Development and Evaluation of Segresmart: An AI-Enabled Mobile Application for Improving Household Waste Segregation Behavior

Abstract

Improper household waste segregation remains a persistent challenge that undermines recycling efficiency and sustainable solid waste management, particularly in urban communities. This study presents the development and evaluation of SegreSmart, an AI-enabled mobile application designed to improve household waste segregation behavior through real-time waste identification and actionable disposal guidance. Guided by a design-and-development research framework, the system integrates image-based artificial intelligence, a user-centered mobile interface, and behavioral analytics to support informed segregation decisions at the point of disposal. The application was evaluated using a quasi-experimental pre-test and post-test design involving urban household participants over a four-week intervention period. Behavioral outcomes were assessed in terms of segregation accuracy, frequency of correct segregation, and perceived behavioral control, complemented by system usage logs and AI performance metrics. Results indicated significant improvements across all behavioral indicators following the intervention, with users demonstrating higher accuracy and consistency in waste segregation and increased confidence in waste classification decisions. The AI model achieved high classification accuracy under real-world conditions, while the human-in-the-loop design, incorporating confidence indicators and manual overrides, enhanced user trust and learning. High usability and acceptance ratings further confirmed that the application was intuitive and suitable for routine household use. Overall, the findings demonstrate that Segre Smart is a functional, usable, and data-driven mobile intervention capable of supporting positive behavioral change in household waste segregation. The study contributes empirical evidence on the effectiveness of AI-enabled mobile applications as decision-support tools for sustainable household waste management. It provides a foundation for future large-scale deployment and longitudinal evaluation.