Optimizing Artificial Intelligence-Based Waste Bank Management

This study examines the implementation of artificial intelligence (AI) technology to optimize waste bank management in West Pamulang, Indonesia. With the national waste volume reaching 68.5 million tons in 2023 and an annual growth rate of 2-4%, sustainable waste management presents critical challenges. West Pamulang accounts for 60% of regional waste, while Indonesia's 8,000 waste banks only reach 1.7% of the contribution to national waste reduction. Using a mixed method approach, the study was conducted in five waste banks in West Pamulang, South Tangerang during January-April 2025, involving 45 participants selected through purposive sampling. Data collection included participatory observations, interviews, questionnaires, and documentation studies. Reliability was assessed using Cronbach's Alpha 0.89, with validity guaranteed through triangulation. Ethical safeguards include informed consent, data anonymization, and institutional ethical approval. The results show significant operational improvements through AI technologies: computer vision-based classification systems, real-time transaction recording, educational chatbots, and volume prediction systems. Quantitative analysis revealed an increase in transaction efficiency by 75%, a 60% decrease in classification errors, and a decrease in data management time from day to minute. The AI predictive model achieves 92% accuracy in volume estimation and 15% fuel savings through route optimization. The classification system shows an accuracy of 89-97%, reducing the sorting time by 70%. Implementation challenges include limited digital literacy, infrastructure gaps, and inadequate policy support. The study recommends training programs, cost-effective platforms, and multi-stakeholder collaboration for a sustainable AI-enhanced waste management system.