The developer has created a navigation module for a strawberry-harvesting robot, designed to move the platform between specific points in a greenhouse. The system is built as a modular architecture using Python, ROS 2, and the Webots simulator.
The navigation module consists of four core components: a graph-based field map, a BFS-based path planner, a motion controller using odometry, and an obstacle detector. The planner prioritizes the shortest path while minimizing turns.
The current prototype operates on a 5x5 grid where each marker is spaced one meter apart. The system calculates routes, executes movement, and detects new obstacles by comparing real-time lidar data against a baseline scan.
The project is currently in the navigation testing phase and does not yet perform fruit harvesting. Future development plans include real-time obstacle detection during movement and integration with computer vision and robotic manipulators.
