Application of LiDAR and SLAM Technologies in Autonomous Systems for Precision Grapevine Pruning and Harvesting

¹ Department of computers Techniques engineering, College of technical engineering, The Islamic University, Najaf, Iraq The Islamic University of Al Diwaniyah, Al Diwaniyah, Iraq The Islamic University of Babylon, Babylon, Iraq
² Ahl Al Bayt University, Karbala, Iraq
³ Department of CSE, GRIET, Hyderabad, Telangana, India

^* Corresponding author: saifobeed.aljanabi@iunajaf.edu.iq

Abstract

Integrating autonomous systems into precision agriculture brings new integrated management in vineyards for operational efficiency and accuracy. This project creates an autonomous system for grapevine pruning and harvesting using LiDAR, SLAM, RGB-D cameras, Convolutional Neural Networks (CNNs), proximity sensors, and Wireless Sensor Networks (WSNs). LiDAR produces detailed 3D vineyard maps that integrate with SLAM algorithms for accurate navigation, ensuring efficient local and global information relay. RGB-D cameras capture visual and depth information of grapevines and fruits, while CNNs process this data to classify different vines and grapes, enabling focused pruning and harvesting decisions. Proximity sensors provide real-time distance measurement for safe operation, allowing obstacle navigation without damaging equipment or vines. WSNs facilitate communication between system components through data exchange, enabling continuous environmental monitoring and real-time adjustments to maximize performance. The project aims to integrate advanced technologies in grapevine pruning and harvesting to optimize these processes. The system improves accuracy and speed, reducing labor costs while enhancing grape yield and quality, representing a promising approach to vineyard management. LiDAR generates detailed 3D maps while SLAM provides navigation with localization accuracy better than 2 cm. RGB-D cameras with CNNs identify grapevines and fruits with 95% accuracy. Proximity sensors ensure obstacle avoidance with 98%> accuracy, and WSNs integrate real-time data with less than 50ms latency. The system has increased harvesting efficiency by 15% and decreased operating costs by 20%o.