T-Twin: Digital Twin Sandbox for Network-Level Traffic Control
Project name: T-Twin: Digital Twin Sandbox for Network-Level Traffic Control
Project leader: Zhenliang Ma , Transport planing
Participants KTH: Jonas Jostmann
Participating universities: VTI, Linköpings universitet
Project period: 2026-01-15 - 2028-04-14
Financing: Trafikverket
Digital Twin (DT) technology offers powerful capabilities for real-time traffic monitoring, forecasting, and control. The GEMINI project at KTH demonstrated the potential of DT-based approaches for urban traffic monitoring using IoT sensors and simulation, piloted in Kista. However, current traffic control systems in Sweden, such as LHOVRA for signal management, are still localized and reactive, with limited coordination across the broader network. Research at VTI and LiU has shown that connected vehicles (CVs) can enhance traffic performance but also highlighted the need for network level
optimization to fully unlock the value of CV data. As traffic conditions become more dynamic and unpredictable, a more adaptive and system-level control strategy is needed. Building on these insights, the T-Twin project aims to develop a modular DT Sandbox for integrated motorway and urban traffic control. The system will combine real-time monitoring, simulation, and AI-based decision-making to support safe and sustainable network-level traffic management. The long-term vision is to complement the TrV’s existing MCS system with DT and AI technologies for traffic flow control (see Appendix for the T-Twin framework).
The project focuses on control method development and testing using the Södra länken corridor in Stockholm (see Appendix for the test example network). This corridor includes both motorway segments and connected urban streets, providing a valuable testbed for evaluating how DT-based control strategies can coordinate across jurisdictional boundaries. Example control strategies will be tested: (1) dynamic speed limits for managing motorway flow, and (2) coordinated signal control for optimizing urban inflow. These strategies are expected to reduce peak congestion and support proactive traffic operations.
The project aims to: 1) Map current practices and stakeholder needs through literature reviews and consultations; 2) Develop a high-fidelity DT sandbox aligned with TDIS data structure; 3) Propose and implement a DT/AI-based network traffic flow control method for improved efficiency, safety, and robustness; 4) Validate the system using the Södra länken use case and develop the scale-up plan for integrating the T-Twin sandbox with the TDIS system for traffic flow control and management.
The project will support the continued PhD research of Jonas Jostmann (GEMINI project) in developing core DT-based traffic control functionality. The T-Twin system is designed to be modular and transferable and aligns with TrV’s ongoing efforts under projects such as ‘Next Generation Motorway Traffic Control’, ‘Connected Vehicles at Signalized Intersections’, and ‘IMTRACS Improving traffic signal control utilizing connected vehicle data and surface detection’. In the long term, the T-Twin will serve as a national testbed for developing and validating next-generation traffic control strategies in both simulation and practice.
Contact
