Determining spoilage level against time and temperature of tomato-based Filipino cuisines using electronic nose

Every person has different perspective on whether a food is spoiled, and this may lead to wrong assumptions on the food's condition. Misinterpretation on food condition could lead to food-borne illness. The need for electronic nose system for classifying if a food is spoiled is still in demand due to this prediction confusion. This study aims to detect food spoilage of tomato-based Filipino cuisines using an electronic nose. Specifically, this research aims to develop a device with an array of sensors to detect the gases emitted by spoiled tomato-based Filipino cuisines and implement Artificial Neural Network as an algorithm for the classification of the data reading of the sensor. The hardware part of the e-nose system makes use of Gizduino 1281, Raspberry Pi 3 model B, 20×4 LCD Screen, 7 MQ gas sensors, and 1 temperature/humidity sensor. Stochastic Gradient Descent together with Back propagation algorithm is used for training the Artificial Neural Network data. The tomato-based Filipino cuisine is placed below the sensor chamber for easy recognition of the gas sensors emitted by the cuisine itself. The study could be useful in classifying whether food is spoiled. This electronic nose system could determine spoilage level of a specific tomato-based Filipino cuisine. The researchers assigned level 0 to 12 as the spoilage level. These levels correspond to the time when the food is observed every four hours starting 7:00 AM of Day 1. As the experiment continued, it was also that the food often spoiled between levels 5 and 6. Based on the confusion matrix, the error rate of the electronic nose system is 3.85%.