Hey Product Hunt! 👋 I'm Dhruv, a B.Tech student at IIMT University, and this is Prithv-E (Sanskrit for "Earth") — a smart bin that sorts your waste for you. The problem: Waste segregation in India (and honestly, most places) mostly fails at the source. People either don't bother separating organic, recyclable, and metal waste, or they don't know how. By the time it reaches a facility, everything's mixed and contaminated, which makes recycling way less effective. What we built: Prithv-E uses a Pi Camera and an on-device AI model (MobileNetV2, running via TFLite on a Raspberry Pi 4) to look at an item as it's dropped in, classify it as Organic, Recyclable, Metal, or General, and mechanically route it into the right compartment using a servo-driven flap — all in real time, no cloud dependency. Where we are: We're at prototype stage, incubated at IIMT University's incubation center and supported by an MSME grant. We've built the hardware, the sorting mechanism, and the data pipeline for training the model on real-world waste images (not just stock datasets, since actual bin conditions — lighting, mess, partial views — look very different). What's next: Fine-tuning the model on real captured data, and eventually working toward the kind of commercial-grade accuracy companies like Bin-e have shown is possible. Would love feedback from this community — especially on the AI/edge deployment side, and on how to think about scaling something like this beyond a single prototype bin. 🙏

Prithv-E Smart Bin

Where Innovation Meets Environmental Responsibility

Where Innovation Meets Environmental Responsibility