In the past decades, the emerging concern about food safety has led to the increasing demand for monitoring food quality across the world. Aiming towards a novel solution for monitoring food, this study proposes a non-destructive method with self-powering capability for online food monitoring, which can be extendable to IoT applications. Furthermore, the study introduces a novel deep neural network model to predict different states of food quality based on the monitoring results. To monitor the variation in food quality, the paper proposes the detection of total volatile organic compounds (TVOCs) inside the food packages, which have been released during food deterioration. A low-power sensor mote comprised of a capacity humidity sensor and a metal-oxide (MOX) gas sensor was manufactured for this purpose. The self-powering capability of the mote is provided through an energy harvester module, which benefits from the far-field Radio Frequency Energy Harvesting (RFEH) technology. The operating frequency of the module was chosen at the 915-MHz ISM band. The analysis of the harvester performance showed that the harvester could generate 3.3-V dc with an RF input power of as low as-8 dBm, which was sufficient for the mote operation. To verify the proposed solutions, a demonstration to monitor the deterioration of packaged pork and fish was conducted in eight days under ambient and refrigerated storage conditions, using the selfdeveloped RF-powered sensor mote. The raw variations in TVOCs were analyzed to evaluate the reliability of the proposed TVOC-based method. A one-dimensional (1-D) convolutional neural network (CNN) model was trained on the TVOCs dataset to predict different states of food quality. To investigate the applicability of the proposed 1-D CNN to multi-class determination of food quality, two other supervised machine learning algorithms using 2-D inputs, including Multilayer Perceptron (MLP), and Support Vector Machine (SVM), are studied. Their classification accuracies based on the confusion matrix are identified and compared. INDEX TERMS Convolutional neural network (CNN), food quality prediction, radio-frequency energy harvesting, total volatile organic compounds, multilayer perceptron (MLP), support-vector machine (SVM).

This study proposes a self-powered food monitoring system that benefits from far-field RF Energy harvesting and deep learning techniques. Recent smart IoT systems for food quality management mainly focused on the appearance of Total Volatile Organic Compounds (TVOCs) during food preservation, which has the limitations of large power consumption, complex configuration, and low accuracy. Different from these methods, we aim to measure the gradual increase in air pressure inside food packages caused by gas emissions during food quality deterioration. With this new approach, the designed sensor circuit's energy consumption is reduced significantly than in conventional systems. The sensor module's operation power is supplied by an RF energy harvester capable of converting electromagnetic waves in space into electrical energy. We adopted a Yagi three elements as a receiver antenna in the RF energy scavenging module to enhance the harvested power and transmission distance. The designed antenna has a high gain of 6.54dBi in the direction of maximum radiation and voltage standing wave ratio (VSWR) better than 1.3 in approximately 60 MHz band (890-950 MHz). To demonstrate the feasibility of the proposed system, a set of experiments were conducted with different sorts of food such as pork, chicken, and fish. Raw data of food storage temperature, air pressure, and storage time obtained from the battery-less sensor module were analyzed and utilized to assess food quality changes. Several classification models, including LSTM, 1D-CNN, MLP, and SVM were developed and trained on the pork, chicken, and fish datasets to predict different food quality states. Experimental results show that the LSTM classifier, which can extract temporal characteristics from the dataset, achieves the best accuracy of above 99% on all three datasets. The food classification performance is investigated based on training accuracy and confusion matrix.

No abstract

No abstract