Virtual Roads, Smarter Safety: A Digital Twin Framework for Mixed Autonomous Traffic Safety Analysis

TL;DR

The paper tackles proactive safety analysis in mixed autonomy by proposing a digital-twin platform that fuses multi-modal data (drone LiDAR, OSM, and vehicle sensors) to build high-fidelity 3D road models and a joint CARLA-SUMO-PhysX simulation environment. It introduces a modular workflow for background generation, map refinement with inclinometer data, and DBSCAN-based junction assignment, followed by a synchronized joint-simulation pipeline that captures macro traffic flow and micro vehicle dynamics. A high-fidelity active safety analysis framework with $t_c^{*}$ for earliest collision time demonstrates superior risk assessment compared with traditional TTC, validated through eight synthetic scenarios and quantitative accuracy improvements (MAE and RMSE reductions). The work advances practical safety evaluation in autonomous and mixed-traffic systems by enabling physics-informed scenario testing, realistic sensor-level perception, and scalable, repeatable experimentation for safety interventions in urban mobility.

Abstract

This paper presents a digital-twin platform for active safety analysis in mixed traffic environments. The platform is built using a multi-modal data-enabled traffic environment constructed from drone-based aerial LiDAR, OpenStreetMap, and vehicle sensor data (e.g., GPS and inclinometer readings). High-resolution 3D road geometries are generated through AI-powered semantic segmentation and georeferencing of aerial LiDAR data. To simulate real-world driving scenarios, the platform integrates the CAR Learning to Act (CARLA) simulator, Simulation of Urban MObility (SUMO) traffic model, and NVIDIA PhysX vehicle dynamics engine. CARLA provides detailed micro-level sensor and perception data, while SUMO manages macro-level traffic flow. NVIDIA PhysX enables accurate modeling of vehicle behaviors under diverse conditions, accounting for mass distribution, tire friction, and center of mass. This integrated system supports high-fidelity simulations that capture the complex interactions between autonomous and conventional vehicles. Experimental results demonstrate the platform's ability to reproduce realistic vehicle dynamics and traffic scenarios, enhancing the analysis of active safety measures. Overall, the proposed framework advances traffic safety research by enabling in-depth, physics-informed evaluation of vehicle behavior in dynamic and heterogeneous traffic environments.

Figures (18)

• Figure 1: Digital Twin Pipeline.
• Figure 2: The multi-modal data-enabled traffic environment reconstruction.
• Figure 3: The reconstruction of the drone-scanned data.
• Figure 4: Georeferenced data and GCPs embedded LiDAR point cloud.
• Figure 5: Workflow diagram of the digital infrastructure.