Inspiration

The idea came from a frustrating moment most of us have experienced: standing at a recycling bin, holding something — a greasy pizza box, a crinkled plastic wrapper, an old battery — and genuinely not knowing where it goes. So you guess. And more often than not, you guess wrong.

What it does

TrashSmart is a browser-based AI recycling assistant with three core features: Scan Any Item — Type any item name and instantly receive the correct bin category (Recycle, Compost, Landfill, or Special Disposal), step-by-step preparation tips, and environmental impact tags. Quick-tap chips cover the most common items for one-click lookups. Personal Waste Tracker — Log every item you dispose of correctly. The tracker shows live stats, a 14-day streak calendar, and a scrollable item history. It turns recycling from a one-off action into a measurable daily habit.

How we built it

TrashSmart is a single HTML file with zero external dependencies and no backend. The entire application — structure, styling, and logic — lives in one file that opens in any modern browser For styling, we committed to a dark forest green aesthetic using CSS custom properties for a cohesive, professional look. Animations and micro-interactions were built entirely in CSS, keeping the experience smooth without any JavaScript overhead

Challenges we ran into

TrashSmart prioritizes simplicity by keeping everything in a single file, making the AI feel real with a well-structured lookup engine, and turning data into meaningful environmental impact metrics. The app also incorporates gamification through badges that reward genuine good behaviors.

Accomplishments that we're proud of

We're proud that the entire application is zero-dependency and zero-install — a teacher could share a single file with a classroom, a municipality could embed it in their website, and anyone in the world could open it on any device without creating an account or downloading anything. We're proud of the disposal logic coverage — the app handles everything from pizza boxes and AA batteries to used cooking oil and old smartphones, with genuinely useful, specific advice for each, not generic platitudes.

What we learned

We learned that constraints breed creativity. Deciding early on to build a no-backend, single-file application forced us to solve every problem with the simplest possible tool — and the result is something more deployable, more accessible, and more resilient than a full-stack alternative would have been.

What's next for Waste Management - TRASHSMART

Geolocation-based drop-off finder. The single biggest gap in TrashSmart today is local specificity. Recycling rules vary by city. Integrating with the Earth911 API would let users find the nearest drop-off for special items like electronics, batteries, and cooking oil. Camera-based item recognition. Using TensorFlow.js for in-browser image classification, users could photograph an item instead of typing it — making the experience even faster and more accessible.

Built With

Share this project:

Updates