Semi-automated robotic arm for oil palm fresh fruit bunch harvester

Over the years, as the oil palm plantation grows bigger, more labourers are required to handle all the work. Over time, labor cost has increased so much that most operations in the field has to be mechanized. Mechanization is commonly recognized as a means of solving increasingly acute shortage of labor in the plantation sector. Mechanized harvesting was attempted with major achievement in the area of mechanized in-field FFB (fresh fruit bunch) collection namely the Mechanical Buffalo, Compact Transporter, and Crabbie. The existing mechanized oil palm harvester located at Malaysian Palm Oil Board (MPOB) Bangi Lama, is inefficient in harvesting fresh fruit bunch. Based on the experiment performed before the automation process, the operator of the harvester takes a long time (3-5minutes) just to adjust the position of the cutter and grabber to perform the harvesting process, compared to a worker who manages to harvest a tree in just a few seconds. Not only was the operation taking a long time, but the operator experiences neck aches and body pain after operating on only one tree. So, the ergonomic of the operator was also an issue here. As a solution to the inefficiency of the harvester, automation of the harvester was carried out. The novelty of this research is to transfer the image of FFB to the Denavit & Hartenberg (D-H) model in determining the position of the FFB to be harvested. The D-H model is a simple way to portray robot links and joints and was used for the harvester configuration. Kinematic analysis was calculated based on the D-H configuration for the position and orientation of harvester arm, where position of harvester can be calculated instantly when all joint variables were known. Consequently, in order to place the harvester arm in a desired location, the amount of each joint movement was calculated through the inverse kinematic analysis. This was possible with the information of the position of the harvester arm with the help of High Resolution Webcam. The webcam feeds the desired position coordinates in the form of pixel which was later converted into meters and was used for the inverse kinematics calculation to obtain the desired angle for the harvester arm movement. The image location and calculations were carried out through Matlab with help of the operator to click on the desired position on the screen. The webcam provides the x and y axes while the ultrasonic sensor provides the z axis measurements. Once the calculations were completed, information for each hydraulic cylinder was transferred to the Programmable Integrated Controller (PIC) controller to be sent to the respective manipulator’s hydraulic cylinder. The PIC was suffixed on to a PIC Circuit Board (PCB) with corresponding relays where two relays were assigned to control one solenoid valve. Thus the cylinders move from their home position to the desired position and clamps on the FFB. Then, the operator will cut the FFB manually using the lever since the cutting system is not efficient to be controlled automatically. This study benefits the oil palm industry by increasing the efficiency of the harvesting process by introducing automation of the oil palm harvester manipulator. By automating the harvester, the work cycle time of harvesting process was reduced, thus indirectly improving the productivity of oil palm harvesting process. The semi-automated harvester was proven to move its arm around 60 to 70 percent faster compared to the mechanical harvester. A successful design, PCB fabrication, testing and implementation of concept of camera vision operation system for FFB harvester with a fully develop a graphical user interface (GUI) for outdoor agricultural activities was achieved