In the context of global challenges faced by agriculture (growing demand for food, limited resources and stricter environmental requirements), domestic crop and livestock production demonstrate a steady increase in innovation activity and high sectoral development rates. Robotization of agriculture provides automation of technological processes while reducing the share of manual labor and is based on the transition to high-precision, resource-saving and sustainable crop production. The aim of this study is to develop a computer vision-based control system and create a functioning prototype of a mobile robot for modern crop production, demonstrating the possibility of autonomous execution of technological operations for tillage based on navigation using Aruco visual marker recognition. The proposed vision system allows solving scientific and technical problems: ensuring stable movement along a trajectory set by markers with the possibility of its operational planning with limited computing resources.; the use of available hardware platforms to solve the problem of cultivating the soil with a screw tool, taking into account the achieved speed of movement of an autonomous mobile robot; reducing the requirements for operator qualifications; using computer vision algorithms and methods that determine the scientific novelty of the study. As a result, a reliable control system for an autonomous mobile robot has been developed with active adjustment of the rotation speed of the screw device for crop production tasks based on computer vision approaches, taking into account hardware limitations and stable retention of the planned trajectory. The implementation of the development during project scaling in the agro-industrial complex of the region will reduce the cost of manual labor and resource consumption; increase the accuracy of agrotechnical operations and switch to environmentally friendly technologies by minimizing chemical stress and soil compaction; create a framework for collecting precision farming data and building digital field models. Further prospects for research on the project involve integrating the development into agrotechnological processes of crop production, taking into account the requirements of regional agricultural producers
Keywords
artificial intelligence, digital transformation, agro-industrial complex, unmanned technologies, vision system, Aruco markers