M-Chäschtli

• 

  Screenshot of the app

Inspiration

Sustainability and carbon footprint has now been a topic for a long time. But still there is big room for improvement for our society. But people usually look for easy ways to decrease their carbon footprint for our earth. Hence, we decided to help with that endeavour! Our novel app helps you to save money and at the same time reduce your carbon footprint! Impossible you think? Then read carefully and let us explain you our idea.

What it does

From the first login on our new app "M-Chäschtli" will track what you buy in Migros through the cumulus system. Together with the selling date the app then knows approximately when the products will expire. Based on this knowledge it will automatically warn you as well as give you suggestions for recipes for the expiring products, thus helping you to avoid food waste and save money at the same time! Amazing right!? But even more, assume you wanna plan what to cook for the next week, then ideally you wanna know what you have left in your food storage, right? With M-Chäschtli you can easily keep track of it. By simply choosing whether you already ate the food or had to trash it, a Food-Wastometer gives you insight about your efficiency while helping you to keep an overview of your food. Furthermore, M-Chäschtli gives you suggestions to for similar products with lower carbon footprint helping you to make conscious choices for the sustainability of earth. And hold on tight, soon there will be a one click food planner which creates menus based on your food storage history and favorite dishes powered by AI helping you to eat more well-balanced with the food you like!

How we built it

As a proof of concept we build a small app with Angular in the front-end and Flask in the back-end. The Application should give the look and feeling of how a user could interact with the platform.

Challenges we ran into

Labeling the products with keepability was the bigger challenge. This data was not available on the ~40000 available products. One approach was to ask ChatGPT to give approximate keepabilities in a formatted way, however due to the size limit of message this was problematic. Though, this must be done only once and hence we avoided spending to much time on this.

Accomplishments that we're proud of

We participated at the HackZurich23 of course! :P And we even build a small demonstrator of our idea in just a few hours! We believe in it! :)

What we learned

Data preparation is key and AI can be useful for data enrichment. Thus AI can be very useful in the design process of an application and not only in the end product.

What's next for M-Chäschtli

A smart food planner would make the app almost undefeatable in how easy one could avoid food waste! Thus a smart algorithm would be necessary to use the users food storage data and propose them menus based on their expiring products and taste. Since the taste is already taken into account by their own choice of product purchase we would propose an approach like a linear program to optimize e.g. the CO2 footprint and food waste. Thus avoiding the need for large user datasets which are usually not available at the beginning.

Built With

• anaconda
• angular.js
• flask
• html
• npm
• pandas
• primeng
• python
• rxjs
• scss
• typescript