Development of A LabVIEW-based Electronic Nose Using ESP32 And Machine Learning for Food Spoilage Detection

Authors: Suraj Kalyanaraman, Nandish S, Aswin N, Dr Saravanakumar R

Abstract: The electronic nose (E-Nose) provides a rapid and non-destructive method for evaluating the quality and safety of food products. In this, we try to address the design and implementation of a low-cost E-Nose for detecting the extent of food deterioration. The hardware architecture comprises an array of metal oxide semiconductor (MOS) gas sensors, including MQ-135, MQ-4, and MQ-2 sensors, which are used to detect volatile organic compounds (VOCs), carbon dioxide (CO₂), methane (CH₄), and ammonia. To ensure high precision, the ADS1115 16-bit ADC is employed, along with the DHT22 sensor for environmental correction. The array is integrated into an acrylic sealed compartment equipped with an ESP32-S3 Xiao microcontroller and an active cleaning fan. The software architecture is constructed using NI LabVIEW and incorporates producer-consumer design patterns and event-based state machine designs. During classification, differential voltages and slopes are extracted and transmitted through a Python node to principal component analysis (PCA) and k-nearest neighbour (k-NN) classification algorithms. This system demonstrates high accuracy in classifying different thresholds of deterioration and offers the food industry a real-time, scalable, and highly efficient surveillance solution. By enabling the early detection of spoilage, this framework directly contributes to environmental sustainability by reducing food waste and lowering the carbon footprint of the agricultural supply chain

DOI: https://doi.org/10.5281/zenodo.23757881