Fruit Freshness Monitoring Employing Chemiresistive Volatile Organic Compound Sensor and Machine Learning

The work highlights the assessment of fruit freshness through the incorporation of a chemiresistive gas sensor and machine learning. For this purpose, SnO 2 nanosheets were synthesized through a low-temperature hydrothermal route. The sensor device was fabricated by integrating a synthesized nanomaterial with interdigitated electrodes. Fruits, including apple, guava, grape, and orange, were taken for analysis, and associated emitted volatiles were considered as the freshness indicator. Fruit samples were stored at room temperature, and systematic sensing measurements were performed at different time intervals (0, 24, 48, and 72 h) to estimate the freshness level. Initially, the responses of the sensor attained from apple, guava, orange, and grape samples were 1.7, 2.5, 6.4, and 2.2, respectively. A neural network-based regression model was employed for the determination of storage duration of fruits in a quantitative manner. The model displayed notable success in prediction, with an average error of less than 10%. Thus, the approach has great potential to develop a smart sensor system to monitor fruit freshness with a real-time outlook.