Development of an Automated Mechanical Intra-Row Weeder for Vegetable Crops

Abstract

<p>Weed management is one of the tedious operations in vegetable production. Because of labor costs, time and tedium, manual weeding is unfavorable. The introduction of chemical weed control methods has alleviated these undesirable factors. However, the emergence of herbicide-resistant weeds, environmental impact and increasing demand for chemical free foods has led to investigations of alternative methods of weed control. Most implements employing mechanical cultivation cannot perform weed control close to the crops, and existing intra-row weeders have limitations. A mechanical weeding actuation system was designed, and a prototype was constructed. This actuator was developed to mechanically control intra-row weed plants. The mechanical weeding actuator consisted of a belt drive system powered by an integrated servo motor and a rotating tine weeding mechanism powered by a brushless dc motor. One of the major challenges in this project was to properly design the actuator and its weeding mechanism for effective intra-row weed control. A prototype actuator was manufactured and a series of tests was conducted to determine actuator efficacy and the corresponding force and speed requirements of the actuator. The actuator would be combined with a machine vision system for detecting crop plant locations and guiding the weeding actuator to execute mechanical weeding operations without damaging crops.</p> <p>In the first field experiment, the performance of the first version of the intra-row weeder was investigated across three factors: working depth, travel speed and tine mechanism rotational speed. There was evidence of differences in weed control efficacy across travel speeds. Using least square means, the slowest travel speed of 0.8 km/h had an average reduction in weed canopy area of 58.2% with standard error of 2.7% compared with the medium travel speed of 1.6 km/h with an average reduction in weed canopy area of 52.6% with standard error of 2.7%. The fastest travel speed of 2.4 km/h had an average reduction in weed canopy area of 42.4% with standard error of 2.7%.There was no statistical evidence of differences in power consumption across working depth, travel speed, or rotational speed. With increasing working depths, reduction in weed canopy area and power consumption tended to increase.</p> <p>With a revised version of the rotating tine weeding mechanism, a second field experiment was also conducted using three factors; tine shape, travel speed and rotational speeds. The results showed that there was no significant difference in reduction in weed canopy area across tine shapes. However, there was some indication that weed control efficacy decreased as travel speed increased. There was evidence of differences in power consumption across rotational speeds. The fastest rotation speed, 536 rpm, had a mean power consumption of 182 W and standard error of 9.4 W. The lowest rotation speed, 350 rpm, had the lowest mean power consumption of 123.5 W and a standard error of 9.4W.</p>