top of page
Concept
Reality
A smart recycling app which helps to segregate waste
A system powered by AI (Computer vision) + AR (Augmented reality) to separate trash from recyclable items.
ROLE
Interaction designer
YEAR
2022
PROJECT DURATION
Since 2022 and still cooking :)
I won't discuss the globally known problem; instead, I'll focus on how my iterations evolved as I matured...
PS: It all started as an academic project where I collaborated with three others on UN's SDG Goal 11.
And the major challenges for people that we learned through our studies are...
-
Awareness: Each material has multiple categories. For example, plastics differ in chemical composition and physical properties, which can confuse people.
-
Education: Not all materials have labels to indicate their recyclability, unless provided by the manufacturer.
-
Effort: The need to mentally remember recycling guidelines can make people less inclined to recycle.
-
Motivation: People often lack motivation to recycle if they don't see others around them doing it.
What's an individual's contribution to sustainability? We felt this could be it.
AI powered scanner
Educational game
Users choose the correct bin for items, like the black bin for a banana peel, to learn waste segregation.
A user can scan the product and it displays the result
But, later, It turned out to be a passion project and I started testing it...
What's the challenge in the previous solution?
It's a hassle scanning each item and waiting for the results
.png){kind=small}
.png){kind=small}
Select scanner
Pick an item
Point it to the scanner
It displays the result
Open app
Senses and interctivity
This game improves cognitive function 🧠, but finding motivation ❤️ to play a recycling game can be challenging.
Aligning goals before brainstorming alternative solutions
To reduce the number of interactions
To make it more gamified
To make it more engaging and fun
I wanted to quickly prototype the idea running in my head.
_gif.gif)
If user clicks on the highlights it displays a popup with more information
A system which shows the highlights in the real-time
Why did I choose this solution?
Plastics are classified into three categories: recyclable, partially recyclable, and non-recyclable based on their properties and the feasibility of recycling processes.
They are distinguished by labeled numbers indicating their recyclability where it's a mental effort to remember.
Likewise, glass, paper, metal, textiles, e-waste, wood, composite materials have unique classification
Refining the design and iterating for version 2
Made the information pop-up sticky on the screen.
I wanted to go further and try something more realistic this time
I used a no-code tool to train the model with multiple datasets and deployed it for real-time detection.
This is what I achieved for version 1 using a no-code tool
And here are the goals which I accomplished
Reduced the number of interactions
Made it fun and engaging for users
What would be my next steps?
Mapping boundaries to ensure precision and accuracy in detection.
Training the model with additional datasets
Training the model with additional materials and their classifications.
Until then, give a try using the QR
Note: The model may not be fully accurate.
bottom of page