Taking responsibility for our planet. turning
Trash to Treasure
Trash to Treasure is a senior capstone project that I developed that uses object scanning and detection alongside a trained data set to create new and creative ways that we can reuse common trash items, turning trash into treasure. The goal of this project was to understand how object detection, AI content generation, and user interactivity could affect individuals’ sustainability habits.
Demo Video
This project is still being worked on, but attached is the demo video of the basic functionalities of the app.
Research Process
To create the application, I went through an extensive research process of the market as well as researching creative decisions.
-
Color Pysychology (figma.com/colors):Neon Yellow/Green:
- Its brightness makes it stand out, representing innovation and forward thinking. It can boost creativity and promote energy, all of which is helpful for the goal of the application: to promote creative ways to encourage sustainable habits.
Black/Ebony:
- Ebony is an excellent background color that can make lighter text and vibrant colors pop, enhancing readability and highlighting key elements. Ebony for is used for luxury brands, high-end products, or services that aim to project an image of exclusivity and premium quality.
Off White (warm):
- With its light beige hue and warm undertones, it feels elegant and rustic. Warm tone off white adds a touch of softness and sophistication to designs.
Type Research:
Header Font (Roc Grotesk Wide Bold):
- Using a bold and wide font is easily readable and draws attention quickly. The modern and versatile type is necessary to capture the viewer's attention and bring them to the need to practice sustainable habits.
Accent Font (Ivy Presto):
- Ivy Presto Display adds a softness and delicate touch to the overall look of the application. It takes away some of the harshness of the intense topic as well as adds a great balance to the bold sans-serif used elsewhere.
Body Copy (Roc Grotesk Medium):
- Roc Grotesk Medium adds cohesion being in the same family as the header text. This font is readable and easy to look at.
-
- Millennials are more willing to spend more money on sustainable products (CBC)
- Recycling rates have increased from about 6% in 1960 to 35% in 2017 (EPA)
- Only about 23% of disposable plastic water bottles are recycled (Pace)
- Plastic water bottles take 450 years to decompose
80% of the plastic water bottles people buy end up in landfills (Pace)- 85% (poll of 1000 Americans) somewhat or strongly agree that plastic waste pollution is a serious and concerning problem that requires immediate political action to solve. (Sullivan)
- 94% said they were “somewhat to much more likely” to recycle plastics (Sullivan)
Direct Competitor: Scrapp Recycling
Strengths:
- Ability to scan a wide variety of objects
- Using barcode for scanning doesn't limit item identification
- Lots of information about the material type and the proper way to recycle that material
- User incentives with correct recycling habits
- Specific to local recycling policies
Weaknesses:
-Bar code scanning functionality extremely unreliable for recycling specification
- Users have to input data about what type of plastic etc products are made out of when that is what the application is supposed to do
- User built data base
- Products have to have
Opportunities:
- Wider range/ more accurate object identification could pair with the object detection used in Trash to Treasure for the widest variety of object identification
- User added database expands and updates regularly if properly implemented
Threats:
- Application for businesses, individuals, and communities
- Understands local needs and recycling habits with automation
- Targeted to a wide audience
Direct Competitor: TOMRA
Strengths:
- Uses Deep learning and AI for object identification
- Extremely detailed information about recycling information for each item
- Can identify virtually any product
Weaknesses:
- No mobile app for scanning yet
- Not as user-friendly UI/UX
- Requires 4 color camera for object identification
Opportunities:
- Local recycling map and community involvement
- Deep Learning tool extremely helpful for AI object identification
- Pairing with AI content generation for unlimited reuse and upcycling opportunities
Threats:
- Extremely sophisticated model
- Not limited to a trained data set
References
Environmental Protection Agency. (n.d.). National Overview: Facts and Figures on Materials, Wastes and Recycling. EPA. https://www.epa.gov/facts-and-figures-about-materials-waste-and-recycling/national-overview-facts-and-figures-materials#:~:text=Over%20time%2C%20recycling%20and%20composting,to%2032.1%20percent%20in%202018.
(CBC), Y. M. (2024, September 19). 38 eco-friendly consumers statistics: A must know in 2024. BusinessDasher. https://www.businessdasher.com/environmentally-conscious-consumers-statistics/#:~:text=66%25%20of%20consumers%20worldwide%20say,believe%20care%20about%20the%20environment.
Pace, M. (n.d.). Important plastic water bottle stats. aquasana.com. https://www.aquasana.com/info/important-plastic-water-bottle-stats-pd.html?srsltid=AfmBOoqgBAkLkNwufF97O3ZB08xo7xbE8Hw4UKNc41-quHfcZJYyKlRU
Sullivan, L. (2022, October 24). Recycling plastic is practically impossible - and the problem is getting worse. NPR. https://www.npr.org/2022/10/24/1131131088/recycling-plastic-is-practically-impossible-and-the-problem-is-getting-worse
Skills developed & software used
-
![]()
TouchDesigner
Prototype Creation
-
![]()
Figma
UI/UX layout, initial wireframing, grayboxing, and prototyping
-
![]()
Python
Object Detection and Identification
-
![]()
OpenAI
Content Generation
