Website Bank Sampah

A community-oriented web platform that digitizes the waste bank workflow with AI-powered waste detection. The system integrates Roboflow AI to automatically classify waste types from photos, enabling faster and more accurate deposit processing. Features include digital savings balance tracking, automatic price calculation based on waste type and weight, and analytics dashboards for monitoring environmental impact and community participation.

The community waste bank relied on paper-based recording and manual waste sorting, leading to frequent calculation errors, disputes about member balances, and inaccurate waste categorization. Staff struggled to identify and separate different types of recyclable materials quickly, causing delays at deposit counters and reducing overall throughput. There was no automated way to verify waste types, resulting in misclassification that affected pricing accuracy and recycling efficiency.

The organization needed a transparent, digital-first system that automates waste identification using AI technology, streamlines the deposit-to-balance workflow, and provides accurate environmental impact reporting.

I developed a web-based waste bank management platform using Flask (Python) for backend logic with PostgreSQL for reliable data storage, and Tailwind CSS with Jinja2 for a responsive, server-rendered interface. The system integrates Roboflow AI detector for automatic waste classification — users simply upload a photo of waste items, and the AI identifies the waste type (plastic, paper, metal, organic, etc.) with high accuracy. This eliminates manual categorization errors and speeds up the deposit process. Staff confirm the AI classification, input weight, and the system instantly calculates the member's balance credit. Analytics dashboards track total recycled volume, community participation, and environmental impact metrics.