Exploring Innovation in Weed Control: Autonomous Mapping and Monitoring of Rumex at CODIT 2024

This year’s CODIT (Control, Decision and Information Technologies) conference, held in the beautiful Mediterranean island of Malta, witnessed many fascinating research presentations. Among them was a paper from our team, presented by Ngoc Thinh Nguyen, titled “Autonomous Mapping and Monitoring of Rumex with Mobile Robot.”

Weed control is a perennial issue in agriculture, traditionally demanding significant time and labor. The paper proposes an autonomous solution for the mapping and monitoring of Rumex, a common and problematic weed. Utilizing advanced technologies like AI and robotics, his approach promises to revolutionize how farmers manage weed infestations.

The autonomous process relies on several sophisticated tools and methodologies:

• YOLOv8 Object Detection: This real-time, one-stage object detector processes images to identify Rumex plants. The model, trained on a custom dataset, ensures accurate detection even under varying environmental conditions.
• Ant Colony Optimization (ACO): A bio-inspired algorithm that optimizes the robot’s path to revisit all mapped Rumex locations efficiently, minimizing travel time and energy consumption.
• ROS-Based Navigation: The Robot Operating System (ROS) provides robust navigation capabilities, integrating various sensors for precise movement and obstacle avoidance.

The presentation at CODIT 2024 highlighted the transformative potential of autonomous systems in agriculture. Our work not only addresses a pressing agricultural challenge but also exemplifies the innovative spirit that CODIT and Malta foster. As we look forward to more advancements in this field, the synergy between cutting-edge technology and practical applications continues to pave the way for a more efficient and sustainable future in agriculture.