Design and development of a robotic arm for hydroponic crop harvesting

The increasing global demand for sustainable agriculture has accelerated the adoption of automation technologies in farming. Addressing challenges such as labor shortages and operational inefficiencies, this project explores the development of an advanced robotic arm system for harvesting hydroponic...

Full description

Bibliographic Details
Main Author: Azman, Amirul Amin
Format: Dissertation
Language:English
Published: Universiti Teknologi Malaysia 2026
Subjects:
Online Access:https://utmik.utm.my/handle/123456789/190850
Abstract Abstract here
Description
Summary:The increasing global demand for sustainable agriculture has accelerated the adoption of automation technologies in farming. Addressing challenges such as labor shortages and operational inefficiencies, this project explores the development of an advanced robotic arm system for harvesting hydroponic crops. By leveraging innovations in camera vision, sensors, and motion control, the proposed solution aims to enhance the precision and reliability of automated harvesting. Key features of the system include the integration of a Webcam camera module to improve real-time object detection and localization, and a newly designed gripping mechanism to ensure secure and efficient handling of crops. The robotic arm incorporates a five degree-of-freedom (DOF) configuration, enabling flexible movement to adapt to various harvesting scenarios. Advanced software algorithms, including the YOLOv8 object detection model, is employed to achieve high detection accuracy and seamless coordination between components. The methodology involves iterative design, develop, manufacture, assemble, integration, testing, and evaluation to refine the system's mechanical, electrical, and software components. The system was successfully assembled and tested at the subsystem level. Independent testing verified the functionality of stepper motors, servo-driven gripper, AS5600 encoders, ultrasonic sensors, and vision modules. Object detection trials using the pretrained YOLOv8 model yielded satisfactory results with confidence scores up to 0.84 in normal lighting conditions, although recognition accuracy declined with cup variations and poor lighting. While full system automation remains a target for future development, the project lays the groundwork for improving harvesting precision and operational efficiency in hydroponic farming. By addressing key design and integration challenges in robotic vegetable harvesting, the system contributes to the advancement of affordable, adaptable solutions in agricultural robotics. This work supports ongoing efforts to promote sustainable, technology-driven farming practices capable of meeting modern agricultural demands.