• 1 Institute of Vehicles – Warsaw University of Technology, Poland
  • 2 Industrial Institute of Agricultural Engineering in Poznań, Poland


The aim of the paper was to test of the concept of the navigation system for the autonomous robot for sowing and wide row planting. Autonomous work of the robot in the field of traction and agronomic processes is implemented based on data from many sensors (cameras, position sensors, distance sensors, and others). The robot is intended for ecologic cultivation requiring mechanical removal of weeds or in crops with application of selective liquid agrochemicals limited to the minimum. The use of a vision system, based on the map coordinates of the position of the sown seeds, allows for their care on an early stage of plant development. Main sensor system is based on a specialized GPS receiver and inertial navigation providing position information with an accuracy of around 10 mm. To determine the angular acceleration the IMU (Inertial Measurement Unit) is used. Additionally, information from the acceleration sensors and wheel encoders is used for navigation purposes. This system is used to: control the speed of the robot, keep the robot on the designated path, and detect the precise position of the seeds. The exact information of the seeds position is used to build maps of seeds, which will be used as supporting information for precision weeding, and to control the position of and operation of key components. The front camera view is used to increase positioning accuracy of the robot. It will allow corrections of the robot path regarding the rows of plants. The vision system is also used for detection of non-moving objects. A structure of requirements for the SQL database has been developed, which is used to store plant and weed geo-data, as well as store data about plants and weeds, based on images recorded by the vision system.


Article full text

Download PDF