm cannot be fully covered, so as to propose the corresponding improved path planning algorithm of industrial robots. In the experimental part, based on the existing industrial robot system, the human-machine cooperation and path planning system proposed in this paper are designed. The experimental results show that the algorithm proposed in this paper improves the accuracy of obstacle avoidance by about 10 points and the corresponding convergence speed by about 5% compared with the traditional algorithm and the experimental effect is remarkable.