KeepItFresh

We provide intelligent software powered by AI for grocery stores to reduce food waste by forecasting demand and optimizing store decisions

Photo of Ayush Raj
9 2

Written by

Please confirm you meet the following criteria

  • We have submitted the supplemental form linked in the description above
  • We are aged between 14 - 20 as of February 11, 2021
  • We live in the United States or its territories (Puerto Rico, U.S. Virgin Islands, Guam, Northern Mariana Islands, and American Samoa)
  • We are not employed by, or directly related (parents or siblings) to a current General Motors (GM) or Ashoka employee
  • We have been working on this project for at least three months
  • We consent to Ashoka and/or GM featuring our work on their website, social media, and in other materials regarding this Challenge using the information in our application

Date You Started Your Project

10/01/2020

Project Stage: Select the description below that best applies to your approach.

  • Start-Up (first few activities have happened)

1. The Problem: What problem are you helping to solve?

Retailers throw out eighteen billion dollars worth of spoiled food each year. 8 million tons of food waste from grocery stores each year ends up in landfills. This waste rots in landfills and produces methane, which is 84 times more potent than Carbon Dioxide. Municipal solid waste landfills are the 3rd largest source of human-related methane emissions in the United States. Our goal is to reduce our carbon footprint by reducing grocery waste.

2. Your Solution: How are you planning to solve this problem? Share your specific approach.

We begin by collecting time-series data (data collected over 2-3 years of time) of the sales of various perishables from our customers, grocery stores. Forecast metrics, such as how many days into the future we want to predict sales, items that are on sale, seasonal changes, and weather patterns are identified with input from customers and from readily available data. We then prepare the obtained data for Machine Learning (ML) models by correlating datasets and searching for anomalies that affect the model’s training process. Once the data is prepared, we predict the sales of perishables using an ensemble of ML models and determine the prediction error. The model with the lowest prediction error, identified through a combination of various error metrics, is selected as the best model. We then analyze the results, and based on our findings, optimize the models to forecast sales of perishables. We work with customers to conduct frequent reviews to ensure the models are fully optimized for given perishables. Forecasted outputs are then interfaced to an application that can be interpreted by any grocery store manager.

3. Please tell us how you are using science, technology, engineering or math to address your environmental challenge.

Retail food stores typically sell thousands of different items across hundreds of stores over the course of multiple years. This generates large amounts of time-series data. At KeepItFresh, we work with grocery stores to collect sales data of perishable items and build highly accurate Machine Learning models that use this historical data to predict sales of items n days into the future, where n can be determined by the grocery store’s perishable ordering schedule. Due to the extremely short shelf-life of perishables, over-ordering or under-ordering has a huge financial implication, and therefore, the accuracy of forecasts of perishable items is critical.

4. Personal Journey: What’s the story behind why you decided to start this project?

Seeing empty shelves in grocery stores during the pandemic led us to research and understand how the supply chain works. Further analysis revealed the challenges in managing perishable items in a grocery store, as the food waste produced by these stores has a very negative impact on our environment. With over 40,000 grocery stores across the US, we started to explore if I could use our knowledge of Machine Learning to help reduce food waste caused by our grocery stores. Food waste that ends up in landfills accounts for almost 25% of the US’s methane emissions. Knowing that methane is 21 times more harmful to the environment than carbon dioxide, I was motivated to build an intelligent solution that can help grocery store managers predict the sales of perishables so that they order enough to keep the shelves stocked while making sure that the food doesn’t end up in landfills.

5. Video (Keep it simple, your phone on selfie-mode is great): Please upload a 1-minute video to YouTube that answers the following “I am stepping up to be a Changemaker because...”

6. Please highlight the key activities you have carried out to bring your project to life.

To demonstrate the effectiveness of our idea and build our backend software, we used a publicly available dataset from Kaggle provided by a large grocery chain in Ecuador. We have built baseline, autoregressive, and decision tree models and compared error metrics. For perishables with the highest sales, our results are comparable to state-of-the-art models and we are working on further model optimization. We have reached out to grocery store managers to try out our prototype.

7. The X Factor: What is different about your project compared to other programs or solutions already out there?

We discovered 3 startups, and no patents, that started in recent years with similar concepts to ours. Loosely, these 3 companies all have the same purpose as ours: to use AI-powered software to predict food sales and demand to decrease waste. Our product is aimed for use in small to mid-sized grocery stores whereas companies like Crisp & Afresh aim at serving large grocery chains. Shelf Engine works directly with vendors, so stores lose control over the pricing of perishables. Our software is a simple tool, runs fast, and predicts sales in seconds.

8. Impact: In the last three months, please detail the impact your project has made.

In the last 3 months, we have focused on building our prototype and doing research on how grocery stores handle waste. We have reached out to grocery store managers in our community to help us get access to their sales and test our prototype. We have started to engage our peers in programming clubs at school to help us build a front-end application and do further research. So far we have 6-8 students who are willing to join in to help us build a front-end application that grocery store managers will use. We also plan to engage additional students who are interested in helping us market our product by reaching out to various grocery stores. By bringing together a community of problem-solvers, we are not only building on various skill sets in engineering, sales, and marketing, but we are actively working towards reducing our carbon footprint, one store at a time.

9. What’s Next: What are your ideas for taking your project to the next level?

We plan to complete prototype development and the front end application and get the software ready for customer demonstrations in Q1 and Q2 of 2021. In Q3 '21, the focus will be partnering with 2-3 customers in our local area and running our software on their sales data. In Q4, based on customer feedback, we will make the necessary tweaks to our software and prepare to scale. In 2022, we will heavily focus on winning customers across the US and aim to reach out to about 200-500 stores. We will continue to invest time and money to explore new ML models to cater to the needs of our customers. We also plan to expand our sales, customer engagement, and engineering teams. Finally, we plan to showcase our product at grocer conferences.

10. Please share how you have influenced other young people to get involved in your project and/or care about environmental sustainability.

We have reached out to our peers in various clubs in our school. Students at programming clubs at our schools are excited to join us and hone their programming skills by working on a meaningful and impactful project. We are partnering with 5-6 students in sales and marketing clubs at school to help us reach out to small grocery stores in our neighborhood to educate grocery store managers of our product and partner with us.

11. How would you partner with other changemakers to make a difference?

We would love to partner with changemakers who are passionate about reducing food waste at all levels of the supply chain. Household food waste is yet another huge source of methane emissions. Engaging with changemakers who have built technology to help people plan meals to make grocery shopping efficient will be a great addition to our technology. Also, raising awareness through social media on the impact of food waste is another way we plan to partner with changemakers.

12. How would you engage others who have never heard about your project to get their buy-in?

We plan to showcase our product at grocer conferences and use email marketing to introduce grocery store managers to our product. Using our prototype, we will demonstrate our software to various grocery store managers and run our ML models on their sales data. Our website will have information on how any grocery store can partner with us to cut down their financial losses due to food waste along with the benefits of running an environmentally conscious business.

13. Finances: If applicable, have you mobilized any of the following resources so far?

  • Family support
  • Mentors/advisors

14. Which of the following types of expertise would be most useful for you? You’ll be able to select only one option.

  • Legal Services

Are you employed, or directly related (grand-parents, parents, sibling) to a GM or Ashoka employee?

  • No

How did you hear about this challenge?

  • Other

Referral: If you discovered the Challenge thanks to an organization or person other than Ashoka or General Motors, who was it?

AI4ALL Newsletter

Evaluation results

4 evaluations so far

1. Overall evaluation

Yes, absolutely! - 25%

Probably - 75%

Maybe - 0%

Probably not - 0%

No - 0%

2. CONNECTION to Environmental Sustainability

5 - Absolutely! It is totally clear that the solution is contributing directly to environmental sustainability and/or addressing climate change - 50%

4 — Yes, it establishes a connection to environment/ climate change but could be stronger - 50%

3 — Somewhat, the entry speaks to this environmental sustainability, but the direct impact is not well established - 0%

2 — Not really, the connection to environment/ climate is very weak - 0%

1 — No. The entry does not have a reference the solution’s impact on environment and/or climate change - 0%

No Answer or No Connection - 0%

3. Is this entry CREATIVE?

5 - Yes, absolutely! - 50%

4 - Yes, I think so - 50%

3 - Maybe - 0%

2 - Probably not. - 0%

1 - No - 0%

No Answer - 0%

4. Does this entry demonstrate a COMMITMENT to changemaking?

5 - Yes, absolutely! - 100%

4 - Yes, I think so - 100%

3 - Maybe - 0%

2 - Probably not - 0%

1 - No - 0%

No Answer - 0%

5. Does this entry value CHANGEMAKING through collaboration with other stakeholders in its approach?

5 - Yes, absolutely! - 50%

4 - Yes, I think so - 25%

3 - Maybe - 25%

2 - Probably not - 0%

1 - No - 0%

No Answer - 0%

6. Is this entry VIABLE financially and operationally?

5 -Yes, absolutely! - 50%

4- Yes, I think so - 25%

3- Maybe - 25%

2- Probably not - 0%

1- No - 0%

No Answer - 0%

7. FEEDBACK: What are the strengths of this project?

CONNECTION: You have a great understanding and personal connection to the problem - 66.7%

CREATIVITY: You have researched existing solutions, and have developed unique, thoughtful new solutions to aid environmental sustainability/combat climate change - 66.7%

COMMITMENT: You have a thoughtful plan for growing your business, and your founding team has a strong combination of leadership and knowledge-based skills - 66.7%

CHANGEMAKER QUALITY: You value thinking around how to activate other changemakers and empower them to care about your cause. You also have a clearly defined plan on how to collaborate across multiple stakeholders - 100%

IMPACT MEASUREMENT: You use specific numbers and evidence to describe what your project has achieved so far (or plan to achieve in the future) and you have a plan for measuring impact - 66.7%

VIABLITY: You have given a great deal of thought to not just the idea itself but how to make it work from a financial perspective in the present and future - 33.3%

Other option - 33.3%

8. FEEDBACK: What are some areas for improvement for this project?

CONNECTION: Why you care about the environment/ climate was unclear – It would be great to elaborate on what this solution means to you, personally and how it affects you and/or your community. - 0%

CREATIVITY: Be more specific in your description of the research you have done into the past solutions to this problem and focus on how your solution is unique and innovative - 0%

COMMITMENT: Your plan for growing the organization can benefit from more specifics. How can you round out the various skills of your current leadership team to make the project a long-term success? - 0%

CHANGEMAKER QUALITY: Try to provide more insights into how you are activating changemakers and empowering them to innovate through your product or programming. How will they care about environment/climate if they currently do not? Think about how to create value for all stakeholders, not just immediate beneficiaries - 50%

IMPACT MEASUREMENT: Provide specific instances of your social impact and how you plan to measure impact – it may be helpful to describe the beneficiaries, products and programming, and provide evidence of (or plan for) how to measure impact - 50%

VIABILITY: Make sure you have provided descriptive information about your financial sustainability plan. Where do the funds come from now and do you have a concrete plan for future sustainability? - 100%

Nothing – I thought everything was great! - 50%

Other option - 50%

9 comments

Join the conversation:

Comment
Spam
Photo of Matthew Marchyok
Team

This is an excellent way computational thinking can have a constructive impact on solving problems in our world.

Spam
Photo of Ayush Raj
Team

Thank you!

View all comments