Traffic prOcessing foR uRban EnvironmentS (TORRES)
Funding:
INNOVIRIS Joint R&D Project
Duration:
2023-2025
Keywords:
digital twin, traffic, simulation
The concept of Smart Transportation has become more predominant over the past decade, encompassing innovative methods for the reduction of traffic congestion, traffic accidents, and air pollution, all of which engender excessive costs to society and impact the general well-being of citizens.
This page collects research papers published as part of TORRES, scientific events where TORRES has been present and main outcomes of the project.
Researchers involved:
Objectives
TORRES project aims at developing both a framework and methodology for the monitoring of heterogeneous traffic data. Specifically:
- Develop necessary dashboards and frameworks for traffic analysis, monitoring, and prediction in a metropolitan city like Brussels.
- Enrich knowledge about urban mobility dynamics through the integration of real and synthetic data.
- Aggregate mobility data from IoT devices to infer useful information for policymakers.
- Create new AI-based methods for interpolating mobility data.
Demos & Innovations
Traffic Management and Simulation Tool: TULIPE project
Urban traffic management struggles to evaluate and mitigate the impact of disruptive events like road closures on traffic flow. An interactive tool was implemented to help experts define, compare, and assess road deviation plans using different traffic models.
Advancing Traffic Management: Digital Twin Model
This demo shows a digital twin of a traffic intersection, integrating real-time data and machine learning to predict traffic conditions. The digital twin enables proactive congestion management and provides a safe environment to test traffic control strategies.
Simulating Traffic Networks: Driving SUMO towards digital twins
This demo presents the development of a digital twin in SUMO, integrating static and dynamic city data to simulate and analyze intermodal mobility networks. Using Osnabrück as a case study, it explores data integration challenges and future improvements.
Resources
Synthetic traffic generation
This project presents a generic tool for generating synthetic traffic demand and shows how it can be used to assess the impact of disrupting events in simulated scenarios where we lack traffic data. Traffic simulation is performed using SUMO software.
Academic Publications
A Simulation Tool to Assess the Impact of Deviation Plans on Disruptive Events of Urban Traffic
Evaluating and mitigating the effects of road disruptions on urban traffic remains a significant challenge. This paper introduces TrafficTwin, an interactive tool that enables experts to define, simulate, and compare road deviation plans using adaptable traffic models and quantitative visualizations.
Calibration of Vehicular Traffic Simulation Models by Local Optimization
This paper presents a novel stochastic simulation-based traffic calibration method that performs local calibration, scales to large environments, and relies solely on traffic count data. Experimental results on a Brussels traffic model show the proposed method improves accuracy by 16%.
Traffic Simulation with Incomplete Data: the Case of Brussels
To address the challenge of incomplete data from sensor failures, this study proposes using the HybridIoT technique to estimate missing data for creating traffic simulation models with SUMO, focusing on Brussels, and evaluates the impact of data sparsity on model accuracy.
Scientific dissemination
OpenMobility meetup: TrafficTwin introduction
Date: June 25, 2025
As a direct result of the Tulipe project, the TrafficTwin tool was implemented to assist public entities during the evaluation of road deviation plans.
Leveraging SUMO for Traffic Twins: Experiences in Urban Traffic Processing
Date: 12-14 May, 2025
The project aims to address current challenges in modern cities by integrating diverse traffic data sources into a unified system for analysis and decision-making. The simulation of traffic flows using SUMO can be used to evaluate policy impacts, supporting decision-making for sustainable traffic management.
TrafficTwin: A Simulation Tool to Assess the Impact of Deviation Plans on Disruptive Events of Urban Traffic
Location: Berlin
Date: 12-14 May, 2025
A novel tool, called TrafficTwin, is proposed for assessing the impact of alternative road deviation plans on vehicular traffic. The proposed tool targets experts in traffic management to make the use of SUMO easier for defining simulation models and performing what-if analysis in case of disruptive events that cause road closures.
Digital twins: modeling reality to make better decisions (Press article)
Date: April 15, 2025
What if we could simulate the effects of a construction site, a change in production, or an energy hazard before they even occur? This is the promise of digital twins, virtual replicas of real objects or systems capable of updating in real time using data from the field. By combining modeling, artificial intelligence, and predictive analysis, these devices open new avenues for decision-making, strategic planning, and the optimization of complex processes.
Integration of simulation and machine learning to predict the traffic in the event of disruptions
Location: Brussels
Date: 26 – 28 June 2024
Decision-making in intelligent transportation systems is complex due to urban traffic’s unpredictable dynamics, particularly when assessing the impact of road works on congestion. To address this, a method combining simulation using SUMO and machine learning with GNN is proposed for traffic prediction in the event of road disruptions.
Smart Decision Making for Cities by Traffic Simulation and Data Science
Location: Toulouse
Date: 15 January 2024
Invited seminar in IRIT lab.
Traffic Simulation with Incomplete Data: the Case of Brussels
Location: Hamburg
Date: 13 November 2023
Event: 1st ACM SIGSPATIAL International Workshop on Methods for Enriched Mobility Data: Emerging issues and Ethical perspectives EMODE 23
Intelligent transport systems support decision-making to reduce traffic congestion, accidents, and pollution, but this process is complicated by the unpredictable nature of urban traffic and incomplete data from sensors. This study uses the HybridIoT technique and SUMO simulation to estimate missing data for creating traffic models, evaluating their accuracy despite varying data sparsity in Brussels.