Agricultural picking robot for intelligent robot application case
At present, the main focus of research on picking robots is on the recognition and localization of fruit targets by visual systems. Cameras capture fruit images, which are then processed using complex image processing algorithms to enable the robot to logically identify and pick fruits automatically. These robots have strong autonomous identification capabilities, allowing them to operate without human intervention. This makes them ideal for agricultural applications, although the technology is still not fully mature and requires significant investment. To address this, the design here adopts a human-machine collaboration approach: humans identify the fruits, while the robot performs the actual picking. The robot is remotely controlled via a wireless remote control, enabling manual observation and operation in the field. This method leverages existing, well-established technologies, reducing development time and cost. Although it cannot completely replace human labor, it significantly reduces the physical workload and is more suitable for current agricultural conditions in China.
In response to these challenges, this paper presents an analog picking robot based on the ATmega32 microcontroller, capable of performing manual mechanical picking. An infrared arm is used to extend the end gripper to the target fruit, completing the grasping task.
The overall design of the robot integrates mechanical manufacturing techniques, electronic circuitry, automatic control, and sensor detection technologies. Sensors and infrared remote controls are connected to the main control board, which processes signals to control a three-degree-of-freedom robotic arm and a crawler chassis. The system’s block diagram is shown in Figure 1.
The robot operates in a direct manipulation mode, where the operator sends commands through a remote controller to move the robot forward, steer left or right, control the robotic arm, and manage the gripper's rotation, tension, and closing. The designed robot features a simple structure, rich functionality, and strong expandability.
The mechanical design includes a two-degree-of-freedom mobile carrier and a three-degree-of-freedom robotic arm with a gripper. The robot body is made from mesh aluminum and engineering plastics, making it lightweight and easy to modify. The crawler chassis is driven by four FAULHABER motors, and the robotic arm uses MG995 servos for precise movement. Servo motors control the gripper's opening and closing, the gripper's rotation, and the arm's vertical movement.
The hardware design uses the ATmega32 microcontroller as the core due to its high performance, low power consumption, and real-time capabilities. It has 32KB of Flash memory, 32 general-purpose registers, and supports advanced RISC architecture, offering high efficiency and flexibility.
The control board includes power, crystal, communication, motor drive, remote control, and input/output modules. It features 8 input and output interfaces, 4 DC motor outputs, ISP programming, and IR receiver ports. Each module is plug-in compatible, allowing easy expansion and use.
A USB-to-UART download circuit using the CP2101 chip enables communication between the AVR microcontroller and the PC. The infrared remote control uses the BL35P12 MCU for key scanning and signal encoding, allowing precise control over the robot’s movements and actions.
Software programming involves structured and modular code written in AVR Studio 4. It includes subroutines for servo motor control, remote reception, sensor processing, and DC motor control. The program is optimized for I/O handling, PWM speed control, and interrupt management.
Testing showed that the robot can move at 0.5 m/s, climb up to 45 degrees, and be controlled within 3 meters. The robotic arm was tested for servo angles, and the final tests confirmed the robot's stability and adaptability in orchard environments.
The modular design allows flexible picking modes and scalability, improving efficiency. The robot is lightweight, reliable, and user-friendly, meeting the expected design goals.
While this is a preliminary exploration, future work should focus on improving accuracy, integrating machine vision, and adding sensors to prevent fruit damage. Overall, the robot offers a practical and cost-effective solution for agricultural automation.
Solar Panel,,High Efficiency Mono Solar Panels,Bifacial Jinko Solar Panels
PLIER(Suzhou) Photovoltaic Technology Co., Ltd. , https://www.pliersolar.com