FruitTech: A Cloud-Based Fruits Grading Machine Using Convolutional Neural Network

Ensuring fruit quality is essential for farmers, especially for rambutan and calamansi, two widely grown fruits in the Philippines. However, traditional manual sorting methods are often inconsistent, time-consuming, and prone to errors, leading to post-harvest losses. This study developed FruitTech: A Cloud-Based Fruit Grading Machine, which uses Convolutional Neural Networks (CNNs) to classify fruits based on their appearance and quality automatically. The system was built using an Agile development approach, combining machine learning, cloud computing, and hardware automation. A CNN model was trained to analyze fruit images, identifying characteristics like ripeness, size, and defects. The system also includes real-time sorting mechanisms, SMS notifications, and data visualization tools, giving farmers instant access to grading results. Testing was conducted through alpha and beta phases, followed by an evaluation using ISO/IEC 25010 and UTAUT to assess system performance and user acceptance. Results showed that FruitTech significantly improved accuracy, efficiency, and ease of fruit grading, reducing the need for manual labor while providing farmers with a more reliable and accessible solution. Users responded positively to the system’s functional suitability, security, and usability, confirming its potential for real-world agricultural applications. For future improvements, researchers could expand the system to grade other fruit types, enhance image analysis in different lighting conditions, and integrate IoT or hyperspectral imaging for detecting internal defects. Developing an offline and portable version would also make FruitTech more accessible to farmers in remote areas.