A Systematic Decade Review of Trip Route Planning with Travel Time Estimation based on User Preferences and Behavior
Nikil Jayasuriya, Deshan Sumanathilaka
TL;DR
This paper addresses the limitations of current navigation systems in personalization, real-time adaptability, and scalability for travel time estimation ($TTE$) in urban networks. It conducts a systematic decade review using AI techniques including reinforcement learning, graph neural networks, and federated learning, alongside emerging methods like explainable AI and meta-learning to characterize adaptive trip route planning. Key contributions include identifying gaps in real-time personalization data integration and ethical deployment, and proposing Industry 5.0–driven pathways with privacy-preserving approaches and lightweight deployments. The work offers a business- and user-centric roadmap to build transparent, sustainable navigation systems that meaningfully improve user satisfaction and urban mobility.
Abstract
This paper systematically explores the advancements in adaptive trip route planning and travel time estimation (TTE) through Artificial Intelligence (AI). With the increasing complexity of urban transportation systems, traditional navigation methods often struggle to accommodate dynamic user preferences, real-time traffic conditions, and scalability requirements. This study explores the contributions of established AI techniques, including Machine Learning (ML), Reinforcement Learning (RL), and Graph Neural Networks (GNNs), alongside emerging methodologies like Meta-Learning, Explainable AI (XAI), Generative AI, and Federated Learning. In addition to highlighting these innovations, the paper identifies critical challenges such as ethical concerns, computational scalability, and effective data integration, which must be addressed to advance the field. The paper concludes with recommendations for leveraging AI to build efficient, transparent, and sustainable navigation systems.