WASSTE

Inspiration

We realized that relying on the general public to memorize and perfectly execute complex recycling rules is a flawed strategy. Public waste management heavily depends on this "honor system", municipal bins are labeled by waste type, and departments expect citizens to meticulously separate their trash before dumping.

However, this expectation clashes with reality. People are often in a rush, confused by local guidelines, or simply unaware of what materials go where. This confusion is heavily compounded by the fact that waste management rules are not universal; they vary drastically from country to country, and even from municipality to municipality. What goes into the recycling bin in one neighborhood might be strictly landfill waste just a few miles away.

This hyper localization of rules creates a massive challenge in vibrant, multicultural hubs like Toronto. With a constant influx of tourists, students, and newcomers from all over the world, each accustomed to entirely different environmental regulations and sorting systems the municipal "honor system" completely breaks down.

This disconnect leads to massive contamination in recycling streams, ultimately ruining perfectly good recyclables and costing cities time and money. We built WASSTE to bridge this knowledge gap, offloading the burden of classification from the everyday person to artificial intelligence.

What it does

WASSTE is an intelligent, AI-powered waste management ecosystem designed to monitor, classify, and analyze public waste streams. Smart Scanning: Through our Scan interface, the system uses computer vision to look at discarded items and instantly determine the correct waste category (Recycling, Compost, Landfill, etc.).Comprehensive Dashboard: We provide a centralized control center for waste management departments featuring interactive maps, live metrics, and deep-dive analytics per bin (BinDetail).AI Sustainability Agent: Instead of just showing raw data, WASSTE features an autonomous agent that actively monitors waste patterns. It uses an |Observe \Reason \ Recommend loop to generate actionable insights. (e.g., noticing a spike in contaminated recycling at a specific park and recommending better signage).

How we built it

Challenges we ran into

Categorization Logic: Standardizing waste categories, colors, and labels across the client (wasteCategories.js) and the server's environment configurations required strict syncing to keep the UI charts rendering correctly.

Accomplishments that we're proud of

Successfully building an autonomous Sustainability Agent that doesn't just read data, but actually reasons about it to provide human-readable, actionable advice for waste management teams.

Developing a beautiful, highly responsive frontend dashboard with complex, synchronized charts that clearly communicate environmental impact.

Creating a robust data-seeding script that generated 30 days of realistic, nuanced waste events across 8 smart bins, allowing us to thoroughly test our analytics engine.

What we learned

What's next for WASSTE

Hardware Integration & Auto-Sorting: Moving our Scan functionality from a web client into a physical Raspberry PiMpowered camera system mounted inside actual public trash cans. This smart hardware won't just monitor the waste the camera will actively "read" deposited items using computer vision and automatically sort the trash into the correct internal compartments, completely removing the human element from the physical sorting process.

Predictive Routing: Using our analytics service to forecast when specific bins will be full, allowing waste management fleets to optimize their collection routes and save fuel.

Public Gamification: Introducing a mobile companion app where citizens can scan their trash, learn about local recycling rules, and earn "sustainability points" redeemable for local rewards.