Gas Sensor-Based Nondestructive Testing for Jackfruit Ripeness Level

Abstract

Jackfruit is a fruit native to South and Southeast Asia that is now cultivated all over the world due to its adaptability, as it can be eaten both ripe and unripe. This study was conducted to accurately identify the Jackfruit’s ripeness. Four MQ families have been identified as the most effective for classifying the ripeness of Jackfruit based on the sensitivity of the sensors. MQ Gas Sensors such as MQ3, MQ5, MQ7, and MQ135 are used to determine the ripeness of the Jackfruit without cutting it open. Jackfruit emits an aroma, which can be used to determine the ripeness of the fruit. The data is collected using 36 Jackfruits, of which 12 Jackfruits each classification. This study evaluates the performance of a device designed to classify Jackfruits into three categories based solely on the gas concentration in their aroma: unripe, ripe, and forcibly ripe. As a result, the system can determine the level of ripeness of the jackfruit. Unripe jackfruit received a score of 1 for precision, recall, F1-Score, and accuracy. Meanwhile, ripe jackfruit received precision, recall, F1-score, and accuracy scores of 0.83333, 1, 0.9091, and 0.9444, respectively. Precision was given a score of 1 for forcibly ripe, while recall and F1-score were 0.8571 and 0.9231, respectively. Similarly, the precision score for forcibly ripe was 0.9444. The device’s overall accuracy rate was 96%, indicating that it distinguished the ripeness level of the jackfruit with high precision. The results of the study provide producers and consumers with a good course of action, allowing them to produce and consume fruit more efficiently and effectively. This not only minimizes food waste, but also ensures that the fruit is consumed at its optimal ripeness, leading to greater consumer satisfaction.