Abstract
The intake of the perishable fruits and vegetables (FVs) in the human diet can contribute to reduce the risk of some chronic diseases. But unfortunately, FVs loss rate is high among all the food produced annually and occurs at storage stage of post-harvest life cycle. One of the key factors contributing to this high loss rate is inability to gauge vital ambient environmental parameters in cold storage. The existing monitoring solutions about cold storage are limited to only gauge temperature, relative humidity and ignore other vital ambient environmental parameters such as luminosity and concentration of gases. This is a critical issue that needs to be addressed to overcome the loss rate of FVs. This paper presents a real-time intelligent monitoring and notification system (RT-IMNS) banked on an Internet of Things (IoT)-enabled approach for real-time monitoring of temperature, relative humidity, luminosity and concentration of gas in cold storage and notifies the personnel on exceeding of dangerous limits of these parameters. Moreover, decision support is implemented in the RT-IMNS using Artificial Neural Network (ANN) with forward propagation to classify the status of commodity into one of three classes i.e. good, unsatisfactory or alarming. The proposed prediction model outperforms Compress Sending (CS), Adaptive Naïve Bayes (ANB), Extreme Gradient Boosting (XGBoost) and Data Mining (DM) with respect to forecasting accuracy. We achieved 99% accuracy using forward propagation neural network model while existing models such as CS, ANB, XGBoost, DM achieved 95.60%, 87.50%, 93.59%, 90% accuracy respectively. Moreover, proposed approach achieved 100% precision, 100% recall, 100% F1-score for good class is achieved, for unsatisfactory class precision is 98%, recall is 99%, F1-score is 98% and for alarming class precision is 100%, recall is 98% and F1-score is 99%.
Highlights
According to FAO (Food and Agriculture Organization), 1.3 billions of tons per year food loss is reported which represents 33% of the total production [1]
MATERIAL AND METHODS In order to reduce the loss of perishable fruits and vegetables (FVs), a cost- effective Internet of Things (IoT)-based solution is presented in this paper that automatically monitors real-time environmental parameters like ambient temperature, relative humidity, light intensity and concentration of CO2
According to World Health Organization (WHO) recommendation, minimum 400g FVs per day should be taken in the human diet to reduce the chances of chronic diseases
Summary
According to FAO (Food and Agriculture Organization), 1.3 billions of tons per year food loss is reported which represents 33% of the total production [1]. The loss of FVs which occur at storage stage is 10% that is higher than all other stages [6] includes harvesting, storage, processing and distribution of post-harvest life cycle [7] These convincing evidences indicate that a critical measure should be immediately taken to reduce the loss of perishable products like FVs. To overcome the loss of perishable FVs, it is vital to monitor ambient environmental parameters that affect the quality of FVs. Traditionally, low temperature is considered as primary factor in cold storage to slow down the ripening process of FVs that does not stop even after picking and continues up to over-ripe or rot of FVs. Traditionally, low temperature is considered as primary factor in cold storage to slow down the ripening process of FVs that does not stop even after picking and continues up to over-ripe or rot of FVs It causes postharvest pathogens like bacteria, fungi etc. Except temperature [8] and relative humidity[9], there are many substantial environmental parameters such as, concentration of gases (CO2 [10, 11], O2, C2H4 [12]), light intensity [13], [14], dust that affect
Sign-in/Register to view highlights & summary Sign-in/Register to access full text options 
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a call
