Robotic system for harvest prediction and targeted herbicide application. The system utilizes autonomous robotic rover and drone equipped with RGB and multispectral cameras for weed and fruit detection. The Faster R-CNN AI method analyzes image datasets for accurate recognition. To enhance production estimates, the system predicts tree canopy volume using drone-captured images. Additionally, the autonomous robotic rover includes a cartesian robotic arm with a sprayer for particular weed-spraying function and a robotic gripper to pick up fallen fruit, contributing to sustainable agriculture and efficient orchard management.