MUMBAI, India, Sept. 22 -- Intellectual Property India has published a patent application (202641110811 A) filed by Anupooja A N; Prof. Bandhavya G; Prof. Maheshn; and Vaijnath Patil on September 15, 2026, for An Ai-Based Bio-Bin System And Method For Automated Detection, Dual-Stage Classification And Sustainable Management Of Waste Using Object Detection And Deep Learning.
Inventors include Anupooja A N; Prof. Bandhavya G; Prof. Maheshn; and Vaijnath Patil.
The application for the patent was published on September 18, 2026, under issue no. 38/2026.
Abstract: Disclosed is an AI-based Bio-Bin system and method for automated waste detection, dual-stage classification and sustainable management. A camera module, optionally triggered by a proximity/distance sensor, captures an image of a waste item, which is pre-processed and passed to a waste detection module employing an object-detection model (e.g., YOLOv8) to localize the waste-item region. A waste classification module employing a convolutional neural network extracts a feature vector from the localized region, classifies the item into one of a plurality of predefined categories (recyclable, reusable, compostable, hazardous or electronic waste), and generates a confidence score. A verification and decision engine compares the feature vector against reference data stored in a centralized database using a similarity metric and compares the confidence score against a predefined threshold, automatically accepting and logging high-confidence, validated classifications and flagging low-confidence or invalid classifications for manual administrator review via a web-based dashboard, with real-time alerts generated for hazardous, unauthorized or uncertain items. A Wi-Fi enabled microcontroller-based hardware embodiment, an e-waste marketplace module and a carbon-footprint estimation module are further disclosed.
Disclaimer: Curated by HT Syndication.
