Decode

Helping people Decode their clothing labels to make more sustainable decisions and prevent greenwashing.

Comment

An Unlikely Culprit for CO2 Emissions

As a whole, the fashion industry is responsible for 10% of global carbon emissions, but Fast Fashion in particular perpetuates the cycle of overproduction and impulse buying.

Have you ever purchased from Shein, Forever21, or Urban Outfitters? These are just some brands are known to contribute to fast fashion.

Fast fashion focuses on mass-producing clothes in a short period of time to take advantage of recent trends. This fast-paced model is cheap and affordable, which is great for businesses looking to make a profit. However, these clothes are made with synthetic materials that break down easier and pollute the environment with harmful chemicals and microplastics. They also degrade faster, meaning you need to replace fast fashion items more often.

As people learn about the dark side of fashion, many people want to turn to sustainable options. Brands know this, and want to cash in on this consumer trend by using greenwashing, AKA, making misleading claims that their products are made sustainably when they probably aren't.

Fight Back Against Fast Fashion

To combat greenwashing, we created Decode. Decode helps analyze pieces of clothing and rates how sustainable they are based on what materials they are made of. Users can use Decode in three ways:

  1. Download the chrome extension and use it on specific clothing pages as they online shop
  2. Copy and paste the URL of a clothing item into our website, or
  3. Manually input the clothing information on our website

How It Works

Given an article of clothing from a website (via one of the 3 methods above), Decode will web-scrape the page to determine clothing information like the blend of materials (80% cotton, 20% polyester), clothing type, brand, and weight.

Using this information, Decode will run its sustainability, quality, and number of washes algorithms based on sustainable clothing research. Decode will then give you:

  1. A sustainability rating based on the materials (score out of 10)
  2. The fabric quality (score out of 10)
  3. The estimated number of washes before product needs to be replaced
  4. Tips to maximize that specific clothing's lifetime
  5. Similar items in the database with high sustainability scores

How We Built It

Like most projects, it all started with a brainstorming session going over different problems tackling sustainability. Fashion just so happened to be the one topic we all agreed upon, and once we decided in this specific direction, we went straight into Figma to start building brand identity, grey-scale and high-fidelity wireframes. We followed the double diamond design process to explore many options in design and synthesizing our findings to build the most compelling solution for our end users.

There was also extensive secondary research done on the sustainability, fabric quality, and number of washes, which was used to create equations in the backend.

Between the two coders, we split responsibilities into frontend and backend. For the frontend, we started by getting the Chrome Extension live, managing the user workflow, and styling the website. The backend was setup with Python/Flask. Then, functions were made using the secondary research to generate the results.

Challenges

One of the challenges our team faced was finding exactly how much energy certain materials require to produce. These numbers took some time to find and some had to be converted into specific units to ensure consistency across all materials.

Software

  1. First time using the Chrome Extensions environment made it difficult to navigate around
  2. Connecting Chrome Extensions to Python Backend created many CORS/permissions issues - we dived deep into Stack Overflow
  3. Chrome API's new update to Manifest JSON V3 doesn't have many resources online yet
  4. Web scraper was getting detected as a bot, so we had to mask requests with rotating browser headers

Accomplishments

The coders had their hands full with building a chrome extension for the first time, as well as setting up a Python backend for the first time. The designers created all the branding, hi-fidelity screens, and user-workflow in a limited amount of time!

What We Learned

Every hackathon is a learning experience, and things such as time management and communication are always key factors. This time around we learned harder skills such as how to make a working chrome extension and developing the backend in python using flask.

In the future, we would setup the connection between the frontend/backend sooner as it created a lot of issues towards the deadline of the hackathon. While using dummy data is OK to get things started, it is important to establish the connection ASAP, as it is the foundation of the codebase.

Next Steps

Improving the WebScraper

Our web scraper could be improved by allowing it to scan more types of HTML elements, as we used the most common ones. Additionally, adding REGEX expressions would increase the fidelity of the webscraper.

Improving the Database

We could also log all user entries into a large database to have a larger collection of recommended products to suggest to users. As the database grows, its sustainability/quality ratings would improve as well.

Getting User Feedback

Our calculations are theoretical based on previous sustainability material research. However, the best way to improve equations is to use real world data to improve the machine learning accuracy.

Share this project:

Updates