SLED (Shelf Life Expiration Date) Tracking System

A solution to the food waste crisis, the SLED tracking system predicts food spoilage using ML, supporting a sustainable environment. 

Food waste is a global issue that results in heavy economic losses and the emission of greenhouse gasses into the environment, worsening the climate crisis the world is already facing. In fact, if food waste was classified as its own country, it would be third in global carbon dioxide emissions, only behind the US and China. Food waste poses economic risks as well - substantial resources are wasted when edible food that could be used to feed others is thrown away. For example, throwing out a jug of milk may not seem to pose a significant problem. However, producing milk requires large areas of agricultural land to support cows. According to Move For Hunger, a third of this land will end up growing wasted food, often replacing the habitats of wild animals, another critical environmental concern often referred to as the greatest threat to animal species. Resources are wasted during production processes and transportation as well, in the form of oil. Every year, millions of gallons of this commodity are used to process and bring milk to grocery stores. Oil, a non-renewable resource, emits a considerable amount of carbon dioxide into the atmosphere, further worsening the problem of global warming. Saving a single glass of milk, or any other food that could have been potentially wasted, can play a part, however small, in saving our planet. 

Although large scale losses do occur, food waste predominantly arises at the consumer level because consumers throw out food prematurely, believing it has expired. Conversely, consumers also become sick by consuming expired food. My passion for this issue stems from a personal incident, as I experienced food poisoning from milk that I had not checked the expiry date on. Becoming sick from consuming expired food is an unpleasant experience that can sometimes turn into a life-threatening situation. While researching ways to prevent this from happening again, I realized that there were almost no ways for me to check if my food was spoiled in an organized way so that I could avoid getting sick while preventing food waste. As a solution, I created the Shelf-Life Expiration Date (SLED) tracking system, which uses machine learning to more precisely predict when foods will spoil and remind consumers to consume food by the predicted spoilage date. To make it user-friendly, I created an app allowing the user to log their food and receive immediate predictions on when it will spoil, becoming increasingly accurate as app usage increases. Moving forward, the key to solving environmental and health issues lies in technological innovation coupled with public awareness, in order to effectively reach as many lives as possible. I believe that implementing technology to combat food mistreatment can do so much more than simply improve consumer habits - it can promote sustainability and protect our world. 

The SLED is a multi-faceted system composed of an app and test kit that runs on machine learning algorithms to allow consumers to take control of how they consume their food. To supply data for my algorithm, I created a dataset for training on using self-recorded observations across numerous trials. Through certain inputs about the sensory characteristics of the food, my algorithm can predict how far along the “spoilage spectrum” the food is, ensuring that food is consumed at the right time. To make it user-friendly, I created an app allowing the user to enter their food and receive immediate results, increasing in accuracy as app usage increases. Users can input their characteristics to see a simple visual graph showing relevant information about their food, but I am also working on an imaging approach that can simply detect features of spoilage, eliminating the need to manually input characteristics. I’ve also developed a test kit that allows users to calculate the spoilage of their food without technology from a mathematical model that I created, allowing people from all walks of life to experience my solution. I believe that implementing technology to combat food mistreatment can do so much more than simply solve a problem - it can improve consumer habits and protect our world. 

This target audience for my solution is people of any age - food waste is a problem that affects people of various backgrounds. My solution is one that targets the everyday consumer, and in the realm of food, very few solutions exist for consumers to test for spoilage in a compact, easy-to-use way. The SLED would have a significant impact on the way consumers treat their food. 

Another one of my main goals with my solution is to promote inclusivity. This pushed me to create a test kit that the consumer could use to store all of the necessary materials and also record their observations. Not only does the test kit make the observation process more efficient by not requiring the user to find their own materials, it also serves as an alternative to people who may not have access to the technology required for the app. Food waste is a problem that affects consumers everywhere, and the addition of a test kit ensures that everyone can measure and check if their food is spoiled, with or without technology. The main reason this solution is the most practical was that a test kit promotes consumer responsibility. I wanted the consumer to look at the foods themselves so they could gain an understanding of what exactly makes a food spoiled or not spoiled. Linking an app with the test kit promotes a system in which the consumer can control their own food, and this leads to a change in consumer habits that is necessary to solve the problem of food waste.

The summer before I started this project, I experienced food poisoning from expired milk that I had not checked the date on. This incident made me very sick for a few days, and I wondered if this was a common occurrence - after all, it was a simple mistake. Becoming sick from consuming expired food is an unpleasant experience that can sometimes turn into a life-threatening situation. Luckily, I was able to get better within a few days, but the dangerous bacteria that thrive in our bodies when we ingest spoiled food can cause a variety of symptoms and physical harm to most people. After recovering from my food poisoning,, I started to explore if there was anything I could do to prevent it from happening again. There were almost no ways for me to check if my food was spoiled in a compact, organized way, so that I could eat it without getting sick and prevent food waste at the same time. However, I also realized that food waste is a two sided problem: consumers are throwing out food prematurely and becoming sick if they ate it too late. This results in not only heavy economic losses, but also the emission of millions of tons of greenhouse gas methane into the environment, worsening the climate crisis the world is already facing. 

As someone who personally experienced the issue that my project aims to solve, I am confident in the impact my solution will have on people around the world, and I cannot wait to bring it to light. 

This project is very important to me because it helps me serve my local community. My school runs various food drives throughout the year, and I have held bake sales to raise money and awareness. I wanted to do more about this issue because we often take food for granted. In fact, according to the World Food Program, all of the food we waste every year is enough to feed 2 billion people, twice the number of malnourished people in the world today. I believe that in order to solve this problem, awareness needs to be created and consumer habits need to be changed from the ground up. 

 After creating the basic app version, I tested my product with potential users to ensure that my design was easy to navigate, and I also collaborated with others to receive valuable feedback about my product. Although the final solution was one that underwent countless iterations, it was made possible through identifying key aspects to program for, as well as considering various alternatives to combine practically. 

The applications of this project are not limited to what it is now. In the future, I hope to expand my project so that consumers can use my app to donate food they will probably not eat in time so that people who need this food can receive it. I believe that the core of this project is giving back to the community, and it is the original reason I was inspired to perform it. 

My solution is innovative in several ways, as it is the first system to specifically focus on preventing food waste on the consumer level. Many existing trackers are built for larger companies and industries that need to manage food on a massive scale, but compact systems that do not require specialized training to operate currently do not exist, especially for the consumer. Furthermore, the application of machine learning algorithms to predict food spoilage is novel, as it takes advantage of the power of large datasets that can be collected from the widespread occurrence of food spoilage. Although I hope that my product will become a leader in the consumer food tracking space, I also hope that it will inspire new innovations to combat the effects of the larger problem my product aims to tackle: climate change. 

The goals of my project were to develop a successful app and a test kit that could predict the spoilage of a food accurately based on sensory characteristics rather than the best-by labels on food products.

The app was successfully powered by a machine learning program that accurately predicted the spoilage of foods using sensory observations. This required Swift, the programming language used in apps in iOS devices. I met this goal by learning Swift and designing a GUI that performed the functions the app was intended to. This was done after the experiment was performed so that the data and observations could be inputted into the app. Within the next 3 months, I plan to publish this app and test it with my local community to receive feedback and raise awareness about the issue of food waste. 

Another goal of my project was to develop the test kit.  The test kit was designed after experimentation to include only the necessary materials. These materials included flashlights, pipettes, and other measuring devices. Most importantly, the test kit contains a checklist that the consumer can reference to measure the spoilage of their food on their own, educating them on how the spoilage of a food is calculated experimentally. Not only does the test kit serve as a way to measure the spoilage of a food, it also serves as a supplement for people who may not have access to the technology necessary for the app. This required the test kit to be affordable and scalable so it can be accessible to everyone. To achieve this goal, I included materials that were inexpensive and available to scale. I am currently working on reducing the cost of the test kit so that it can be distributed widely. 

This project explored the issue of food waste by examining the spoilage of different foods over time based on sensorial observations. I conducted my own experiments that tested the spoilage of foods that people either waste the most or become sick from most often. After some research, the five foods I decided to test were bread, milk, eggs, bananas, and leafy greens. Each food was assigned 3-5 characteristics that could be measured to determine if the food was spoiled or not. Based on an estimate of how long these foods were known to stay safe to eat for, I recorded observations on these foods over time, noting how these characteristics changed and when it was no longer safe to eat these foods. When handling potentially unsafe food, I needed to learn how to handle it safely. This helped me learn about the processes involved in food safety and how my community treats food safely. 

After all of this data was collected, I programmed a machine learning algorithm that calculated a “spoilage score” for each food that was tested. These scores were then plotted against the day that they were observed to reveal a regression graph. After calculating the line of regression, I allowed the user to input their observations to check if their food was spoiled or not. The program outputted the number of days the food likely had left in addition to whether the food was spoiled or not. To present this algorithm in a more user friendly manner, I created an app that would allow the user to submit their observations and receive immediate results that would become more accurate as app usage increased over time. In addition, the app also serves as a reminder system that alerts the user when their food is close to expiring so they can consume it before it is unsafe and needs to be thrown away. However, I realized that I needed to have a way for the user to test all of the characteristics that were used in the experiment, because some required additional materials. This pushed me to create a test kit that the consumer could use to store all of the necessary materials and also record their observations. Not only does the test kit make the observation process more efficient by not requiring the user to find their own materials, it also serves as an alternative to people who may not have access to the technology required for the app. Food waste is a problem that affects consumers everywhere, and the addition of a test kit ensures that everyone can measure and check if their food is spoiled, with or without technology. The main reason this solution is the most practical was that a test kit promotes consumer responsibility. I wanted the consumer to look at the foods themselves so they could gain an understanding of what exactly makes a food spoiled or not spoiled. Linking an app with the test kit promotes a system in which the consumer can control their own food, and this leads to a change in consumer habits that is necessary to solve the problem of food waste. 

My solution is currently in the prototype phase, as I would like to make sure that it is accurate and consumer-friendly before I bring it to market. So far, I am testing my prototype with friends and family (about 10-15) to gather feedback that can improve the functionality of my product. Within the next year, I hope to expand the reach of my product to my school and community to 50+ people. Farther into the future, I would like to market my product through social media and other avenues to further my product’s usage. 

One of the most notable barriers that I currently face is assimilating large amounts of data into a compact system. Since data feeding the algorithm increases with usage over time, too much data causes slowdown, and sometimes the crashing of the app itself. To avoid data overloads like these, I am experimenting with different approaches to handle data within the iOS framework. Another barrier I face is the normalization of data. Evidently, one consumer’s banana is different from another’s, as it is dependent on the source and handling of the food. I am finding ways to collect data such that the outputs are reflective of the specific food, and image recognition is proving to be a viable avenue to solve this problem.

I currently do not partner with any organizations, but I hope to partner with food banks and food pantries in the near future. My vision for the future is a system in which users can donate food that may not be consumed before it is wasted to areas where others can benefit from them, creating a community in which everyone is benefited. 

My business model is currently centered around two strategies. The first is my app, which provides a way for consumers to track the spoilage of their food, along with serving as a reminder system. I would like to make my app free of charge initially, to expand the reach and impact of my product. The second strategy is my test kit, which is a physical component that can be used to determine spoilage using simple mathematical models that do not need technology. The test kit serves as a supplement for people who may not have access to the technology necessary for the app. This required the test kit to be affordable and scalable so it can be accessible to everyone. To achieve this goal, I included materials that were inexpensive and available to scale. I am currently working on reducing the cost of the test kit so that it can be distributed widely. 

I plan to fund my project by appealing for grant programs. In order to widely distribute my product, I will need to perform more research testing a wider variety of foods, and securing a grant will be instrumental in providing me the facilities and funding to add necessary data to my algorithm. To cover costs that may occur during this process, I will continue to sell my app and test kit to bring in revenue. I believe that the primary avenue to reach financial sustainability is to achieve healthy growth year over year - therefore, increasing the reach of my product through low cost measures such as social media will be essential to keeping costs low and maximizing revenue.