Performance evaluation of automated implement for vineyard mechanical weed control

Abstract

Weed control is an important management practice for sustainable grape production since grass coverage can change nutritional and water balance of vines. Although inter-row grass is used to reduce soil erosion and reduce vine vigour, under-row grass is usually undesirable. Chemical, mechanical, and thermal weeding systems are commonly used for under-row grass control. Over the last few years, the technological advance in vineyard mechanization led to new solution as an alternative to chemical-based weed control. Such systems take advantage of different type of sensors and feeler to manage the tillage implement close to vine trunk. In this study, a sensor-driven tillage system was evaluated and compared to manual management of the same tillage implements for weed control. Mechanical weeding was performed with a rotary star tiller controlled by a laser scanner system (LIDAR). Automated and manual working configurations were evaluated in terms of effective field capacity (ha·h-1) and uniformity of tilled soil from vines trunk. Effects of the automated system on tillage implements were evaluated with accelerometers. According to the result, the sensor driver implement allows in increments of working speed of 29.4%. At the same time, the average variation of tilled soil was 3.8 cm with the automated system and 8.1 with manual management of tillage implements. The results of this study showed that the automated system can be used to increase the field capacity and uniformity of mechanical weeding.