A Reward-Oriented Smart Waste Disposal System with Automated User Identification and Waste Classification

Improper waste segregation continues to be a significant issue in urban and institutional environments, leading to reduced recycling efficiency and increased operational overhead. Although several smart waste management solutions have been proposed, many of them primarily focus on waste classification and do not address user accountability or behavioral motivation. This paper presents a reward-oriented smart waste disposal system that integrates automated user identification with image- based waste classification to promote responsible disposal behav- ior. In the proposed approach, users are identified at the time of disposal using a face recognition module based on deep facial embeddings, while the disposed waste is classified into material categories using a MobileNetV2-based deep learning classifier trained through transfer learning. Based on the classification outcome, reward points are assigned and maintained through a centralized web application, allowing users to view their disposal history and accumulated incentives. The system architecture is based on a modular design that allows for clear separation of the vision processing, the application logic, and the data storage. The system also differs from previous smart bin systems in that user identification, waste classification, and reward schemes are not separate entities but are combined in a single workflow that ensures real-time correlation with specific user actions and rewards. The system's performance in user identification and waste classification is demonstrated through experiments conducted under controlled indoor conditions across multiple waste categories.