International Journal of Scientific Research and Engineering Development

The rapid urbanization of cities across India and the globe has intensified the challenge of managing municipal solid waste efficiently. Traditional garbage collection systems depend on manual inspection and fixed-schedule pickups, leading to overflowing bins, delayed responses, and inequitable distribution of sanitation resources. This paper presents the design, development, and experimental evaluation of a Smart City Garbage Collection Management System that integrates Artificial Intelligence (AI) with a web-based administrative platform to automate and optimize urban waste management. The core of the system employs the YOLOv8 (You Only Look Once version 8) deep learning object detection model, trained on a custom dataset of 3,847 images spanning 8 garbage categories including plastic bottles, polythene bags, cardboard, food waste, glass bottles, metal cans, electronic waste, and mixed garbage, to identify and classify waste items in real time from both citizen-uploaded photographs and live camera feeds. The system is built using the Python Flask web framework for backend processing, MySQL for relational database management, and a responsive HTML5/CSS3/JavaScript frontend with Jinja2 templating. An intelligent round-robin autoassignment algorithm distributes garbage reports equitably among approved sanitation staff. The trained YOLOv8n model achieved an overall mAP@0.5 of 90.6%, with Metal Cans reaching the highest per-class AP of 96.3%

Abhishek A Gulhane, Sahil P Raipure, Radha U Bhelkar, Sachin S Tipare, Satyak R Gulhane,"AI- Based Garbage Detection System: Integrating YOLOv8 Deep Learning, Flask Web Framework and Intelligent Staff Auto-Assignment" International Journal of Scientific Research and Engineering Development, V9(2): Page(2020-2027) Mar-Apr 2026. ISSN: 2581-7175. www.ijsred.com. Published by Scientific and Academic Research Publishing.