Jia Hao ChenMemoona SaleemMuhammad Fahad KhanChing Seong TanMuhammad Jahangir2026-10-142026-10-142025-11-25979833157136810.1109/SCOReD68498.2025.11399120https://dspace-cris.utar.edu.my/handle/123456789/11746This study introduces an artificial intelligence-driven humanoid robotic system designed for intelligent waste sorting, with the dual objective of enhancing waste management efficiency and fostering greater environmental awareness among the public. The robot engages autonomously with individuals, gathering demographic and behavioral data through interactive dialogues and quizzes. Confidence scores are computed in real time using a multi-factor weighted approach that considers variables such as age, gender, and quiz performance. Decision-making is handled by a hybrid engine that integrates logistic regression with a random forest classifier to determine whether to execute physical actions, including opening the lid of a smart waste container. An online incremental learning framework is incorporated to enable the system to adapt continuously to evolving user behaviors. An analysis of 40 user interaction samples showed that the hybrid approach maintained stable and slightly conservative scoring patterns, with most values falling between 0.6 and 0.8 (average = 0.699, standard deviation = 0.125). Across evaluations, the Random Forest model consistently outperformed Logistic Regression, reaching a validation accuracy close to 81.25%. Overall, the system demonstrates robust output stability, effective detection of anomalous behavior, and dependable incremental updates, thereby addressing the constraints of conventional static waste classification methods while enhancing the intelligence, adaptability, and public engagement in urban waste management. © 2025 IEEE.en-USGarbage classificationHumanoidLogistics regressionRandom forestSmart Waste ManagementAnthropomorphic robotsBehavioral researchClassification (of information)Environmental managementIntelligent robotsLogistic regressionLogisticsRandom forestsSortingWaste managementEnvironmental awarenessHumanoid roboticsLogistics regressionsManagement efficiencyRobotic systemsSmart waste managementWaste sortingDecision treestext::conference output::conference proceedings::conference paper