Inspiration
As a steadfast group of friends, it was almost natural that we'd decide to work on a project of this scale together. Beyond being just a super fun event for us, it was also a chance to allow us to do something that would protect something that us team members cherish dearly- the environment. And of course, Umer's insistence that we implement a machine learning model in anyway (rolls eyes). But on a more serious note, we believe that the best way to raise awareness is to make things accessible and help educate people. Thus, EcoEats was born.
What it does
On the surface, our project appears to be a search engine simply suited to providing you the best guidance on how to satisfy your cravings, at home. Instead, it also implements a highly complex algorithm model, a pretrained one that we somewhat altered to suit our needs better, and does efficient webscraping through it. What it does is collect a list of 4 recipe links, based on keywords that are put into our android app, and then collects and ingredient list for each recipe (as well as their quantities, the unit for said quantity, and so on) through parsing the recipe on each link. After, our webscraper program checks the carbon footprint of each of our ingredients, by searching each ingredient in the website www.carboncloud.com, which gives the carbon footprint. Next, we take the quantities for each ingredient, in units such as cups, tablespoons, etc., converts them to kilogram and multiplies them by the carbon footprint we obtained for each ingredient. Then, we add up the appropriately scaled carbon footprint and output them to the android app interface.
How we built it
We began by brain storming ways to do what we wanted to do and how to accomplish it during our given time constraint. We then split up to build separate parts of the program which allowed us to each develop our skills in different, but equally interesting areas. In particular, areas of full-stack app development, machine learning algorithms, and graphics design were all explored.
Challenges we ran into
As our first attempt to create an Android app, we initially utilized the Android Studio application as it seemed to be simpler for implementation. However, this quickly became a massive roadblock given that Kotlin and Java was needed to effectively create the app. Additionally, the back-end machine learning algorithm also proved to be more difficult to create as well. However, through teamwork and perseverance, we were able to overcome these challenges as a whole.
What we learned
Among these different areas of of the project, we were able to be exposed to different materials such as using HTML or Java to implement our original idea into a working app.
Log in or sign up for Devpost to join the conversation.