Liangliang Yang

Leader-follower system using two robot tractors to improve work efficiency

A leader-follower system using two robot tractors was developed for field work.This system was designed for practical application.The experiment results showed the two robots can work safely to complete the work.The systems efficiency improved by 95.1 percent compared with using a single robot. Two robot tractors were used in a leader-follower system for agricultural field work. Each of the robots is fully independent and can conduct field work alone. They can also work together to form a certain spatial arrangement during the operation. During the headland turn, to make the best use of headland, the two robots coordinate to turn to next path and do not keep the spatial arrangement. Each robot is simplified as a rectangular zone, and the two robots cooperate and coordinate to turn to the next path without collision. This system is designed for practical application, and the system gains the ability to tolerate most of the disturbances in a real field. Fault tolerant methods in accordance with agricultural work are illustrated to solve the common disturbances from the GPS and the IMU. Field experiments were conducted to determine the effectiveness of the system. The results of the experiments showed that the two robot tractors can work safely together to complete the field work. The average lateral error of the navigation system of the robots was less than 0.04m, and the efficiency of the leader-follower system was improved by 95.1% compared with that of a conventional single robot.

Development of a Wheel-Type Robot Tractor and its Utilization

Abstract This article described the development and utilization of a CAN-bus based robot tractor. A RTK-GPS and an IMU were utilized as a position and a posture sensor for the development of a navigation system. The path planning and navigation control method were described in this article. Three field tests were conducted to show the performance and stability of the robot on straight path, logged path following crop rows and long time work. The RMS value of lateral and heading errors of three tests were 0.05 m and 0.6 degree, 0.02 m and 0.77 degree, 0.04 m and 0.75 degree, respectively. The results revealed the CAN-bus based robot tractor has high and stable navigation accuracy. The accuracy was acceptable for all tests.

Development of a Robot Combine Harvester for Wheat and Paddy Harvesting

Abstract Currently, the number of farmers in Japan is decreasing at an especially high speed. In addition, the average age of the farmers available is increasing. In this trend, it is entirely possible that several decades later, the self-sufficiency rate of agricultural product in Japan will drop sharply. To solve this problem, one possible solution should be to employ agricultural robot in actual use, including robot tractor, robot combine harvester and so on. Recognizing the importance and urgency of developing such agricultural robots, this study proposed and developed a robot combine harvester. This robot relies on an AGI GPS receiver and IMU for position and posture data. It is controlled via CAN BUS. In this study, its accuracy was tested, and it shows that this combine harvester can be applied in Honshu and Hokkaido area of Japan with this accuracy.

An Active Safety System Using Two Laser Scanners for a Robot Tractor

Abstract A 2D laser scanner was used for obstacle detection for autonomous vehicles for its high accuracy at both indoor and outdoor environment on the measurement of distance. Traditionally, the obstacle detection method was depended on the distance from an object to a laser scanner, and an obstacle was defined if the measured distance was less than a threshold. A shortage of the traditional obstacle detection method for a robot tractor is that a low obstacle cannot be detected when the height of the obstacle is lower than the height of the scanning plane of the 2D laser scanner. In this article, obstacle detection method using two 2D laser scanners is discussed. One laser scanner is mounted at the front and the other is mounted at the rear of the vehicle in order to cover the whole surroundings of the vehicle. The laser scanner was mounted look-at-down in order to detect a low obstacle. A test was carried out at an outdoor environment to verify the developed safety system. The test results indicate the developed system could detect obstacles with accuracy around 10 cm.

Apple Internal Quality Classification Using X-ray and SVM

Abstract Classify the apples internal quality is important for apples quality. Mould core apple is not eatable and can reduce the juice quality if used for making juice. The core tissues color is brown and the density of mould core apple is lower than a normal apple. This paper describes the detection of apples one internal quality (mould core disease) on the physical basis – density. X-ray technology can provide us internal image of an object that reflects the objects density. This paper presents the method of processing the X-ray image and creating the feature vector for using support vector machine (SVM) method to classify the apples internal quality. The experiment results show that the classification method could detect the mould core apple at the accuracy of 91.03%, and detect un-mould core apple at the accuracy of around 95.95%.

Development of a Human-driven tractor following a Robot System

Abstract Automated agricultural equipment such as a robot tractor has been developed to solve the problem of agriculture labor force shortage. A human-driven tractor following a robot system is useful for saving time on a large-scale farm. In this kind of system, a robot tractor does farm work while a human-driven tractor is following and doing a different operation. The human operator can control the robot through a controller. To monitor work conditions of the leading robot tractor, four cameras are mounted on the robot tractor, and images from the cameras are sent to the human-driven tractor through a video transmission system. The human operator can check the surroundings of the robot from a monitor. Also, a laser scanner is mounted in the front of the robot for safety. The results of an experiment using the system showed that the precision of the robot tractor, which had an RMS error of about 0.03 m, was better than that of an experienced tractor operator, for which the RMS error was about 0.04 m, while the range of lateral errors of the human-driven tractor improved from 0.25 m to 0.20 m by following the robot.

Development of Robot Tractor Associating with Human-drive Tractor for Farm Work

Abstract A single agricultural robot was developed to solve the problem of labor shortage; however, it cannot meet the demands as the farmland is becoming larger and larger. This article focused on a robot tractor associating with human-drive tractor system. A robot tractor is doing farm operation while a human-drive tractor following and doing different operation. Human can adjust the parameters of the robot by a microcontroller. To ensure safety, four cameras were mounted around the robot tractor, and the images of the cameras were sent through wireless video transmission system. Human can keep a watchful eye on the surroundings of the robot tractor. The results of the system indicated that the performance of the robot tractor behaves better than an experienced farmer driver, in the mean while the efficiency is increased compared with a single robot system.