Design and Implementation of Microscopic Traffic Simulation for Digital Twin of Bandung

Abstract

Traffic congestion remains a critical challenge in large cities. This paper presents a microscopic traffic simulation using a multi-agent approach that integrates an OpenStreetMap-derived road network, Intelligent Driver Model (IDM) for car-following, and MOBIL for lane-changing. The simulator targets a Bandung-scale Road network and records performance metrics such as density, average speed, delay, and computational throughput. On commodity hardware, a performance study indicates the simulator can sustain approximately 40,395 active vehicles per 1,000 ms simulation step, suggesting feasibility for city-scale what-if analysis and as a building block for a Digital Twin pipeline and smart traffic control. We further specify that pipeline as six stages coupling the simulator to the real-time CCTV network of the Bandung City Transportation Agency (Dishub Kota Bandung), and report a prototype trial at the Buah Batu-Samsat-Dayeuhkolot intersection in which observed per-approach queues drive the simulated state and the resulting indicators feed back into signal-timing decisions.