Smart Containers for Leftover Food Tracking for Packed and Unpacked Food ^†

Abstract

Due to busy schedules and a lack of tracking of food that is stored, a lot of food is wasted every day in households. According to the UNEP Food Waste Index Report 2021, India’s household food waste amounts to 50 kg per person per year, or over 68 million tons. In 2023, India would have produced over 68 million tons of food waste. Food waste is rising quickly every year. The following variables will affect food loss and waste (FLW) at the consumer level: improper food storage, including not using it before it goes bad. Partially used ingredients, preparing meals beyond necessity, and poor visibility of food in freezers are the main causes of food spoiling at home. According to data from the Food Safety and Standards Authority of India (FSSAI), one-third of India’s food is wasted or spoils before it is eaten. This can be minimized by tracking food in smart containers, which work for both packed and unpacked food. This can be used with or without a refrigerator.

1. Introduction

There is a huge wastage of food in refrigerators across the world due to inefficient tracking methods and lack of visibility. Among all these issues, the smart container makes use of sensors and IOT technology and helps advise users to consume the food before it spoils. Although there are some models which reflect the smart container ideology, they were not of sufficient quality. Using only weight sensors does not meet the needs of the customers. On the other hand, our smart containers use multiple technologies and contain the features most needed by users.

The smart containers use IOT technology. The Internet of Things (IoT) connects physical objects with the virtual world. Smart devices and machines are connected to each other through the Internet using IOT technology. They record information about their direct environment with the assistance of the sensing elements (sensors), then analyze and link it and make it available in a network.

2. Literature Review

A.E. Gallo conducted research on consumer food waste in the US in 1980; the study focused on food measurement using the dairy method. The fruit may be identified by its color and texture, according to 2010 research [1]. Internet Refrigerator, 2015: An Internet of Things (IoT) typically consists of sensors, control units, Bluetooth/Wi-Fi, RFID tags, and communication modules for real-time food monitoring in the refrigerator. The 2016 paper “Determinants of consumer food waste behavior: Two routes to food waste” in [2] includes information on food waste based on buying habits and leftover reuse practices. In 2018, we saw the release of an IoT-based novel smart refrigerator designed to reduce food waste by Gaurav Anand and Lucky Prakash. The refrigerator is equipped with an Arduino AT Mega 2560 board, a camera, sensors, and RFID tags. When the door is opened, it takes pictures of the food using the image recognition system. The IoT-Based Food Wastage Management System [3] makes extensive use of IOT devices and sensors. Weighbridges are used to measure the weight of trash boxes containing food waste, and the weight of each trash box is saved in the cloud. In Meo’s Electronic Nose for Fruit and Vegetable Spoilage Monitoring and Detection in the Refrigerator Using Principal Component Analysis, Marian N. Fernando Carrillo, Roselle, Febus Reidj G. Cruz, Vincent C. Caya Mark B. Malonzo, Marian M. Lafuente, and Wen-Yaw Chung show that the fruits and vegetables within the refrigerator can be identified by the electronic sniffing system as to whether or not they are spoilt. The PCA Algorithm and KNN classifier are two methods the system uses to classify spoiling.

Silo and Ovie are two examples of smart containers. Ovie uses electronic tags that are affixed to the containers and each tag is linked to an app. It has three LED lights that change color depending on the food’s freshness—red, yellow, and green [4]. The food in the silo is kept fresher for longer thanks to vacuum seals. Ananya Tadigadapa’s 2022 Smart Containers for Food Storage in Refrigerators: The load sensor in the smart container measures the weight of the food and notifies the user after 48 h if it is left undisturbed. The primary disadvantage was that fruits and vegetables could not be packaged with it.

3. Methodology

The smart container contains Raspberry Pi to control the sensors as well as other electronic parts; it is a very flexible platform for programming where scripts can be written in any language compatible with the devices that are linked, including C, Python, and C++. The load sensor measures the weight. The load sensor is placed at the bottom of the smart container [5]. The load sensor will measure the weight of the food that is kept in the smart container. If the food is not disturbed for 48 h, i.e., the weight is the same, then it will remind the user to consume the food before it spoils. The MQ3 gas sensor has a variable resistor that changes its value according to the gas concentration [6]. The smart container contains the MQ3 gas sensor, which will detect the ripening of the fruits and vegetables. If the fruits and vegetables are ripening, then they will have a rotten smell, and the MQ3 gas sensor will detect it and remind the user to consume the food before it spoils.

For packed food, RFID tags are used. RFID tags are a type of system that uses smart barcodes to identify items. RFID stands for “radio frequency identification”, and as such, RFID tags utilize radio frequency technology. With the use of RFID tags, the packed food barcode will be scanned, and when its expiry date is near, it will remind the user to consume the food before it expires [7].

Wi-Fi or Bluetooth is used to connect devices. Wi-Fi is the most common way to connect a device to the internet without any physical link. Fitted to the device is the Wi-Fi interface that communicates with a wireless router, which provides access to the internet. Raspberry Pi will control all the sensors. The load sensor and MQ3 gas sensor are for unpacked food, and the RFID tags are for packed food. By using the temperature sensors, we can use it for food stored outside the refrigerator, and at different temperatures we can use the smart container. There are many existing devices, but they are not widely used due to cost and other reasons; this smart container can be used inside and outside the refrigerator at different temperatures and different types of food can be stored in this smart container. The existing system contains only a load sensor, which will consider all food similarly, but this smart container contains different sensors and tags to identify the freshness of the different food. The mq series gas sensor and temperature sensor are added to the smart container, which increases the accuracy of detection of freshness of both packed and unpacked food [8].

4. Results and Comparisons

The researchers gathered a total of three samples—mango, sambar, and chip packet—for testing the system. The data acquired from sample testing are used as a reference to test whether it is spoiled or unspoiled.

This smart container is useful for both packed and unpacked food. The gathered food items, like leftover food, fruits and vegetables, and packed food, were tested, and accurate results were obtained.

In the case of leftover food like sambar, the results are very accurate, and it classified spoiled and unspoiled food very accurately, as shown in

Table 1. For leftover food—sambar.

S. No	Human Classification	Smart Container Classification
1	spoiled	spoiled
2	unspoiled	unspoiled
3	unspoiled	unspoiled
4	spoiled	spoiled
5	spoiled	spoiled
6	unspoiled	unspoiled

. Out of seven test cases, all the classifications are correct.

In the case of fruits and vegetables, the results are very accurate, and it classified riped and unripe food very accurately, as shown in

Table 2. For fruits and vegetables—mango.

S. No	Human Classification	Smart Container Classification
1	unripe	unripe
2	ripe	ripe
3	spoiled	ripe
4	ripe	ripe
5	ripe	ripe
6	unripe	unripe

. Out of seven test cases, six test case classifications are correct.

In the case of packed food, the results are very accurate, and it classified expired and not expired food very accurately, as shown in

Table 3. For packed food—chip packet.

S. No	Human Classification	Smart Container Classification
1	expired	expired
2	Not expired	Not expired
3	expired	expired
4	Not expired	Not expired
5	expired	expired
6	Not expired	Not expired

. Out of seven test cases, all the classifications are correct.

5. Conclusions and Future Enhancement

In this study, a smart container is developed with IOT technology with a load sensor, MQ series gas sensor for unpacked food, and RFID tags for packed food, to track the freshness of the food placed in the container and remind the user to consume it before it spoils through the dedicated app. The smart container results were the most accurate.

The aim of this smart container is to develop sensor-based products that help to implement the solutions most effectively. Future improvements should also detect the freshness of the fruits and vegetables by the structure of their rind using image recognition machine learning with neural networks and increase accuracy by using more sensors, like temperature sensors, for monitoring food outside the refrigerator.