RATIONAL AGENTS FOR AGRICULTURAL CROP SURVEYING: ADAPTIVE TASK AND MOTION PLANNING

The farmer requires detailed information on crop stress conditions in order to manage the crop growth. This information may be acquired by low granular satellite images to coarsely identify areas of interest within the field followed by ground measurements carried out by robotic surveying platforms for plant identification and segmentation. A novel concept for creating rational behaviors of surveying agents within agricultural row crops is outlined. Global and local adaptive planning strategies combined with structural knowledge are used to focus on the automated generation of internodes, edges, and bi-directed edge costs. Furthermore, path bands are introduced as part of the task and motion planning.