At present, the technical level of the review counting section of the bundle of bars in China is very backward, the high output of the Chinese steel industry and the demand for automation. According to the World Iron and Steel Association 2024 data, China's crude steel production accounts for 49% of the world, especially in Tianjin, Hebei and Shandong, where the total steel production reached 0.273 billion tons, but the traditional manual review sorting has problems such as low efficiency and high labor intensity. Aiming at the pain point of this industry, the project team has developed an intelligent review and sorting manipulator system. Through machine vision and RPP robot technology, the project team integrates X/Y/Z three-axis screw module and 360 rotating reducer to realize automatic counting and sorting functions. The system workflow includes five-step state switching, which can effectively replace manual operation. The manual point check counting method is widely used, which is influenced by the complex and changeable shape of the baled bar, the physiological and psychological factors of the workers, and the counting error caused by the review error is very prominent. The manual review count does not match the high productivity of the bar mill, which becomes the production bottleneck of the bar mill. This product is an intelligent review and sorting manipulator. It integrates the end face image of bundled bars through the machine vision system, and carries out image processing and image analysis to realize automatic counting and determine the count information of bundled bars. Then it controls the intelligent review and sorting manipulator through PLC to complete online automatic review and sorting, which can greatly improve labor productivity, reduce labor intensity of workers, improve automation degree of production line, reduce review counting error and improve economic benefits.