Optimizing safe and sustainable public transit for wellness tourism: Southeast Asian case studies

This study presents the development of the Reinforcement-Learning Multi-Objective Artificial Multiple Intelligence System (RL-MOAMIS) framework, aimed at optimizing public transport systems specifically designed for wellness tourism in Southeast Asia. The framework is tailored to enhance the safety, sustainability, and economic viability of bus routes that connect key wellness destinations across Thailand, Laos, and Vietnam. By integrating safety and environmental sustainability principles through a mixed-integer programming model, the RL-MOAMIS framework introduces two innovative improvement methods inspired by metaheuristic operators and employs a reinforcement-learning-based approach to dynamically select enhancement strategies. The empirical results demonstrate that the RL-MOAMIS framework significantly outperforms conventional methods, such as the Genetic Algorithm and Whale Optimization Algorithm, achieving a 27.45% increase in optimization efficiency, a 15% improvement in safety metrics, and a sustainability index enhancement ranging from 7.35% to 13.48% across all routes. The findings underscore the potential of RL-MOAMIS as a comprehensive and adaptable solution for optimizing public transport in wellness tourism. These improvements not only provide safer and more sustainable transportation options but also support economic development in rapidly urbanizing regions. The implications of this study are significant for policymakers, urban planners, and tourism developers, offering critical insights for the design and implementation of public transport systems that cater to the unique demands of wellness tourism.