Weed control should be regarded as one of the key technological operations in organic crop production, along with the application of biological plant protection agents and the localized placement of organic fertilizers. Mechanized equipment currently available for row-crop cultivation is primarily designed to remove weeds from the inter-row space. However, 15–20 percent of weeds may remain within the crop rows, in the spaces between cultivated plants. (Research purpose) The study aimed to develop a structural and technological diagram and an operating algorithm for an automated sectional working unit for in-row weed removal in potato crops, using digital solutions based on machine learning and computer vision. (Materials and methods) A distinctive feature of in-row weed control in potato crops is that the weeding blade must periodically move into and out of the ridge during operation. This imposes specific requirements on the blade design, its geometric parameters, and the drive system. A system for controlling the blade working depth is also required to maintain stable operation under variable soil-contact conditions. (Results and discussion) An information model of the weed removal process was proposed. The model describes how design, technological, technical, and environmental parameters affect weed removal efficiency, crop damage, and energy consumption during the operation. A structural and technological diagram was developed, and an operating algorithm for the automated sectional working unit was compiled. (Conclusions) The study determined how the force acting on the side cutting edges of the weeding blade depends on the inclination angle of these edges and the working depth during blade entry into and exit from the ridge. Under the optimal blade parameters, this force should not exceed 38.2 newtons. The proposed process model enables accurate estimation of the loads acting on the weeding blade drive and supports the rational selection of its operating parameters.