SERICA: Sericulture Evaluation and Real-Time IoT-Based Cocoon AI-Classification

Abstract

SERICA, which stands for Sericulture Evaluation and Real-Time IoT-Based Cocoon AI-Classification, is an automated system developed to assist the Sericulture Research and Development Institute in sorting silkworm cocoons. The system was designed to reduce the time, effort, and inconsistency involved in manual cocoon sorting. It uses a Raspberry Pi 5 with camera modules and a Convolutional Neural Network model to classify cocoons as good or defective, while an ESP32 microcontroller controls the sorting mechanism. The system also includes real-time monitoring through a dashboard and SMS notification. Based on the evaluation, SERICA results showed high effectiveness, achieving an accuracy of 91.65% to 94.25%. This means that SERICA can be trusted to sort cocoons reliably. However, some mistakes still occur, indicating that the system can be improved. The users found it very easy to use and suitable for real-life application, giving it a rating of 4.61, which is considered very high. Thus, it reduces the time and effort required from workers, and helps ensure that the cocoons are sorted accurately. It improves the efficiency and sustainability of the silk production process by using technologies such as the Internet of Things and artificial intelligence. The SERICA system is beneficial to the people who work with silkworm cocoons at the SRDI.