An IoT-Enabled Tomato Sorting System with Dual Opposing Ultrasonic Sensors and Multispectral Color Detection for Five-Class

Abstract

Tomato sorting plays an important role in maintaining crop quality, especially in terms of ripeness and size, which affect market value. Tomato sorting is generally still done manually, which is time-consuming, labor-intensive, and often results in misclassification. To address this issue, this study aims to design an automatic tomato sorting device using a TCS34725 color sensor and an ultrasonic sensor integrated with the Internet of Things (IoT) using ESP32 as a remote monitoring system. The research method used is research engineering, which includes hardware and software design. The hardware consists of a DC motor to drive the conveyor, a PCA9685 module to control four servos, a TCS34725 sensor to detect RGB values, and two opposing ultrasonic sensors to measure diameter. The software was built using Arduino IDE, with Blynk IoT integration as a medium for monitoring RGB values, diameter, category, and number of sorting results. The system was developed to classify tomatoes into five categories, namely large ripe, small ripe, large semi-ripe, small semi-ripe, and mixed unripe. Test results show that the sorting tool is capable of classifying ripeness levels based on RGB values and diameter with an average accuracy of 90% to 99% and an average error of less than 10%. Sorting data can be monitored remotely via Blynk, and the system can still operate offline without an internet connection. Thus, this tomato sorting tool is expected to facilitate the community, especially farmers, in the process of sorting tomatoes more accurately and efficiently.