Sensorless Reduction of Cane Oscillations Aimed at Improving Robotic Grapevine Winter Pruning

Agricultural sector faces challenges like high labour costs and a shortage of qualified workers for repetitive tasks, leading to increased interest in agricultural robotics. Pruning is a focus for automation efforts worldwide. However, pruning robots struggle with slow and inaccurate vision systems, resulting in slow, costly, and potentially harmful operations for plants. This study aims to provide a reproducible and reliable method for detecting contact with grapevines during pruning, minimising potential damage, and improving vision system speed by reducing cane oscillations. The proposed approach uses a momentum-based observer, eliminating the need for force sensors. Experiments on Vitis vinifera cv. Pinot Noir canes validated this methodology.