Final Report: June 30, 2021
Principal Investigator: Qin Zhang
Center for Precision and Automated Agricultural System, Washington State University (WSU CPAAS)
24106 N Bunn Road, Prosser, WA 99350
509-786-9360
qinzhang@wsu.edu
Co-PI: Manoj Karkee
WSU CPAAS
24106 N Bunn Road, Prosser, WA 99350
509-786-9208
manoj.karkee@wsu.edu
Project Summary: Canopy management is one of the major production activities in the annual lifecycle of the vineyard. Green shoot thinning is one of the many field operations used to create and maintain healthy and productive canopies, which, in another term, is a component of overall pruning activity. Shoot thinning is used to improve uniformity, spacing and direction of shoot growth. This operation can help improve light penetration and air movement through a canopy, adjust crop load (by thinning fruiting shoots to reduce the crop), and adjust the leaf-area-to-crop ratio (by thinning non-fruiting shoots). Shoots growing from the base of spurs, multiple shoots from the same node, shoots growing from non-spur positions or originating in the head region or on the trunk are all candidates for removal, unless needed to replace an old or poorly positioned spur or an old cordon. Shoot thinning is a labor-intensive task and is costly. Dean (2016) found that the cost for shoot thinning by hand is about $650 per hectare. If a mechanical shoot thinner is used, the cost could be reduced to about $25 per hectare. In addition, one machine can replace up to 25 workers (productivity 25 hrs/ha vs 1 hr/ha; Dean, 2016). Therefore, mechanical shoot thinning is essential for the profitability and sustainability of wine grape production. In the past, a few machines have been developed and tested in vineyards for shoot thinning. Some of them only focus on removing suckers from the trunk (Clemens Vineyard Equipment Inc., Rotary Brush), while others can remove both shoots in the cordon and suckers in the trunk (Oxbo Cordon Brush: Model 62731; Vine Tech Equipment, Shoot Thinner). However, based on tractor ground speed and the thinning head speed (number of head rotations per minute), cluster removal may vary between 10% and 85% (Dokoozlian, 2013). Such variability generates an issue of either too many shoots being removed or not being removed sufficiently. In this project, we developed a technique for automated positioning and orientation of the thinning heads for precise operation of the machine, which is expected to improve usability and commercial adoption of shoot thinners. We have successfully developed the artificial intelligence-based machine vision system that can identify the location of the cordon, trunk, and shoots in the real vineyard environment. This machine vision was integrated with the 3DoF (Degrees of Freedom) prototype machine for lab and field evaluations. Based on the laboratory and field evaluations of the integrated systems, the system has potential to estimate cordon trajectories and move the thinning end-effectors to the desired position with precision in the vineyards. This study proved the proposed concept, and the developed technology is ready for being transferred to commercial entities for adoption.
Read more by downloading the full report above.