DTUMOS, digital twin for large-scale urban mobility operating system

Area of Science:

• Urban planning and smart city technologies
• Transportation systems engineering
• Computer science and simulation

Background:

• Digital twin technology is increasingly vital for smart cities and urban mobility.
• Existing digital twins and simulations for mobility systems face challenges in scalability and speed.
• There is a need for advanced frameworks to test and develop urban mobility solutions.

Purpose of the Study:

• To introduce DTUMOS, an open-source digital twin framework for urban mobility operating systems.
• To present a novel architecture combining AI for ETA prediction and vehicle routing.
• To demonstrate DTUMOS' advantages in performance, scalability, and visualization.

Main Methods:

• Developed DTUMOS, a versatile and open-source digital twin framework.
• Integrated an AI-based estimated time of arrival (ETA) model.
• Incorporated an advanced vehicle routing algorithm.
• Validated performance and scalability using real-world data from Seoul, New York City, and Chicago.

Main Results:

• DTUMOS achieves high-speed performance and accuracy in large-scale urban mobility simulations.
• The framework demonstrates superior scalability, simulation speed, and visualization capabilities.
• Validated effectiveness using real data from major metropolitan areas.

Conclusions:

• DTUMOS provides a robust platform for developing and testing urban mobility systems.
• Its lightweight, open-source nature facilitates the creation of simulation-based algorithms.
• Enables quantitative evaluation of policies for future urban mobility infrastructure.