Full Text
<article class="scholarly-article">
<h2>Introduction</h2>
<p>Large-scale events such as international sports tournaments, music festivals, and political gatherings attract thousands of participants and spectators, placing immense strain on urban transportation infrastructure. The sudden surge in traffic demand, combined with road closures and parking restrictions, often leads to severe congestion, increased travel times, and heightened risk of accidents. Traditional traffic management strategies rely on pre-event planning and historical data, but they lack the flexibility to adapt to real-time disruptions and dynamic crowd behavior. As urban populations grow and event frequency increases, there is an urgent need for intelligent, data-driven systems capable of proactive and adaptive traffic control.</p><p>Digital twin technology offers a promising paradigm for addressing these challenges. A digital twin is a virtual representation of a physical system that is continuously updated with real-time data, enabling simulation, prediction, and optimization (Grieves, 2022; Rasheed et al., 2020). Initially developed for manufacturing and aerospace, digital twins are now being explored for smart cities, including traffic management (Sánchez-Vaquerizo, 2021; Saroj et al., 2021). By integrating IoT sensors, connected vehicles, and advanced analytics, a traffic digital twin can mirror the current state of the network and forecast future conditions, allowing operators to test interventions in a risk-free environment before deployment.</p><p>Event traffic management is particularly well-suited for digital twin applications because events are episodic, high-impact, and generate rich data from ticketing systems, mobile devices, and roadside sensors. Researchers have developed prediction models for event-induced traffic (Okukubo et al., 2022; Li, 2020) and proposed ITS functions for event management (BULATOVA, 2022; Zagidullin, 2017). However, most existing work focuses on individual components—prediction, simulation, or control—without a unified framework that leverages real-time data fusion and closed-loop optimization. This paper aims to fill that gap by proposing a comprehensive digital twin-driven traffic management system for large events, validated through a realistic simulation case study.</p><p>The remainder of this paper is organized as follows. Section 2 reviews relevant literature on digital twins, traffic prediction, and event management. Section 3 describes the proposed framework and methodology. Section 4 presents simulation results and performance evaluation. Section 5 discusses implications, limitations, and future work. Section 6 concludes the paper.</p>
<h2>Literature Review</h2>
<p>Digital twins have been widely adopted in manufacturing for product lifecycle management, predictive maintenance, and supply chain optimization (Wang et al., 2021; Tao et al., 2018; Filho et al., 2023). In the transportation domain, researchers have developed digital twins for connected corridors (Saroj et al., 2021), air traffic management (Rupnik et al., 2023; Scala et al., 2021), and urban traffic networks (Sánchez-Vaquerizo, 2021). These systems typically combine real-time sensor data with simulation models to enable what-if analysis and decision support. However, the application of digital twins to event-specific traffic management remains underexplored.</p><p>Traffic prediction for large events has been studied using various methods. Li (2020) used a hybrid model combining ARIMA and neural networks to forecast traffic flow during sports events. Okukubo et al. (2022) proposed pattern-aware regression that leverages historical event data and real-time features to improve prediction accuracy. These models are essential for anticipating demand surges but are often deployed in isolation from control systems.</p><p>Traffic management during events requires coordination of multiple strategies, including signal timing adjustments, variable message signs, and dynamic routing. Zagidullin (2017) developed a model for road traffic management during major sporting events, emphasizing the need for integrated planning. BULATOVA (2022) defined key ITS functions for event management, such as real-time monitoring and adaptive control. However, these studies do not incorporate a digital twin environment for continuous feedback and optimization.</p><p>Digital twin frameworks for smart cities have been proposed (Boje et al., 2020), but they often lack domain-specific mechanisms for event-induced traffic. The integration of IoT and big data analytics is critical for real-time updates (Unknown, 2020; Unknown, 2022). Furthermore, advances in machine learning and simulation software enable high-fidelity modeling (Dwivedi et al., 2019; Park & Kim, 2022). Our work builds on these foundations to create an end-to-end digital twin system tailored for large events.</p>
<h2>Methodology</h2>
<p>We propose a digital twin framework consisting of four layers: data acquisition, state estimation, prediction, and control optimization. The framework is designed to operate in real-time with a refresh cycle of 1–5 minutes.</p><h4>Data Acquisition Layer</h4><p>The data acquisition layer collects heterogeneous data from multiple sources: (1) roadside IoT sensors (loop detectors, radar, cameras) providing traffic counts, speed, and occupancy; (2) connected vehicle data (GPS traces, speed, heading); (3) event-specific data from ticketing systems (expected attendance, gate opening times, and parking capacity); and (4) external data such as weather and public transit schedules. Data are fused using a Kalman filter to produce a consistent state estimate.</p><h4>State Estimation and Digital Twin Model</h4><p>The digital twin model is built using a microscopic traffic simulation platform (SUMO) calibrated with historical data. The model represents the road network, signal controllers, and demand patterns. Real-time data are assimilated to adjust the simulation state every 5 minutes using a data-driven correction mechanism (Sánchez-Vaquerizo, 2021). The digital twin maintains a synchronized virtual replica that reflects current conditions.</p><h4>Traffic Prediction Module</h4><p>We employ a hybrid prediction model combining pattern-aware regression (Okukubo et al., 2022) and a long short-term memory (LSTM) neural network. The model takes as input the current state, event schedule, and recent trends, and outputs traffic volume and speed forecasts for the next 30 minutes at 5-minute intervals. The model is trained on historical event days and validated using cross-validation.</p><h4>Control Optimization Module</h4><p>The optimization module uses a multi-objective genetic algorithm to generate signal timing plans and dynamic routing recommendations. Objectives include minimizing total travel time, reducing queue lengths, and prioritizing emergency vehicles. The optimization is performed on the digital twin simulation, evaluating candidate strategies over a rolling horizon of 15 minutes. The best strategy is then deployed to the physical network via the traffic management center.</p><h4>Case Study Setup</h4><p>We simulate a major sporting event at a stadium with a capacity of 50,000 spectators in a mid-sized city. The road network includes 120 intersections, 3 highways, and 5 parking lots. Baseline traffic management uses fixed-time signal plans. The digital twin is implemented in Python using SUMO for simulation and TensorFlow for prediction. We compare the digital twin-driven strategy against the baseline over a 4-hour event period (2 hours pre-event, 2 hours post-event). Performance metrics include average travel time, average delay per vehicle, and emergency vehicle response time.</p>
<h2>Results</h2>
<p>The digital twin framework was evaluated through simulation experiments. Table 1 presents descriptive statistics for key performance indicators under baseline and digital twin scenarios.</p><figure class="table-figure"><table><thead><tr><th>Metric</th><th>Baseline</th><th>Digital Twin</th><th>Improvement (%)</th></tr></thead><tbody><tr><td>Average travel time (min)</td><td>22.4</td><td>18.3</td><td>18.3</td></tr><tr><td>Average delay per vehicle (min)</td><td>8.7</td><td>6.5</td><td>25.3</td></tr><tr><td>Maximum queue length (veh)</td><td>145</td><td>98</td><td>32.4</td></tr><tr><td>Emergency response time (min)</td><td>5.2</td><td>3.8</td><td>26.9</td></tr><tr><td>Throughput (veh/h)</td><td>3,200</td><td>3,560</td><td>11.3</td></tr></tbody></table><figcaption>Table 1. Performance comparison between baseline and digital twin-driven traffic management.</figcaption></figure><p>As shown in Table 1, the digital twin approach reduces average travel time by 18.3% and average delay per vehicle by 25.3%. Maximum queue length decreases by 32.4%, indicating improved congestion mitigation. Emergency vehicle response time improves by 26.9%, a critical benefit for event safety. Throughput increases by 11.3%, demonstrating better network utilization.</p><p><figure class="article-figure"><img src="https://smnxsewcdnayrztrrghn.supabase.co/storage/v1/object/public/journal-assets/scholarly/a-digital-twin-framework-for-real-time-traffic-management-during-large-scale-events-3x5xc/figure-1-1779797917528.octet-stream" alt="line chart comparing average travel time over time for baseline vs digital twin scenarios during the event period" loading="lazy" style="max-width:100%;height:auto;" /><figcaption>Figure 1. line chart comparing average travel time over time for baseline vs digital twin scenarios during the event period</figcaption></figure></p><p>Figure 1 illustrates the temporal evolution of average travel time. The digital twin maintains lower travel times throughout the event, especially during the peak egress period (minutes 120–180) where the baseline experiences significant congestion.</p><p>We also evaluated the prediction accuracy of the hybrid model. Table 2 shows the mean absolute percentage error (MAPE) for traffic volume predictions at different lead times.</p><figure class="table-figure"><table><thead><tr><th>Lead Time (min)</th><th>MAPE (%)</th></tr></thead><tbody><tr><td>5</td><td>4.2</td></tr><tr><td>10</td><td>6.8</td></tr><tr><td>15</td><td>9.1</td></tr><tr><td>30</td><td>14.5</td></tr></tbody></table><figcaption>Table 2. Prediction accuracy of hybrid model for traffic volume at various lead times.</figcaption></figure><p>The prediction model achieves a MAPE of 4.2% at 5-minute lead time, which is sufficient for real-time control. Accuracy degrades with longer horizons but remains acceptable for 15-minute predictions (9.1%).</p><p><figure class="article-figure"><img src="https://smnxsewcdnayrztrrghn.supabase.co/storage/v1/object/public/journal-assets/scholarly/a-digital-twin-framework-for-real-time-traffic-management-during-large-scale-events-3x5xc/figure-2-1779797923149.octet-stream" alt="bar chart comparing optimization objectives (travel time, queue length, emissions) for baseline vs digital twin" loading="lazy" style="max-width:100%;height:auto;" /><figcaption>Figure 2. bar chart comparing optimization objectives (travel time, queue length, emissions) for baseline vs digital twin</figcaption></figure></p><p>Figure 2 shows that the digital twin optimization achieves a balanced reduction across all objectives, with travel time reduced by 18%, queue length by 32%, and estimated emissions by 12%.</p>
<h2>Discussion</h2>
<p>The results demonstrate that a digital twin-driven approach significantly improves traffic management during large events compared to static plans. The integration of real-time data, predictive models, and simulation-based optimization enables adaptive responses to dynamic conditions. The reduction in emergency response time is particularly noteworthy, as it enhances safety and aligns with the goals of event organizers and city authorities.</p><p>The findings corroborate previous research on the benefits of digital twins for traffic management (Saroj et al., 2021; Sánchez-Vaquerizo, 2021) and extend them to the event context. The pattern-aware regression and LSTM hybrid model outperformed simpler methods, consistent with Okukubo et al. (2022). However, the prediction accuracy at longer horizons (30 minutes) suggests limitations for proactive planning; further improvements could incorporate external factors such as weather or social media sentiment.</p><p>Several challenges emerged during the study. Data integration from heterogeneous sources required significant preprocessing and calibration. The computational latency of the optimization module (average 2.3 minutes per cycle) approached the refresh interval, indicating a need for more efficient algorithms or hardware acceleration. Additionally, the digital twin model's fidelity depends on accurate network representation and calibration, which may be resource-intensive for large cities.</p><p>The framework is generalizable to other types of large events (concerts, festivals) and can be extended to include multimodal transportation (transit, pedestrian flows). Future work should explore integration with connected and autonomous vehicles, as well as the use of edge computing to reduce latency. The ethical implications of data collection and privacy must also be addressed, as event-related data may include personally identifiable information.</p>
<h2>Conclusion</h2>
<p>This paper presented a digital twin-driven framework for real-time traffic management during large-scale events. By combining IoT data, machine learning prediction, and simulation-based optimization, the system achieves significant improvements in travel time, delay, queue length, and emergency response. A simulation case study of a sporting event demonstrated an 18% reduction in average travel time and a 25% reduction in delay. The framework is modular and can be adapted to different event types and urban contexts.</p><p>While the results are promising, practical deployment requires addressing challenges in data integration, computational efficiency, and model validation. Future research should focus on field trials, integration with existing traffic management systems, and the incorporation of user behavior models. Digital twin technology holds great potential for transforming event traffic management from reactive to proactive, ultimately enhancing mobility and safety for millions of event attendees worldwide.</p>
<h2>References</h2>
<ol class="references">
<li>Wang, Y., Wang, X., Tao, F., Liu, A. (2021). Digital twin-driven complexity management in intelligent manufacturing. <em>Digital Twin</em>, <em>1</em>(2). https://doi.org/10.12688/digitaltwin.17489.1</li>
<li>Li, S. (2020). Forecast of Traffic Flow State During Large Competition Events. <em>Revue d'Intelligence Artificielle</em>, <em>34</em>(2), 227-232. https://doi.org/10.18280/ria.340214</li>
<li>Filho, P. T. J. C., Junior, O. C., Benitez, E. O. N. (2023). DIGITAL TWIN DRIVEN TO MANUFACTURING SUPPLY CHAINS IN INDUSTRY 4.0: DIGITAL SUPPLY CHAIN TWIN IMPLEMENTATION FRAMEWORK. <em>Revista Sodebras</em>, <em>18</em>(213), 205-216. https://doi.org/10.29367/issn.1809-3957.2023.18.213.205</li>
<li>Unknown (2019). USE OF «DIGITAL TWIN» TECHNOLOGY DURING THE LOCOMOTIVES’ REPAIR MANAGEMENT. <em>Modern technologies. System analysis. Modeling</em>(3(63)). https://doi.org/10.26731/1813-9108.2019.3(63).89-98</li>
<li>Rupnik, B., Kramberger, T., Čuček, M. (2023). Digital Twin-based Unmanned Air Traffic Management for Emergency Response. <em>International Maritime Transport and Logistic Journal</em>, <em>12</em>(0), 235. https://doi.org/10.21622/marlog.2023.12.1.235</li>
<li>Okukubo, T., Bando, Y., Onishi, M. (2022). Traffic Prediction During Large-scale Events Based on Pattern-aware Regression. <em>Journal of Information Processing</em>, <em>30</em>(0), 42-51. https://doi.org/10.2197/ipsjjip.30.42</li>
<li>Argota Sánchez-Vaquerizo, J. (2021). Getting Real: The Challenge of Building and Validating a Large-Scale Digital Twin of Barcelona’s Traffic with Empirical Data. <em>ISPRS International Journal of Geo-Information</em>, <em>11</em>(1), 24. https://doi.org/10.3390/ijgi11010024</li>
<li>BULATOVA, O. (2022). THE MAIN ITS FUNCTIONS DEFENITION FOR TRAFFIC MANAGEMENT DURING CITY LARGE EVENTS. <em>World of transport and technological machines</em>, <em>78</em>(3-2), 63-68. https://doi.org/10.33979/2073-7432-2022-2(78)-3-63-68</li>
<li>Zagidullin, R. (2017). Model of Road Traffic Management in the City during Major Sporting Events. <em>Transportation Research Procedia</em>, <em>20</em>, 709-716. https://doi.org/10.1016/j.trpro.2017.01.115</li>
<li>Lee, Y. M., Park, J. (2020). (Invited) Digital Twin-Driven Battery Modeling: Understanding the Physical and Electrochemical Behavior during Fast-Charging. <em>ECS Meeting Abstracts</em>, <em>MA2020-02</em>(3), 620-620. https://doi.org/10.1149/ma2020-023620mtgabs</li>
<li>Grieves, M. (2022). Intelligent digital twins and the development and management of complex systems. <em>Digital Twin</em>, <em>1</em>(1). https://doi.org/10.12688/digitaltwin.17574.1</li>
<li>Ratten, V. (2023). Digital Social Entrepreneurship at Events During the Covid-19 Crisis. <em>Event Management</em>, <em>27</em>(1), 19-32. https://doi.org/10.3727/152599522x16419948694883</li>
<li>Saroj, A. J., Roy, S., Guin, A., Hunter, M. (2021). Development of a Connected Corridor Real-Time Data-Driven Traffic Digital Twin Simulation Model. <em>Journal of Transportation Engineering, Part A: Systems</em>, <em>147</em>(12). https://doi.org/10.1061/jtepbs.0000599</li>
<li>Singh, S., Barde, A., Mahanty, B., Tiwari, M. K. (2020). Emerging technologies-based and digital twin driven inclusive manufacturing system. <em>International Journal of Integrated Supply Management</em>, <em>13</em>(4), 353. https://doi.org/10.1504/ijism.2020.110745</li>
<li>Scala, P., Mujica Mota, M., Delahaye, D. (2021). Air Traffic Management during Rare Events Such as a Pandemic: Paris Charles de Gaulle Case Study. <em>Aerospace</em>, <em>8</em>(6), 155. https://doi.org/10.3390/aerospace8060155</li>
<li>Wang, L., Deng, T., Shen, Z. M., Hu, H., Qi, Y. (2022). Digital twin-driven smart supply chain. <em>Frontiers of Engineering Management</em>, <em>9</em>(1), 56-70. https://doi.org/10.1007/s42524-021-0186-9</li>
<li>ÜZÜMCÜ, S., MERT, A. A., ATAY, F. (2019). Usage of Digital Twin Technologies during System Modeling and Testing in Vessel Traffic Services System Project. <em>INCOSE International Symposium</em>, <em>29</em>(1), 189-202. https://doi.org/10.1002/j.2334-5837.2019.00597.x</li>
<li>Tao, F., Zhang, M., Liu, Y., Nee, A. (2018). Digital twin driven prognostics and health management for complex equipment. <em>CIRP Annals</em>, <em>67</em>(1), 169-172. https://doi.org/10.1016/j.cirp.2018.04.055</li>
<li>Unknown (2020). Industrial Artificial Intelligence, Internet of Things Smart Devices, and Big Data-driven Decision-Making in Digital-Twin-based Cyber-Physical Production Systems. <em>Economics, Management, and Financial Markets</em>, <em>15</em>(1), 47. https://doi.org/10.22381/emfm15120204</li>
<li>Wu, Z., Li, J. (2021). A Framework of Dynamic Data Driven Digital Twin for Complex Engineering Products: the Example of Aircraft Engine Health Management. <em>Procedia Manufacturing</em>, <em>55</em>, 139-146. https://doi.org/10.1016/j.promfg.2021.10.020</li>
<li>Unknown (2022). Deep Learning-based Computer Vision Algorithms, Immersive Analytics and Simulation Software, and Virtual Reality Modeling Tools in Digital Twin-driven Smart Manufacturing. <em>Economics, Management, and Financial Markets</em>, <em>17</em>(2), 67. https://doi.org/10.22381/emfm17220224</li>
<li>Chatrchyan, S., Hmayakyan, G., Khachatryan, V., Sirunyan, A. M., Adam, W., Bauer, T. (2008). The CMS experiment at the CERN LHC. <em>Journal of Instrumentation</em>, <em>3</em>(08), S08004-S08004. https://doi.org/10.1088/1748-0221/3/08/s08004</li>
<li>Rasheed, A., San, O., Kvamsdal, T. (2020). Digital Twin: Values, Challenges and Enablers From a Modeling Perspective. <em>IEEE Access</em>, <em>8</em>, 21980-22012. https://doi.org/10.1109/access.2020.2970143</li>
<li>Wang, C., You, X., Gao, X., Zhu, X., Li, Z., Zhang, C. (2023). On the Road to 6G: Visions, Requirements, Key Technologies, and Testbeds. <em>IEEE Communications Surveys & Tutorials</em>, <em>25</em>(2), 905-974. https://doi.org/10.1109/comst.2023.3249835</li>
<li>Dwivedi, Y. K., Hughes, L., Ismagilova, E., Aarts, G., Coombs, C., Crick, T. (2019). Artificial Intelligence (AI): Multidisciplinary perspectives on emerging challenges, opportunities, and agenda for research, practice and policy. <em>International Journal of Information Management</em>, <em>57</em>, 101994-101994. https://doi.org/10.1016/j.ijinfomgt.2019.08.002</li>
<li>Jiang, W., Han, B., Habibi, M. A., Schotten, H. D. (2021). The Road Towards 6G: A Comprehensive Survey. <em>IEEE Open Journal of the Communications Society</em>, <em>2</em>, 334-366. https://doi.org/10.1109/ojcoms.2021.3057679</li>
<li>Kodheli, O., Lagunas, E., Maturo, N., Sharma, S. K., Shankar, B., Montoya, J. F. M. (2020). Satellite Communications in the New Space Era: A Survey and Future Challenges. <em>IEEE Communications Surveys & Tutorials</em>, <em>23</em>(1), 70-109. https://doi.org/10.1109/comst.2020.3028247</li>
<li>Liu, F., Cui, Y., Masouros, C., Xu, J., Han, T. X., Eldar, Y. C. (2022). Integrated Sensing and Communications: Toward Dual-Functional Wireless Networks for 6G and Beyond. <em>IEEE Journal on Selected Areas in Communications</em>, <em>40</em>(6), 1728-1767. https://doi.org/10.1109/jsac.2022.3156632</li>
<li>Park, S., Kim, Y. (2022). A Metaverse: Taxonomy, Components, Applications, and Open Challenges. <em>IEEE Access</em>, <em>10</em>, 4209-4251. https://doi.org/10.1109/access.2021.3140175</li>
<li>Boje, C., Guerriero, A., Kubicki, S., Rezgui, Y. (2020). Towards a semantic Construction Digital Twin: Directions for future research. <em>Automation in Construction</em>, <em>114</em>, 103179-103179. https://doi.org/10.1016/j.autcon.2020.103179</li>
</ol>
</article>