Vol. 83, No. 5
Abstract
Draft Reduction without Increase in Power Requirement Using a Vertically Oscillating Model Subsoiler
[Keywords: subsoiling, oscillatory tillage, draft, velocity ratio, draft ratio, power ratio]
Draft and power ratios of a vertically oscillating model subsoiler was determined. The subsoiler was oscillated at varying amplitudes (5 mm, 10 mm, and 15 mm), frequencies (2 Hz, 3 Hz and 4 Hz) and forward speeds (0.05 m s－1 and 0.20 m s－1) resulting in velocity ratios ranging from 0.3 to 7.5. The mean draft and power ratios stood at 0.65 and 0.89, respectively. Draft ratio decreased with an increase in amplitude but increased with an increase in forward speed. Power ratio reduced with an increase in forward speed but was not significantly affected by frequency and amplitude. Velocity ratio of about 4 gave the optimum result. The main reason inducing draft and power reduction was observed to be soil pulverization and crack propagation caused by the lifting of the soil by the chisel ahead of the shank.
Investigation of Vibration Characteristics of Hand Tractor Based on Frequency-domain and Vibration Intensity
[Keywords: hand tractor, acceleration, frequency analysis, r.m.s value, anti-vibration]
The use of hand tractors on small and mid-sized Cambodian farms has increased remarkably and provides a labor saving and improvement of work efficiency for the operator. The present study investigated the translational accelerations of hand tractors at 7 different measurement points, as well as their engine speeds under a stationary condition. The results showed that the vibration magnitude was the greatest at the top of the engine along the lateral axis at low frequencies, followed by the handgrip along the vertical axis at different frequencies corresponding to the engine speed. The RMS of hand-arm vibration exposure was extremely higher than those stated in the health guidance zone at all engine speeds. Thus, anti-vibration measures should be introduced.
Automatic-Driving Control System to Exit a Garage and Travel on a Farm Road for a Robotic Combine
――Exiting a Garage by using SLAM――
[Keywords: agricultural robot, robot operating system, light detection and ranging, simultaneous localization and mapping, point cloud]
In order for agricultural robots to work more manpower-saving and more efficient, they need not only to work in the fields, but also to automate from exiting a garage to traveling on a farm road. This research aims that a robotic combine exits a garage automatically. SLAM by LIDAR is adopted for localization in the garage. It is tested that the robot can travel from the initial position (about 4 m far and ±0.8 m shifted from side to side) to the exit center at 0.2 m/s based on the estimated poses. As the result, the robot’s pose could be estimated by SLAM at the rate of 10 fps and the robot could travel toward the exit center within±0.12 m error.
A Study on the Process of Farming Accidents in Paddy Fields and Upland Fields
[Keywords: farming accidents, farming safety, bodily injury accidents, property damage accidents, characteristics of accidents]
This study focuses on the process of farming accidents to determine the trend of accidents in paddy fields and upland fields. The patterns of frequent accidents in paddy fields and upland fields included a vehicle collides with a stationary object during tilling and leveling work, which accounted for 17 % of all property damage accidents, and 80 % of the damaged stationary objects included “water supply and drainage equipment” and “walls or fences.” The victims of farming accidents were not limited to farmers; non-farmers were also involved in the accidents and suffered bodily injuries. It is necessary to change the target of caution according to the nature of work to reduce accidents.
Rice Tiller Number Estimation by Field Robot and Deep Learning (Part 1)
――Exploring Infield Tiller Detection with YOLOv4――
[Keywords: crop sensing, deep learning, field robot, precision agriculture, rice tiller, YOLOv4]
The monitoring of rice tiller number is one of the most tedious and the time consuming task of rice cultivation. In this work, we use deep learning as an alternative method to estimate tiller number from images captured by a field robot at three tillering stages: early stage, active stage, and maximum stage, for the two Japanese rice varieties of Fukuhibiki and Haenuki. Three types of YOLOv4 models were trained to estimate the tiller number: models aimed at estimating actual tiller numbers, models trained on classes of grouped tiller numbers, and models trained with classes based on a tiller number histogram. In the experiments, the tiller number histogram based models achieved the highest scores of mAP at the three tillering stages of early, active, and maximum: 62.3, 67.5, and 73.5 for Fukuhibiki variety, 61.3, 63.5, and 49.8 for Haenuki variety.