Integrating Geographic Information Systems and Machine Learning to Improve Traffic in Smart Cities

NOURHAN BACHIR | In Progress

Summary

Modern cities face increasing challenges in maintaining safe and efficient transportation systems amidst disruptions such as traffic congestion, accidents, natural disasters, and intentional disturbances. Addressing these issues requires innovative and adaptable solutions that transcend the limitations of traditional methods. Existing approaches often rely on static measures or case-specific frameworks, making them difficult to generalize across diverse urban environments.

This research focuses on developing scalable and generic methodologies to enhance the safety and resilience of transportation systems in smart cities. Introducing the PEMAP (Post-Event Management of Transportation Systems) framework, this work integrates Geographic Information Systems (GIS), Artificial Intelligence (AI), and Vehicular Ad-Hoc Networks (VANETs) to provide a comprehensive strategy for managing disruptions and ensuring the swift recovery of traffic networks. By leveraging advanced data-driven techniques, the proposed solutions aim to be adaptable across different urban contexts, addressing the dynamic nature of traffic patterns and network vulnerabilities.

A significant contribution of this research is the use of microscopic simulations, particularly SUMO (Simulation of Urban Mobility), to model traffic dynamics with a high degree of precision. This approach provides an efficient and realistic representation of urban traffic systems, enabling researchers to evaluate road link criticality and network performance without the need for on-ground studies. By simulating traffic scenarios under various disruption conditions, this method offers valuable insights into network behavior and facilitates data collection for predictive models.

Central to the work is the identification and evaluation of road link criticality, a key step in understanding the vulnerabilities of transportation networks. The research proposes innovative methods that prioritize critical road links using data-driven analyses and machine learning frameworks. These approaches not only enhance predictive accuracy but also ensure that the solutions are scalable to larger networks and transferable across different cities.

This research underscores the importance of generic, adaptable, and efficient frameworks for managing urban transportation systems in the context of smart cities. By combining the strengths of GIS, AI, and microscopic simulations, it contributes to the development of sustainable, resilient, and intelligent transportation systems that can adapt to the challenges of modern urbanization.

Publications

• Bachir, N., Harb, H., Zaki, C., Nabaa, M., Nys, G.-A., & Billen, R. (September 2023). PEMAP: An intelligence-based framework for post-event management of transportation systems.
• Computers and Electrical Engineering, 110, 108856. doi:10.1016/j.compeleceng.2023.108856 https://hdl.handle.net/2268/305270
• Bachir, N., Zaki, C., Harb, H., & Billen, R. (28 November 2024). SAMO: A Sequential Pattern Mining Model for Evaluating Road Criticality in Urban Traffic Networks [Paper presentation]. 2024 IEEE 100th Vehicular Technology Conference (VTC2024-Fall), United States. doi:10.1109/VTC2024-Fall63153.2024.10757675 https://hdl.handle.net/2268/325236
• Bachir, N., Zaki, C., Harb, H., & Billen, R. (December 2024). VeTraSPM: Novel Vehicle Trajectory Data Sequential Pattern Mining Algorithm for Link Criticality Analysis. Vehicular Communications, 51, 100869. doi:10.1016/j.vehcom.2024.100869 https://hdl.handle.net/2268/325235