From Empirical Judgment to Data-Driven Approaches: A Survey of Traffic Reorganization and Management During Urban River-Crossing Corridor Construction

Abstract

Urban river-crossing corridors serve as critical bottlenecks within urban transportation networks, where traffic management during construction periods directly influences urban operational efficiency and socioeconomic activities. Traditional management approaches based on empirical judgment exhibit fundamental limitations when confronting large-scale infrastructure construction projects, including low prediction accuracy, delayed response times, and insufficient systematic coordination. This survey aims to synthesize existing data-driven approaches, identify research gaps, and establish a roadmap for intelligent traffic management advancement. Unlike previous surveys focusing on individual technologies, this review constructs a complete technical chain from data sensing to intelligent decision-making and systematically reveals implementation pathways for paradigm transformation. The research establishes technical architecture encompassing data sensing, intelligent analysis, predictive warning, and decision support systems, while elucidating the application mechanisms of cutting-edge technologies such as multi-source data fusion, artificial intelligence, and digital twins in urban traffic management. Through analysis of six representative engineering case studies from China, the United States, Republic of Korea, Russia, and Europe, including bridge construction, emergency repair, and highway reconstruction projects, the investigation reveals that data-driven approaches not only achieve improvements in technical performance but also facilitate fundamental paradigm shifts in traffic management philosophy from passive response to proactive prevention, and from localized optimization to systematic coordination. The findings enable policymakers to develop standardized frameworks for data-driven traffic systems, assist urban planners in selecting appropriate technologies based on project characteristics, and guide engineers in implementing integrated traffic management solutions during critical infrastructure construction.

1. Introduction

1.1. Research Background and Urgency of Practical Demands

Urban river-crossing corridors function as critical nodes and significant bottlenecks within urban transportation systems, bearing the core responsibility of connecting north–south urban traffic flows. Their operational status directly correlates with the efficiency and stability of entire urban transportation networks. With China’s rapid urbanization and sustained growth in transportation demand, numerous existing river-crossing corridors face pressing challenges including severe capacity inadequacy and structural deterioration, necessitating large-scale reconstruction or renovation projects. The Qianjiang Third Bridge in Hangzhou exemplifies this situation as a central hub within the city’s cross-river corridor system, serving not only as a vital artery for north–south traffic but also as critical infrastructure ensuring normal urban operations. The anticipated two-year major reconstruction and closure of Qianjiang Third Bridge will significantly reduce the overall capacity of Hangzhou’s cross-river corridor system, directly affecting daily commutes of approximately 500,000 residents and substantially impacting urban functions including commercial offices and retail services along the corridor. Concurrently, the forced redistribution of traffic flows will impose enormous pressure on alternative corridors, potentially triggering regional traffic congestion and causing significant disruptions to urban economic activities and social order. The severity of this practical challenge extends beyond Hangzhou, representing a common predicament faced by numerous major and medium-sized cities. Traditional work zone traffic management primarily relies on engineers’ empirical judgment and analogical reasoning from historical cases. This experience-driven management approach has revealed numerous deep-seated issues and technical bottlenecks when confronting such complex and large-scale urban infrastructure construction projects.

Marzouk and Elsayed [1] proposed a Bridge Information Modeling (BrIM)-based assessment framework for analyzing bridge construction impacts on work zone traffic. This framework integrates four functional modules—construction data module, traffic data modeling module, traffic simulation module, and visualization module—to intuitively demonstrate construction phases and corresponding traffic conditions. Their case study results from Cairo’s El-Nahas Bridge indicate that construction phase impacts on work zone traffic exhibit significant variations throughout the project lifecycle, with initial phases producing the most severe impacts accompanied by high user costs, while subsequent impacts moderate based on the specific construction activities in each phase. This finding provides crucial guidance for precision traffic management during urban river-crossing corridor construction, demonstrating that traditional static management approaches cannot effectively address dynamically changing construction impacts and necessitating the establishment of real-time data-based dynamic response mechanisms.

Ullman et al. [2] systematically emphasized the critical role of ITS technology in work zone traffic management through their work zone intelligent transportation systems implementation guide. The guide indicates that work zone ITS represents a comprehensive system encompassing extensive communication, information, and electronic technologies aimed at enhancing traffic safety and mobility in work zones and surrounding areas. Work zone ITS deployment can focus on safety or mobility objectives, typically supporting both goals while improving productivity. These systems predominantly feature portable and temporary characteristics, although certain deployments may utilize existing fixed infrastructure or become permanent installations. The systematic implementation methodology provided by this guide offers new technical pathways and methodological foundations for addressing traffic reorganization challenges during urban river-crossing corridor construction.

1.2. Current State of Technology Development and Core Challenges

Current international research on work zone traffic management and responses to urban critical transportation node disruptions demonstrates clear developmental trends from single technology applications toward comprehensive system integration, from passive responses toward proactive prevention, and from empirical decision-making toward data-driven approaches. In traffic flow rerouting and congestion mitigation, Wang et al. [3] proposed a dynamic adaptive vehicle rerouting strategy based on grid network topology, innovatively combining k-shortest path algorithms with dynamic congestion rerouting algorithms to real-time identify traffic congestion conditions and automatically generate optimized alternative routes. The core contribution of this research lies in establishing an adaptive traffic management framework capable of dynamically adjusting route recommendation strategies according to real-time traffic conditions, demonstrating effective congestion mitigation performance in simulation experiments. This technological achievement provides significant theoretical foundations and technical references for intelligent management of large-scale traffic flow redistribution during urban river-crossing corridor construction.

In the domain of urban river-crossing corridor system vulnerability assessment, Ding et al. [4] conducted forward-looking research that innovatively proposed vulnerability assessment methods considering disruption impacts on travel, specifically targeting the systematic characteristics of urban river-crossing corridors as specialized transportation infrastructure. The research employed complex network theory and trip chain analysis methods to systematically analyze the impact mechanisms of different types of river-crossing corridor disruptions on urban transportation networks. The study revealed a significant counter-intuitive conclusion: river-crossing corridors in peripheral urban areas exhibit higher systematic vulnerability compared to core area corridors, with their disruptions causing broader impact ranges and longer duration effects. This finding provides crucial practical guidance for developing differentiated emergency management strategies and priority ranking systems, indicating that urban river-crossing corridor construction planning and emergency response plan development cannot simply rely on geographical location or traffic volume as criteria for impact assessment, but must establish scientific evaluation systems based on systematic vulnerability analysis.

In decision support system construction, Gong et al. [5] developed a comprehensive decision support system for traffic management during large-scale road network construction. The system’s core architecture incorporates deep integration of three key technical components: first, a traffic state modeling module based on user equilibrium traffic assignment models, capable of accurately simulating and predicting the impacts of large-scale construction activities on regional traffic flow distribution; second, an origin-destination (OD) matrix calibration module employing bi-level optimization models that achieve precise estimation of dynamic traffic demand through nested structures optimizing OD parameters in the upper level and solving traffic assignment in the lower level; finally, a traffic management strategy generation module that automatically generates optimal traffic management strategies based on analysis results from the preceding modules. The system achieved significant effectiveness in practical applications, with recommended lane addition strategies reducing average peak-hour commuting delays from 6.57 min to 5.82 min, representing an 11.42% improvement. This achievement fully demonstrates the enormous potential of data-driven approaches in enhancing traffic management efficiency and decision-making scientific rigor.

The technological evolution of data-driven intelligent transportation systems represents a significant developmental direction in the traffic management field. Zhang et al. [6] proposed milestone perspectives in their comprehensive survey of data-driven intelligent transportation systems, indicating that the availability of vast amounts of data is driving revolutionary changes in ITS development, transforming traditional technology-driven systems into more powerful multifunctional data-driven intelligent transportation systems. These novel systems exhibit three distinctive characteristics: high visualization capability, displaying complex traffic state information through intuitive visualization interfaces; diversified data sources, integrating heterogeneous data from different sensors and platforms; learning algorithm-driven functionality, possessing the capability to continuously enhance system performance through ongoing learning and optimization. Regarding traffic control system development, Wang and Yang [7] conducted a systematic review of key technologies for next-generation urban traffic control systems, identifying new trends and characteristics in current Chinese urban traffic control technology development. With the large-scale development and engineering applications of emerging traffic technologies such as vehicle-infrastructure cooperation and autonomous driving, real-time interaction between traffic controllers and controlled objects has gained unprecedented technical support. These technological advances provide solid foundations for creating next-generation intelligent traffic systems, with new traffic optimization control systems requiring four core characteristics: high refinement, achieving precise control at lane or even vehicle levels; precision, making decisions based on real-time high-accuracy data; rapid responsiveness, responding to traffic condition changes within second or minute timescales; enhanced intelligence, possessing autonomous learning and decision optimization capabilities. This technological development provides entirely new technical frameworks and implementation pathways for addressing complex traffic problems during urban river-crossing corridor construction.

1.3. Review Methodology and Technical Approach

This review employs a structured narrative literature review methodology specifically designed to synthesize the emerging and interdisciplinary field of data-driven traffic management during urban river-crossing corridor construction. Given the heterogeneous nature of technologies, diverse implementation contexts, and rapidly evolving methodologies in this field, a narrative approach enables comprehensive synthesis and expert interpretation of complex technological paradigms that cannot be meaningfully subjected to statistical meta-analysis.

The literature search strategy was designed to ensure comprehensive coverage of international transportation engineering, computer science, and intelligent systems literature. Primary search databases included Web of Science Core Collection, IEEE Xplore Digital Library, ScienceDirect, Transportation Research International Documentation (TRID), and Engineering Index (EI Compendex) to capture peer-reviewed academic publications. Supplementary searches were conducted in China National Knowledge Infrastructure (CNKI) for Chinese language publications and manual searches of proceedings from significant academic conferences including IEEE Intelligent Transportation Systems conferences, Transportation Research Board Annual Meetings, and International Intelligent Transportation Systems conferences to ensure inclusion of recent developments and gray literature.

The search strategy employed comprehensive combinations of multi-domain keywords structured around core concepts. Traffic management terms included “traffic reorganization”, “work zone management”, “construction traffic control”, and “bridge construction traffic”. Technology-focused terms encompassed “data-driven”, “intelligent transportation”, “machine learning”, “artificial intelligence”, and “traffic flow prediction”. Context-specific terms covered “river-crossing passage”, “bridge construction”, “urban infrastructure”, and “traffic impact assessment”. Advanced technical terms included “multi-source data fusion”, “digital twin”, “traffic simulation”, “sensor fusion”, “big data analytics”, “edge computing”, and “vehicle-road cooperation”. Boolean operators (AND, OR) were employed to construct targeted search expressions, with language filters applied to include both Chinese and English research outputs.

The temporal scope primarily focused on literature from 2010 to 2025 to comprehensively capture the complete technological development process from basic traffic management techniques toward intelligent data-driven approaches, while selectively including historically significant foundational works from 1998 to 2009 that are important for understanding technological development trajectories. This timeframe ensures coverage of both the theoretical foundations and the most recent technological advances in the field.

The literature screening process followed multi-stage assessment criteria designed to ensure both comprehensiveness and quality. Initial automated screening was conducted based on title relevance and abstract content to remove obviously irrelevant literature. Subsequently, full-text review was performed for studies meeting inclusion criteria, focusing specifically on data-driven traffic management technology applications in urban traffic contexts. Quality assessment prioritized peer-reviewed high-impact factor journal papers, renowned international conference papers, and technical reports from authoritative research institutions. Additional high-quality relevant research was identified through backward and forward citation tracking methods to ensure comprehensive literature coverage.

Inclusion criteria specifically targeted research domains encompassing data-driven traffic management methods and technologies, urban river-crossing corridor traffic impact analysis, multi-source heterogeneous data fusion technologies, innovative applications of artificial intelligence in transportation systems, actual implementation cases of intelligent transportation systems, traffic simulation and prediction technologies, and integration solutions for emerging technologies with traditional traffic management systems. The screening process resulted in the inclusion of 68 primary studies that directly address the research scope, supplemented by additional foundational works that provide important theoretical context.

Quality assessment was conducted using evaluation criteria adapted for the diverse study types encountered in this interdisciplinary field. Studies were evaluated based on methodological rigor including appropriate research design and clear methodology description, practical relevance demonstrated through real-world application or validation evidence, technical soundness with adequate technical detail and proper validation procedures, publication credibility through journal impact factor or established conference reputation, and contribution significance through novel insights or substantial advancement to field knowledge.

The synthesis approach follows a thematic logical organizational framework covering seven core technical domains and application directions. These include paradigm shift mechanism analysis from traditional empirical methods to data-driven approaches, traffic model construction and impact assessment methodological systems during urban river-crossing corridor construction, theoretical foundations and implementation pathways for multi-source data fusion and intelligent sensing technologies, key algorithms and application modes for traffic prediction and intelligent decision support technologies, integrated platform architecture design and systems engineering implementation solutions, analysis and effectiveness evaluation of typical engineering application practice cases, and technological development trend predictions with future research direction prospects.

Data extraction and analysis were structured to capture both quantitative performance metrics where available and qualitative insights regarding implementation challenges, success factors, and technological evolution patterns. The comparative analysis framework enables systematic understanding of the intrinsic connections between technical theoretical foundations and practical engineering applications, while accurately identifying current research status and future development opportunities in data-driven traffic management during urban river-crossing corridor construction.

This review conducts systematic analysis according to a thematic logical organizational framework, covering seven core technical domains and application directions: (1) paradigm shift mechanism analysis from traditional empirical methods to data-driven approaches; (2) traffic model construction and impact assessment methodological systems during urban river-crossing corridor construction; (3) theoretical foundations and implementation pathways for multi-source data fusion and intelligent sensing technologies; (4) key algorithms and application modes for traffic prediction, early warning, and intelligent decision support technologies; (5) integrated platform architecture design and systems engineering implementation solutions; (6) analysis and effectiveness evaluation of typical engineering application practice cases; (7) technological development trend predictions; (8) future research direction prospects. This organizational structure facilitates systematic understanding of the intrinsic connections between technical theoretical foundations and practical engineering applications, while accurately identifying the current research status and future development opportunities in data-driven traffic management during urban river-crossing corridor construction.

2. From Empirical Methods to Data-Driven: Paradigm Transformation and Intelligent Computing Technology Foundations

The historical evolution of traffic management methods during urban river-crossing corridor construction profoundly reflects the paradigm transformation in transportation engineering from qualitative empirical analysis to quantitative scientific analysis, from static passive management to dynamic proactive control, and from individual empirical decision-making to collective intelligent decision-making. This transformation not only manifests innovations in technical means and tools but more importantly represents systematic changes in management philosophy, methodology, and technical architecture. We need to understand how this transformation actually works. This chapter examines the limitations of traditional methods, then analyzes the advantages of data-driven approaches. Finally, we explore how holographic sensing and intelligent computing can support modern traffic management systems.

2.1. Technical Bottlenecks and Fundamental Limitations of Traditional Empirical Judgment Methods

Traditional work zone traffic management methods are primarily established on the foundations of engineering practice experience accumulation, historical case analogical reasoning, and expert knowledge judgment. This experience-driven management approach has exposed a series of deep-seated technical bottlenecks and fundamental limitations when addressing complex and variable urban traffic environments, particularly large-scale infrastructure construction projects.

2.1.1. Systematic Issues of Insufficient Prediction Accuracy

Dickerson et al. [8] systematically identified and analyzed core technical challenges faced by traditional empirical judgment methods through in-depth case studies of work zone management in Washington, D.C. The research revealed that traffic state predictions based on historical experience often exhibit significant deviations from actual conditions, with such low-precision predictions severely affecting the scientific rigor and effectiveness of traffic management decisions. More critically, traditional methods demonstrate significantly delayed response times, typically requiring hours or even days from problem identification to effective countermeasure implementation. This lag often leads to further problem deterioration in rapidly changing urban traffic environments. Additionally, various management departments make independent decisions based on their respective experiences, lacking systematic coordination capabilities and struggling to achieve synergistic effects, frequently resulting in scenarios where local optimization leads to overall performance degradation. These issues are particularly pronounced in the management of critical traffic nodes such as urban river-crossing corridors, as corridor interruptions often trigger regional or even citywide traffic chain reactions, with traditional method limitations being further amplified in such complex scenarios.

2.1.2. Technical Deficiencies in Capacity Estimation Methods

In the core technical component of work zone capacity estimation, traditional method limitations manifest more obviously and profoundly. Jehn and Turochy [9] discovered through specialized research on rural freeway work zones that current work zone capacity estimation methods in the sixth edition of the Highway Capacity Manual, while significantly improved compared to historical versions, still exhibit fundamental technical deficiencies and methodological problems. For instance, existing methods over-rely on deterministic parameter settings, employing fixed average queue discharge rates to define capacity. This deterministic approach completely ignores the stochastic characteristics of traffic flow operation and breakdown occurrence, failing to reflect the complexity and uncertainty of actual traffic systems. Traditional methods severely neglect the influence of dynamic changing factors, inadequately considering the significant impacts of various dynamic factors such as weather condition changes, driver behavior variations, construction schedule adjustments, and incident occurrences on actual capacity. Unified calculation formulas and parameter settings lack personalized adaptation capabilities, struggling to accommodate different regional traffic characteristics, specific features of different project types, and traffic demand patterns across various time periods.

Mashhadi et al. [10] provided a comprehensive review of work zone capacity estimation methods that further confirmed and deepened the aforementioned perspectives. The research systematically examined existing parametric methods, non-parametric methods, and traffic simulation approaches, discovering that while these methods each possess technical characteristics and application advantages, they all share common technical limitations: method effectiveness highly depends on historical data completeness, accuracy, and representativeness, with method reliability declining sharply once foundational data exhibits gaps or quality issues; methods demonstrate severely insufficient generalization capabilities, with prediction accuracy often deteriorating rapidly when confronting new engineering scenarios, abnormal traffic conditions, or emergency events, failing to provide reliable decision support; existing methods generally lack adequate real-time performance capabilities, unable to provide real-time dynamic predictions and optimization recommendations, struggling to meet the urgent demands for rapid response in modern urban traffic management. These fundamental limitations render traditional methods inadequately equipped to effectively address the complex, variable, and dynamically evolving traffic management requirements during urban river-crossing corridor construction.

2.1.3. Fundamental Deficiencies in System Complexity Handling Capabilities

From in-depth analysis through systems science and complexity theory perspectives, the fundamental limitations of traditional empirical judgment methods lie in their rigid linear thinking patterns and static analysis frameworks. Traffic systems during urban river-crossing corridor construction represent typical complex mega-systems, exhibiting multiple levels of system complexity characteristics: nonlinear features manifested as complex nonlinear relationships between system outputs and inputs, where small disturbances may trigger large system responses; dynamic characteristics reflected in continuous system state changes over time with obvious time-varying properties; stochastic features demonstrated in system behaviors influenced by multiple random factors with inherent uncertainty; emergent characteristics showing that overall system behavior cannot be simply predicted through superposition of partial behaviors, exhibiting obvious emergent phenomena.

Traditional empirical methods, based on simplified assumption linear analysis frameworks, obviously cannot effectively handle such multi-dimensional and multi-level system complexity. This fundamental methodological deficiency determines that their limitations when confronting complex urban traffic management problems are structural and insurmountable. Therefore, resolving these limitations must rely on paradigm shifts and methodological innovations, with data-driven intelligent methods representing important manifestations of such innovation.

2.2. Technical Advantages and Revolutionary Breakthroughs of Data-Driven Methods

Data-driven methods provide systematic solutions and revolutionary technical breakthroughs for the fundamental limitations faced by traditional methods through introducing advanced computing technologies, intelligent algorithms, and big data analytics technologies, achieving improvements across multiple key dimensions including prediction accuracy, response speed, adaptation capability, and system coordination.

2.2.1. Significant Advantages and Breakthrough Progress of Intelligent Prediction Technologies

In the key application domain of work zone traffic delay prediction, the hybrid machine learning model developed by Du et al. [11] represents a typical manifestation of data-driven method technical advantages. This model constructs an intelligent system specifically designed for predicting work zone spatiotemporal delays through innovative integration of Artificial Neural Network (ANN) and Support Vector Machine (SVM) technologies. The model’s technical innovations are primarily manifested in multiple aspects: first, the system comprehensively considers multi-dimensional influencing factors including road geometric characteristics, lane closure quantities, and work zone duration across different time periods, overcoming the limitation of single-factor consideration in traditional methods; second, through the nonlinear modeling capabilities of neural networks and the generalization performance of support vector machines, precise modeling of complex traffic phenomena is achieved; finally, the model possesses adaptive learning capabilities, continuously optimizing prediction performance based on new observational data, which represents an important characteristic unattainable by traditional static methods. Experimental results demonstrate that this hybrid model significantly outperforms traditional single-model approaches in terms of root mean square error, providing work zone planners with more precise and reliable decision support.

The application of deep learning technology in traffic flow prediction represents another important breakthrough direction for data-driven methods. The big data-based deep learning traffic flow prediction method proposed by Lv et al. [12] successfully applies deep architecture models to the traffic flow prediction domain, achieving significant methodological innovation. This method employs stacked autoencoder models to learn universal traffic flow features, effectively handling complex spatiotemporal correlations inherent in traffic data. The core of technical innovation lies in using autoencoders as basic building blocks to represent and learn deep-level feature patterns of traffic flow. This deep feature learning approach can automatically discover hidden complex patterns and regularities in data without relying on manual feature engineering. Experimental results demonstrate that this method achieves high levels of accuracy in traffic flow prediction, providing powerful technical tools for traffic state prediction during urban river-crossing corridor construction. Wang et al. [13] further validated and extended the technical advantages of data-driven methods through systematic empirical comparative studies of deep learning, ensemble strategies, and performance evaluation. The research discovered through large-scale experiments that deep learning models generally outperform traditional shallow learning methods in urban road travel time prediction, with LSTM-DNN models particularly achieving optimal MAPE values across all prediction scenarios with 30 min sliding time windows. This finding holds important practical significance for traffic management during urban river-crossing corridor construction, as accurate travel time prediction constitutes the key foundation for formulating effective traffic guidance strategies, implementing dynamic route planning, and optimizing signal control schemes. Research results indicate that deep learning methods not only excel in prediction accuracy compared to traditional methods but also demonstrate obvious advantages in handling complex spatiotemporal dependencies, adapting to data distribution changes, and processing missing data.

2.2.2. Collaborative Optimization Capabilities of Intelligent Transportation Systems

The Multiagent Reinforcement Learning for Integrated Network of Adaptive Traffic Signal Controllers (MARLIN-ATSC) system proposed by El-Tantawy et al. [14] provides a successful case study for intelligent computing technology applications in large-scale urban traffic management. This system specifically targets integrated adaptive traffic signal controller network design and underwent comprehensive validation on a large-scale complex simulation network containing 59 intersections in downtown Toronto. The system’s technical innovation lies in adopting a multi-agent reinforcement learning framework, enabling each intersection’s signal controller to function as an independent agent that learns optimal control strategies through environmental interaction while achieving network-level collaborative optimization through inter-agent information exchange and coordination mechanisms.

Experimental results demonstrate network-level average intersection delay reductions of 27–39%, with travel time savings of 15–26% along Toronto downtown’s busiest routes. These quantitative achievements showcase the potential and value of intelligent computing technology in addressing large-scale complex traffic network coordination control problems. From the perspective of traffic management during urban river-crossing corridor construction, the successful application of this system proves that reinforcement learning-based intelligent algorithms can effectively address complex optimization problems arising from traffic flow redistribution, providing feasible technical pathways for achieving regional traffic network collaborative optimization.

2.2.3. Core Technical Advantages of Data-Driven Methods

The fundamental sources of data-driven method technical advantages lie in their unique technical characteristics and methodological innovations, primarily manifested in the following core aspects:

Nonlinear modeling capabilities: Through advanced algorithms such as deep neural networks, support vector machines, and random forests, data-driven methods can effectively handle complex nonlinear relationships in traffic systems, capturing complex feature patterns unidentifiable by traditional linear methods, achieving more accurate and in-depth modeling of traffic phenomena.

Adaptive learning mechanisms: Unlike traditional static methods, data-driven methods possess capabilities for continuous learning and self-optimization, continuously adjusting model parameters and prediction strategies based on real-time observational data and historical experience, maintaining excellent prediction performance when confronting new traffic scenarios and changing environmental conditions.

Multi-source data fusion characteristics: Data-driven methods can comprehensively utilize multi-source heterogeneous information from different sensors, platforms, and spatiotemporal scales, achieving effective information integration through advanced data fusion algorithms, thereby enhancing prediction and decision-making accuracy and robustness.

Real-time response capabilities: Based on efficient computing architectures and optimization algorithms, data-driven methods can achieve rapid responses at second or minute levels, meeting the stringent real-time requirements of modern urban traffic management.

From the specific requirements of traffic management during urban river-crossing corridor construction, the technical advantages of data-driven methods are concentrated across three levels: at the sensing level, comprehensive, multi-dimensional, and real-time traffic state perception is achieved, accurately capturing the complex impacts of construction activities on traffic systems; at the analysis level, precise, personalized, and predictive traffic impact assessment is provided, accurately predicting traffic impact degrees and spatial distributions across different construction phases; at the decision level, scientific, collaborative, and adaptive traffic management strategy formulation is supported, enabling dynamic adjustment of management measures based on real-time traffic conditions.

2.3. Technical Implementation Pathways for Holographic Sensing and Intelligent Computing

The deep integration of holographic sensing and intelligent computing technologies represents the latest direction and highest level of data-driven traffic management technology development, providing powerful technical support and systematic solutions for traffic management during urban river-crossing corridor construction.

2.3.1. Large-Scale Intelligent Transportation Network Collaborative Control Technology

Regarding short-term traffic prediction technology development, research by Vlahogianni et al. [15] indicates that since the early 1980s, short-term traffic forecasting has been a core component of intelligent transportation systems research and applications. With substantial improvements in computing technology capabilities and continuous refinement of mathematical modeling methods, researchers have gained unprecedented technical opportunities to expand research horizons and application boundaries. Through systematic analysis of current technology development status and application requirements, the research identified 10 challenging yet relatively under-researched important development directions, including multimodal traffic prediction, heterogeneous data fusion, real-time adaptive prediction, uncertainty quantification, and explainability enhancement. These development directions provide important guidance value for improving traffic prediction accuracy during urban river-crossing corridor construction, enhancing system reliability, and user acceptance.

Sayed et al. [16] provided an in-depth analysis of innovative opportunities and technological transformations brought by Internet of Things (IoT) technology expansion in their comprehensive review of artificial intelligence-based traffic flow prediction. The research emphasizes that with the rapid development and widespread deployment of IoT technologies, a series of creative solutions represented by smart cities have emerged, making people’s daily lives more efficient, convenient, and intelligent. Intelligent transportation systems, as core components of smart city ecosystems, have been deeply integrated into multiple smart city application scenarios, playing important roles in improving transportation efficiency, enhancing travel service quality, and optimizing resource allocation. This development trend provides more abundant technical choices and flexible implementation pathways for traffic management during urban river-crossing corridor construction, enabling managers to select the most suitable technical solutions based on specific requirements and constraint conditions.

2.3.2. Four-Layer Technical Architecture for Holographic Sensing and Intelligent Computing

From the systematic perspective of technical implementation pathways, the integrated application of holographic sensing and intelligent computing is embodied in a complete four-layer technical architecture, with each layer possessing unique technical characteristics and functional positioning:

Data Collection and Sensing Layer: This fundamental layer of the entire system is responsible for constructing comprehensive and three-dimensional sensing networks through deployment of multi-type sensor networks, high-definition video surveillance systems, mobile communication equipment, and vehicle-mounted terminal devices, achieving real-time collection of multi-dimensional information including traffic flow, vehicle speed, lane occupancy, traffic incidents, and environmental conditions. In urban river-crossing corridor construction scenarios, technical innovation at this level is particularly important, requiring establishment of stable and reliable data collection systems within limited spatial constraints and complex construction environments. Technical challenges primarily include sensor interference resistance, data transmission stability, and equipment maintenance convenience.

Data Processing and Fusion Layer: This layer employs modern computing architectures including edge computing, cloud computing, and fog computing to conduct real-time processing, intelligent cleaning, feature extraction, data fusion, and standardization processing of massive multi-source heterogeneous data, providing high-quality and standardized data foundations for upper-level intelligent analysis. Technical challenges include achieving low-latency real-time processing while ensuring data quality, effectively handling spatiotemporal data inconsistencies, and ensuring reliability and security of data processing procedures.

Intelligent Analysis and Decision Layer: This core system layer employs artificial intelligence technologies including machine learning, deep learning, reinforcement learning, and complex network analysis to establish intelligent analysis model clusters such as traffic state recognition models, traffic flow prediction models, impact assessment models, and risk warning models, providing scientific foundations and intelligent support for traffic management decisions. This level represents the concentrated manifestation of data-driven method technical advantages, representing the system’s core competitiveness and technical level.

Application Service Support Layer: This user and manager-oriented service layer forms scientific traffic management strategies, emergency response plans, and user service information based on intelligent analysis results, combined with traffic management business rules, expert knowledge, and best practices. This level requires achieving organic integration of intelligent algorithms and expert experience while ensuring decision scientific rigor and solution operability and practicality.

The organic integration and collaborative operation of these four levels constitute a complete data-driven traffic management technology system, providing systematic and scientific technical solutions for traffic reorganization and management during urban river-crossing corridor construction. Through comprehensive application of this technology system, transformations from passive response to proactive prevention, from empirical decisions to scientific decisions, and from local optimization to system coordination can be achieved, thereby enhancing the scientific, precise, and intelligent levels of traffic management.

2.3.3. Key Success Factors for Technical Implementation Pathways

To ensure successful application of holographic sensing and intelligent computing technologies in traffic management during urban river-crossing corridor construction, the following key success factors require focused attention:

Technical Standardization: Establishing unified data standards, interface specifications, and communication protocols to ensure interoperability and compatibility between different subsystems and products from different manufacturers, providing guarantees for long-term stable system operation and continuous upgrades.

Data Quality Assurance: Establishing comprehensive data quality control mechanisms, including data collection quality monitoring, data transmission error detection, data processing anomaly identification, and data storage integrity verification, ensuring high-quality data foundations for intelligent analysis.

Algorithm Model Optimization: Continuously optimizing and improving intelligent analysis algorithms, enhancing model prediction accuracy, generalization capabilities, and computational efficiency, ensuring systems can adapt to continuously changing traffic environments and management requirements.

Human–Machine Collaboration Mechanisms: Establishing effective human–machine collaboration mechanisms, fully leveraging the computational capabilities of artificial intelligence systems and the experiential advantages of human experts, achieving organic unity between intelligence and humanity.

Security Assurance Systems: Constructing comprehensive network security and data security assurance systems, ensuring system and data security and reliability while preventing various security threats and risks.

Through comparative analysis in

Table 1. Comparison of Technical Characteristics Between Traditional Empirical Methods and Data-Driven Methods.

Technical Dimension	Traditional Empirical Methods	Data-Driven Methods	Technical Breakthroughs and Advantages
Data Foundation	Historical statistical data, expert experiential knowledge	Multi-source real-time big data, intelligent analysis algorithms	Rich data dimensions, broad coverage, strong real-time capability
Modeling Approach	Linear regression, empirical formula derivation	Deep learning, machine learning, neural networks	Nonlinear modeling capability, adaptive learning mechanisms
Prediction Accuracy	Insufficient prediction accuracy in traditional methods	Advanced models achieving accuracy above 80%	Significantly improved prediction accuracy, substantially reduced errors
Response Speed	Hour-level to day-level response time	Minute-level, second-level real-time response	Fundamental enhancement in real-time response capability
Adaptation Capability	Static fixed parameter mode	Dynamic adaptive parameter adjustment	Stronger adaptability to complex changing scenarios
System Coordination	Relatively independent subsystem operation	Multi-system deep integration collaborative optimization	Significantly enhanced systematic coordination capability
Decision Mechanism	Qualitative analysis and empirical judgment dominated	Quantitative analysis combined with intelligent reasoning	Substantially improved decision scientific rigor and precision
Learning Capability	Relying on manual experience accumulation and inheritance	Automated continuous learning optimization mechanisms	Continuous improvement and knowledge accumulation capability

, it can be clearly observed that data-driven methods demonstrate significant advantages over traditional methods across all key technical dimensions, providing more scientific, efficient, precise, and intelligent technical solutions for traffic management during urban river-crossing corridor construction. This transformation represents not merely simple innovation in technical means and analytical tools, but systematic changes in traffic management philosophy, methodological systems, and technical architecture, possessing theoretical significance and practical value for constructing future-oriented modernized urban traffic management systems.

The comparative analysis presented in Table 1 reveals consistent patterns of improvement across all technical dimensions when transitioning from traditional empirical methods to data-driven approaches. Response speed improvements are particularly notable, with data-driven systems achieving minute-level or second-level real-time responses compared to hour-level or day-level responses in traditional methods. Prediction accuracy enhancements represent another significant advancement, with data-driven models achieving accuracy rates above 80% compared to insufficient accuracy levels in traditional approaches. These quantitative improvements translate directly into practical benefits for traffic management during urban river-crossing corridor construction, enabling more responsive and effective management strategies that can adapt to rapidly changing conditions.

The analysis in this chapter sets the stage for examining traffic models, data fusion technologies, and prediction systems. By comparing traditional and data-driven approaches, we establish a foundation for the technical discussions that follow.

3. Traffic Model Construction and Impact Assessment Analysis During Corridor Construction

Following the analysis of data-driven method technical foundations in Section 2, this chapter enters the core component of traffic management implementation—traffic model construction and impact assessment. During the transformation from experience-driven to data-driven approaches, traffic models have evolved from static empirical parameter models to dynamic data-driven intelligent models, providing scientific foundations for precise assessment of traffic impacts during river-crossing corridor construction.

3.1. Traffic Model Construction for Urban River-Crossing Corridor Closure Construction Scenarios

3.1.1. Multi-Source Heterogeneous Data-Driven Modeling Methodological Innovation

The core breakthrough in data-driven traffic modeling lies in the organic fusion and deep mining of multi-source heterogeneous data. Regarding the balance between privacy protection and collaborative modeling, federated learning technology provides innovative solutions for cross-departmental data collaboration involved in urban river-crossing corridor construction. The distributed statistical model training method by Li et al. [17] achieves unification of data localization and global model optimization, with this technical pathway being particularly suitable for data collaboration requirements among multiple departments including transportation, traffic police, and urban construction during river-crossing corridor construction.

The technological innovation in spatiotemporal traffic flow modeling is equally remarkable. The advantages demonstrated by deep spatiotemporal residual networks (ST-ResNet) in citywide crowd flow prediction have laid foundations for precise prediction of complex traffic flow changes during river-crossing corridor construction [18]. This technology can simultaneously capture temporal closeness, periodicity, and trend characteristics of traffic flows, which holds important value for understanding redistribution patterns of traffic demand after corridor closure. The development of traffic control coordination optimization theory also faces challenges of “practice preceding theory.” Research by Wang and Yang [19] reveals that while non-conventional lane applications have achieved positive results in practice, deficiencies still exist in theoretical systems, resource utilization, and safety assessment, precisely illustrating the important role of data-driven methods in bridging the gap between theory and practice.

3.1.2. BrIM-Based Traffic Impact Visualization Modeling

The introduction of Bridge Information Modeling (BrIM) technology has pioneered new paradigms in traffic impact assessment. The framework developed by Marzouk and Elsayed [1] achieves comprehensive visualization from engineering construction to traffic states through four core modules: construction progress tracking, traffic data modeling, simulation analysis, and virtual reality display. This integrated approach not only achieves dynamic correlation between engineering and traffic but more importantly provides decision-makers with intuitive impact assessment tools.

Results from the El-Nahas Bridge case study indicate that construction phase impacts on work zone traffic exhibit obvious time-varying characteristics: initial project phases produce maximum impacts accompanied by high user costs, with impacts gradually diminishing as construction progresses. This finding provides important foundations for formulating phased traffic management strategies while demonstrating the unique value of BrIM methods in revealing complex spatiotemporal change patterns.

3.2. Application of Mixed Traffic Flow Simulation Technology in Traffic Reorganization

3.2.1. Multi-Level Collaborative Simulation Technology System

The development of mixed traffic flow simulation technology exhibits significant characteristics of multi-level collaboration. The Urban Traffic Mobility Optimization Model (UTMOM) achieves cross-dataset traffic dynamic analysis through deep fusion of data mining and mathematical modeling [20]. This method particularly emphasizes reducing variability and identifying traffic intensity changes, holding important value for reconstructive analysis of peak-hour traffic patterns during river-crossing corridor construction.

The technical support for traffic engineering solutions exhibits multi-dimensional development trends. From resilience-perspective traffic network analysis and pollutant distribution modeling to camera-based intelligent monitoring system development, the comprehensive application of these technologies provides full support for simulation analysis in complex traffic environments [21]. Particularly noteworthy is the application of these technologies in environmental impact assessment and safety risk prevention, providing important tools for comprehensive impact assessment during urban river-crossing corridor construction.

The systematic compilation of work zone management experience in developing countries provides important references for simulation technology applications. Through in-depth analysis of three core indicators including work zone travel delays, queue lengths, and user costs, researchers have identified key factors affecting simulation accuracy [22]. These findings hold important guidance significance for constructing more accurate simulation models for urban river-crossing corridor construction.

3.2.2. Simulation Accuracy Enhancement and Validation Method Innovation

The improvement of simulation technology precision relies on innovations in modeling methods. The application of Improved Cellular Automata (ICA) models in work zone traffic modeling demonstrates the potential of technological progress [23]. The innovation of this model lies in introducing more realistic lateral speed and position update rules, making vehicle lateral movement simulation closer to actual conditions. This enhancement in modeling precision holds important significance for accurately predicting traffic state changes during river-crossing corridor construction.

The capability expansion of microscopic traffic simulation technology is equally noteworthy. The latest developments in SUMO simulator regarding multimodal transportation solutions, simulator coupling, and model development and validation provide powerful tools for precise modeling of complex traffic environments [24]. These technological advances enable simulation to extend beyond single transportation modes, comprehensively considering behavioral characteristics of multiple traffic participants including pedestrians, bicycles, public transit, and private vehicles.

3.3. Travel Pattern Change Prediction and Impact Assessment During Corridor Construction

3.3.1. Deep Mechanisms of Travel Pattern Evolution Patterns

The theoretical foundations of travel pattern change prediction are undergoing profound transformations. Research on work zone scheduling problems based on day-to-day traffic assignment models incorporates consideration of day-by-day progression from non-equilibrium to equilibrium states [25]. This theoretical breakthrough holds important significance for understanding traveler behavioral adaptation processes during urban river-crossing corridor construction, particularly in revealing dynamic conversion mechanisms between short-term and long-term equilibrium.

The research discovered temporal patterns of travel mode adaptation: when construction duration is relatively short, optimal scheduling strategies change significantly with duration; when duration extends to specific thresholds, optimal construction sequences tend to stabilize. This finding not only reveals temporal characteristics of traveler learning adaptation but also provides scientific foundations for formulating phased traffic management strategies. Simultaneously, guiding travelers to change original travel habits can reduce total network travel costs during construction periods, providing theoretical support for proactive traffic demand management.

3.3.2. Systematic Innovation in Traffic Impact Assessment Methods

The development of traffic system simulation technology exhibits four important trends: variable granularity simulation, deep simulation, digital twins, and simulation as a service [26]. These technological trends not only represent technical progress in assessment methods but more importantly reflect fundamental transformations from single functions to comprehensive services, from offline analysis to real-time decision-making, and from professional tools to widespread applications.

Variable granularity simulation technology achieves dynamic balance between assessment precision and computational efficiency, enabling flexible switching among microscopic, mesoscopic, and macroscopic levels according to specific requirements. Deep simulation technology significantly enhances modeling capabilities for complex traffic phenomena through machine learning algorithm integration. Digital twin technology constructs real-time digital mirrors of physical traffic systems, providing novel platforms for dynamic assessment and prediction. Simulation as a service models substantially reduce application thresholds for advanced assessment technologies through cloud computing platform support.

The fundamental theories of traffic simulation are also continuously deepening. The enhancement of computer technology capabilities, software engineering development, and emergence of intelligent transportation systems have driven the transformation of simulation technology from academic tools to professional applications [27]. The unique capability of simulation technology to model temporal variability of traffic phenomena makes it an ideal tool for capturing traffic system complexity, which holds irreplaceable value for traffic impact assessment of complex projects such as urban river-crossing corridor construction.

The traffic model construction and impact assessment technologies surveyed in this chapter constitute the core technical support for the transformation from empirical judgment to data-driven approaches. Multi-source heterogeneous data-driven modeling methods address the limitations of traditional empirical methods in data acquisition and processing capabilities, BrIM visualization modeling technology achieves intuitive correlation between engineering construction and traffic impacts, mixed traffic flow simulation technology provides multi-level collaborative analysis capabilities, and travel pattern change prediction methods reveal intrinsic patterns of complex traffic phenomena.

The common characteristics of these technological innovations lie in transitions from static to dynamic, from single to integrated, and from empirical to intelligent. They not only enhance the scientific rigor and accuracy of traffic impact assessment during urban river-crossing corridor construction but more importantly establish solid foundations for subsequent intelligent traffic organization and decision support. Within the overall paper thematic framework, this chapter assumes a crucial bridging role: deepening the data-driven methodologies proposed in Section 2 while providing technical preparations for the intelligent organization strategies in Section 4 and system implementation in Section 5.

4. Intelligent Traffic Organization and Guidance Management Strategies

Based on the technical foundations of traffic model construction and impact assessment from Section 3, this chapter focuses on innovation and implementation of intelligent traffic organization and guidance management strategies. This stage represents the critical transformation of data-driven methods from theoretical analysis to practical application, embodying the evolution of modern traffic management from passive response to proactive prediction, from single-point control to system coordination, and from empirical decision-making to intelligent optimization.

4.1. Traffic Organization Optimization Based on Supply-Demand Matching

4.1.1. Technical Integration and Collaborative Optimization of Intelligent Transportation Systems

The theoretical foundations of intelligent traffic organizations are undergoing profound transformations. Traffic simulation technology, with its unique capability to model temporal variability, has become a core tool for capturing traffic system complexity [27]. This enhancement in technical capability stems not only from improved computer performance and software engineering development but more importantly from the continuous maturation of intelligent transportation system technologies, enabling complex dynamic optimization to transform from theoretical concepts to practical applications.

The potential demonstrated by connected vehicle technology in work zone management is remarkable. Speed harmonization optimization systems achieve significant delay reduction effects when penetration rates reach specific thresholds through precise vehicle state prediction and optimization algorithms [28]. The important significance of this breakthrough lies not only in technical effect validation but more importantly in proving the feasibility of intelligent transportation system transformation from concept verification to scaled applications.

The innovation in traffic estimation technology is equally noteworthy. Addressing the widespread challenge of limited continuous count station sensor coverage, the combination of custom regression models and data augmentation techniques provides effective solutions for precise traffic estimation in sparse sensor environments [29]. The application of variational autoencoders not only fills data gaps but also significantly improves estimation accuracy through synthetic data generation. This technical pathway holds important practical value for situations with limited sensor deployment during urban river-crossing corridor construction.

4.1.2. Intelligent Evolution of Traffic Control Strategies

Traffic control strategies are experiencing evolution from simple feedback control to intelligent collaborative control. The capability of decentralized feedback loop control strategies to approximate user-optimal traffic flow distribution in mixed network environments demonstrates the advantages of distributed control architectures [30]. This control philosophy is particularly suitable for complex traffic environments during urban river-crossing corridor construction, achieving rapid response while avoiding single-point failure risks.

The introduction of multi-objective optimization concepts marks an important transformation of traffic control toward sustainable development directions. Two-class freeway traffic regulation methods consider not only traditional congestion mitigation objectives but also incorporate emission minimization into optimization frameworks [31]. This multi-objective collaborative optimization methodology holds important implications for modern urban traffic management, particularly under increasingly stringent environmental protection requirements.

The development of adaptive traffic signal control systems further embodies the development trend of intelligent control. Systems can perform real-time adjustments according to seasonal variations and short-term fluctuations in traffic demand, with this adaptive capability holding important value for addressing highly uncertain environments during urban river-crossing corridor construction [32]. The advantages of multi-agent reinforcement learning-based signal control methods in real-time performance, accuracy, and self-learning capabilities foreshadow future development directions of traffic control technology.

4.2. Traffic Guidance and Route Optimization During Corridor Construction

4.2.1. Theory and Practice of Multi-Path Collaborative Optimization

The integrated framework for urban sustainable traffic management provides systematic solutions for alternative route management. This framework achieves coordinated unification of short-term optimization and long-term adaptation through organic combination of within-day traffic control methods and day-to-day behavioral route choice models [33]. This bi-level optimization concept holds important guidance significance for traffic management during urban river-crossing corridor construction: ensuring current traffic efficiency while guiding traffic flows toward reasonable long-term distribution patterns.

The theoretical foundations of route guidance effectiveness are continuously deepening. The application of evolutionary game theory in driver decision-making modeling reveals dynamic evolution patterns of traffic flow distribution under route guidance [34]. The characterization of driver decision temporal evolution processes by replicator dynamic equations provides scientific tools for understanding and predicting guidance effects. The two types of stable points discovered in research not only enrich the theoretical connotations of route guidance but also provide important references for guidance strategy design in practical applications.

4.2.2. Systematic Innovation in Regional Collaborative Control

The development of real-time control technology for multi-regional traffic networks embodies the evolution of traffic management toward systematic and intelligent directions. Joint route guidance and demand management strategies achieve more comprehensive traffic system optimization through combination of spatial distribution optimization and temporal regulation optimization [35]. Route guidance is responsible for identifying optimal inter-regional transfer flows to maximize trip completion rates, while demand management controls traffic flows entering the network through origin waiting, with this collaborative mechanism providing effective frameworks for regional traffic coordination during urban river-crossing corridor construction.

The application of Model Predictive Control (MPC) frameworks further enhances the intelligent level of control systems. By simultaneously optimizing demand flows and inter-regional transfer flows, systems can minimize the total time of all vehicles in the network, with this global optimization concept holding important value for addressing traffic redistribution problems during urban river-crossing corridor construction.

4.3. Emergency Response and Management for Sudden Situations

4.3.1. Dynamic Network Modeling and Intelligent Decision Support

The enhancement of emergency response capabilities relies on dynamic network modeling technology support. Analysis of new real-time driver information system impacts on traffic congestion patterns requires extended dynamic network modeling frameworks [36]. This framework not only contains driver route and departure time choice behavioral models but also explicitly incorporates the modeling process of driver information acquisition and integration. This human–machine combined modeling approach can more accurately predict the impacts of sudden events, providing scientific foundations for formulating effective emergency strategies.

The development of control technology for mixed operational environments of multi-technology level vehicles embodies forward-thinking in emergency management. Research on intersection control for coexisting environments of conventional vehicles, connected vehicles, and automated vehicles constructs intelligent control systems adaptable to technological evolution through integration of technology phase switching signal control and branch-and-bound solution methods [37]. This technology development-oriented design philosophy provides important references for emergency management during urban river-crossing corridor construction.

4.3.2. Application of Reinforcement Learning in Emergency Control

The application of reinforcement learning technology in traffic signal control represents an important breakthrough in emergency management technology. Compared to traditional control methods requiring pre-specified environmental models, the core advantage of reinforcement learning lies in control agents’ ability to autonomously learn relationships between control actions and environmental impacts while pursuing objectives [38]. This self-learning capability holds special value for highly uncertain environments during urban river-crossing corridor construction. The value of reinforcement learning in emergency management is reflected not only at the technical level but more importantly in the management philosophy transformation it represents from passive response to proactive learning, from fixed strategies to dynamic adaptation, and from experience dependence to intelligent decision-making. This philosophical transformation points to the direction for constructing more intelligent urban traffic emergency management systems.

The intelligent traffic organization and guidance management strategies elaborated in this chapter represent key components in the transformation of traffic management from traditional empirical modes to data-driven intelligent modes. Supply-demand matching-based organizational optimization achieves transformation from static configuration to dynamic equilibrium, traffic guidance and route optimization embodies upgrades from single-point control to network coordination, and emergency response management demonstrates evolution from passive response to proactive prevention.

The core characteristics of these strategic innovations are embodied in breakthroughs across three dimensions: in the technical dimension, achieving leaps from traditional control to intelligent learning; in the spatial dimension, completing expansion from local optimization to system coordination; in the temporal dimension, accomplishing transformation from post-event processing to predictive early warning. These breakthroughs not only address practical challenges in traffic management during urban river-crossing corridor construction but more importantly provide technical pathways and methodological support for constructing modern urban intelligent traffic management systems.

Within the overall logical framework of this paper, this chapter assumes core functions of technical implementation: transforming the model construction achievements from Section 3 into operational management strategies while providing functional requirements and technical preparations for the integrated platform implementation in Section 5. This bridging role embodies the complete technical chain of data-driven methodologies in traffic management practice, achieving organic unification from theoretical construction to strategy design and system implementation.

5. Integrated Platform Architecture and System Implementation

Based on the in-depth exploration of traffic modeling, impact assessment, and intelligent organization strategies in preceding chapters, this chapter focuses on the critical component of transforming theoretical methods into engineering practice—integrated platform architecture and system implementation. This transformation process not only embodies the important leap of data-driven methods from concept verification to scaled applications but also represents the construction of technological assurance systems for modernizing traffic management during urban river-crossing corridor construction.

5.1. System Architecture Design of Data-Driven Traffic Management Platform

The architecture design of modern traffic management platforms is experiencing transformation from single functions to comprehensive integration. The technology combination of work zone intelligent transportation systems demonstrates this development trend: from dynamic lane merging to queue warning, from variable speed limits to automated enforcement, from real-time traveler information to incident management systems, the organic integration of nine core technologies constitutes complete technical solutions [39]. This integrated design philosophy provides systematic technical support for complex traffic management requirements during urban river-crossing corridor construction.

The balance between system architecture stability and intelligence represents key challenges in technical implementation. The application of maximum pressure controllers in signalized arterial networks validates the effectiveness of decentralized control methods [40]. This method models arterial traffic networks as queuing systems through extended versions of store-and-forward models, stabilizing all queues in the system under specific conditions. This decentralized architecture design provides stability assurance for large-scale traffic management systems.

The systematic enhancement of data processing capabilities represents the technical core of platform architecture. Traffic parameter extraction methods based on vehicle trajectory data achieve complete transformation from raw data to decision support through collaborative operation of five modules [41]. Innovative trajectory mapping algorithms effectively solves the technical challenges of mapping GPS points to road network nodes, multi-source data fusion enhances parameter estimation accuracy, and highly automated processing tool chains reduce system operation and maintenance costs. This modular data processing architecture not only ensures system scalability but more importantly provides technical assurance for precise state perception in complex traffic environments. In areas lacking traditional detection equipment, this technical solution provides feasible alternative solutions, which holds important practical value for temporary monitoring requirements during urban river-crossing corridor construction.

Based on the aforementioned technological innovation achievements, this research constructs an integrated platform architecture for traffic reorganization management during urban river-crossing corridor construction, as shown in Figure 1. This platform architecture adopts layered design philosophy, containing four core levels from bottom to top: sensing access layer, data processing layer, intelligent analysis layer, and application service layer, achieving a complete technical chain from data collection to intelligent decision-making.

The innovation of this platform architecture lies in achieving organic unification of technical integration and application requirements: through comprehensive data collection at the sensing access layer, addressing the problem of single data sources in traditional traffic management; through unified fusion platforms at the data processing layer, overcoming technical challenges in multi-source heterogeneous data processing; through dual-engine architecture at the intelligent analysis layer, establishing intelligent conversion mechanisms from data to decisions; through specialized services at the application service layer, meeting diverse management requirements during urban river-crossing corridor construction. The entire architecture embodies the technical philosophy of transformation from empirical judgment to data-driven approaches, providing complete technical solutions for modernizing traffic management during urban critical infrastructure construction.

5.2. Cross-Departmental Collaborative Traffic Management Decision Support System

The technical implementation of cross-departmental collaborative decision-making requires balancing professionalism and coordination. Knowledge-based decision support architectures provide intelligent decision support for traffic management centers through distributed blackboard architecture integration of multiple real-time expert systems [42]. This architectural design is particularly suitable for addressing non-recurrent congestion problems in large or complex networks, with distributed characteristics ensuring independence of various professional domains while blackboard mechanisms achieve effective cross-domain information sharing.

The challenges of massive real-time data processing faced by modern traffic management systems have catalyzed innovative applications of event-driven architectures. Event-driven architectures can achieve efficient processing of continuous event streams compared to traditional polling-based processing methods [43]. This architectural mode holds particularly prominent application value during urban river-crossing corridor construction: frequent traffic state changes and higher probability of sudden events make traditional processing methods difficult to meet real-time requirements.

The integrated application of deep learning technology in decision support further enhances system intelligence levels. Integrated frameworks combine data denoising and deep learning models, suppressing data outliers through techniques such as empirical mode decomposition while employing long short-term memory networks to complete prediction tasks [44]. The excellent performance of LSTM + EEMD solutions in prediction accuracy provides technical foundations for constructing high-precision decision support systems.

5.3. Platform Key Technologies and Operational Assurance Mechanisms

Innovation in platform operational assurance is reflected not only at the technical level but more importantly in social considerations. The application of social media data in public sentiment analysis for roadway work zones has opened new research domains [45]. Through collection and analysis of social media data, researchers can real-time grasp public opinions about work zones and identify key factors affecting work zone experiences. This social perception capability provides important public opinion feedback mechanisms for traffic management departments.

The application of machine learning methods in social media data classification makes automated analysis of large-scale public sentiment possible. This technical capability holds important value for public relations management during urban river-crossing corridor construction: through real-time monitoring of public sentiment changes, management departments can timely discover potential problems, adjust management strategies, and improve public satisfaction.

Travel time prediction technology for important corridors in urban agglomerations provides important support for platform core algorithms. The integrated solution of genetic algorithm-based segmented travel time calculation and long short-term memory networks embodies the development trend of combining optimization algorithms with deep learning technologies [46]. The advantages of genetic algorithms in global optimization compensate for neural network deficiencies, while the pattern recognition capabilities of neural networks enhance the learning effects of optimization algorithms.

This algorithmic fusion technical pathway provides effective solutions for complex prediction problems during urban river-crossing corridor construction. Through refined road segment division and accurate capture of spatiotemporal dependency relationships, systems can maintain high prediction accuracy in highly uncertain environments.

Systematic methods for socioeconomic impact assessment of transportation infrastructure projects provide important evaluation frameworks for platform operational assurance. Through systematic examination of assessment theoretical foundations, influencing factors, and analysis methods, particularly methodological innovations in data analysis and socioeconomic environmental impact assessment, scientific tools are provided for comprehensive quantification of construction project benefits [47].

Potential economic benefits identified by the evaluation framework include multiple dimensions such as travel time reduction, road capacity increases, and enhanced regional investment attractiveness. This comprehensive evaluation method not only provides objective foundations for project decision-making but more importantly provides effect feedback mechanisms for continuously optimizing management strategies.

The integrated platform architecture and system implementation constructed in this chapter represents key achievements in the transformation of data-driven traffic management methods from theoretical research to engineering applications. Multi-technology fusion system architectures achieve transformation from dispersed functions to integrated services, cross-departmental collaboration mechanisms embody upgrades from single management to comprehensive governance, and the combination of key technologies with assurance mechanisms ensures reliable transitions from concept verification to scaled applications.

The technological innovations of platform implementation are primarily manifested in four aspects: architectural innovation achieves organic integration of multi-level technologies, algorithmic innovation enhances prediction accuracy in complex environments, mechanism innovation ensures efficient operation of cross-departmental collaboration, and evaluation innovation establishes comprehensive effect feedback systems. These innovative achievements not only address technical challenges in traffic management during urban river-crossing corridor construction but also provide complete technical solutions for constructing modernized intelligent traffic management systems.

Within the overall thematic framework of this paper, this chapter assumes key functions of technical integration and application implementation: serving as both an integrated manifestation of the aforementioned theoretical methods and technical strategies and as technical assurance for the complete application of data-driven methodologies in urban traffic management. Through systematic design of platform architecture and engineering implementation of key technologies, the fundamental transformation from empirical judgment to data-driven management is truly completed, providing replicable and promotable technical pathways and practical experience for intelligent development of modern urban traffic management.

6. Application Practice, Challenges, and Development Trends

The preceding chapters have elucidated the technical architecture and methodological framework of data-driven approaches in traffic reorganization and management during urban river-crossing corridor construction from theoretical perspectives. However, gaps often exist between theory and practice, with the actual application effectiveness of technical solutions, challenges encountered, and future development trends directly determining whether data-driven methods can truly replace traditional empirical judgment approaches. This chapter systematically analyzes the application status, technical challenges, and development prospects of data-driven traffic management methods from engineering practice perspectives, providing practical guidance for advancing in-depth development in this field.

6.1. Analysis of Typical Engineering Application Cases

The actual application effectiveness of data-driven traffic management methods serves as an important criterion for testing their feasibility and validity. In traffic management practices during river-crossing corridor construction both domestically and internationally, a series of representative engineering cases have emerged, providing crucial support for evaluating the actual effectiveness of data-driven methods.

The I-495 bridge repair project in Delaware represents a typical case of traffic management during bridge construction. Withers [48] conducted an in-depth case study of this project, analyzing multimodal and multi-attribute tradeoff decisions during the recovery process following the emergency closure of the I-495 bridge in 2014. The outstanding characteristics of this case lie in the rapid emergency response and scientific nature of the decision-making process, demonstrating the important role of data-driven decisions in incident management through detailed examination of changes in user travel patterns, institutional planning to minimize disruptions, and innovative construction methods to accelerate bridge service restoration. This case fully illustrates the scientific decision-making value of data-driven methods when addressing complex engineering problems, providing important reference for implementing similar projects.

The bridge construction project in Sejong City, Republic of Korea, exemplifies the application value of agent-based modeling in traffic dispersion effect analysis. Yun et al. [49] developed traffic simulation models describing individual movement behaviors of entire urban populations based on agent-based urban management models, generating urban traffic system demands through aggregating individual-level movement behaviors and conducting statistical validation against real data. The innovation of this approach lies in its capability to capture the diversity and complexity of individual behaviors, providing new perspectives for understanding traffic flow redistribution patterns during construction periods. Compared to traditional macroscopic traffic analysis methods, agent-based modeling approaches better reflect the heterogeneous characteristics and decision-making behaviors of traffic participants, providing support for formulating more precise traffic management strategies.

European explorations in construction traffic management and planning decision support hold equally important reference value. Brusselaers et al. [50] studied methods for addressing spatiotemporal impacts of construction transport on urban traffic networks through traditional traffic and transport simulation, with research results emphasizing the important role of rigorous construction transport planning in avoiding peak traffic hours and significantly alleviating traffic congestion. This research particularly noted the need to comprehensively consider all simultaneously ongoing construction project demands when evaluating city-level disruptions, a finding that holds important guidance significance for cities with multiple parallel construction projects. These practical cases indicate that successful application of data-driven methods requires not only advanced technical means but also systematic planning thinking and coordinated management mechanisms.

The case studies reveal three key advantages of data-driven methods. Decision-making becomes more evidence-based, relying on quantitative analysis rather than intuition. Response times improve significantly, allowing faster adaptation to changing conditions. Finally, management becomes more targeted, with solutions tailored to specific contexts.

To provide systematic comparison of data-driven methods’ effectiveness, we synthesized quantitative performance metrics from studies reporting comparable outcomes. Traffic management system implementations demonstrate substantial improvements across multiple performance dimensions. Signal control optimization systems show particularly promising results, with the MARLIN-ATSC system achieving 27–39% reduction in average intersection delay and 15–26% travel time savings along major urban routes. Similarly, comprehensive decision-making systems for large-scale construction demonstrate practical benefits, with reported reductions in average peak-hour commuting delays from 6.57 min to 5.82 min, representing an 11.42% improvement.

Traffic prediction accuracy represents another critical performance dimension where data-driven methods show clear advantages. Advanced machine learning approaches consistently outperform traditional methods, with hybrid models achieving root mean square errors, mean absolute errors, and mean absolute percentage errors of 0.79, 0.60, and 2.14, respectively. Deep learning applications in traffic flow prediction demonstrate superior performance particularly in handling complex spatiotemporal dependencies, with LSTM-DNN models achieving optimal MAPE values across prediction scenarios with 30 min sliding time windows.

Technology penetration effects reveal important thresholds for system effectiveness. Connected vehicle applications in work zone management achieve significant delay reductions exceeding 13% when penetration rates reach 80% or higher, indicating clear requirements for widespread adoption to realize full benefits. These quantitative findings consistently demonstrate that data-driven approaches provide substantial performance improvements over traditional empirical methods, though specific benefits vary considerably based on implementation context, technology maturity, and local conditions.

6.2. Implementation Barriers and Critical Challenges

6.2.1. Critical Analysis of Implementation Barriers and Limitations

While data-driven methods demonstrate significant potential advantages, their practical implementation faces substantial challenges that must be critically examined. These barriers span technical, economic, and organizational dimensions, often determining the success or failure of intelligent traffic management deployments.

• Technical Implementation Challenges

Data quality represents the most fundamental barrier to effective data-driven traffic management. Zambrano-Martinez et al. [51] discovered through experimental analysis of realistic vehicular traffic trajectories in Valencia, Spain, that “only some street segments fall under the general theory of vehicular flow theories and can be well-fitted using quadratic regression, while numerous street segments fall under other categories.” This finding reveals the heterogeneous characteristics of urban traffic flows, indicating that traditional traffic flow theories have significant applicability limitations in complex urban environments. The complexity of multi-source heterogeneous data integration creates additional technical challenges, with significant differences in temporal synchronization, spatial alignment, and format standardization that complicate data fusion and analysis processes.

System integration complexity represents another critical technical barrier. The deployment of comprehensive data-driven traffic management systems requires integration across multiple technological platforms, sensor networks, communication systems, and legacy infrastructure. Many implementations encounter significant difficulties in achieving seamless operation across these diverse technological components, often resulting in performance degradation or system failures that are not adequately documented in published literature.

2.

Economic and Organizational Constraints

Economic evaluation and cost–benefit analysis present substantial implementation difficulties that are often underestimated in theoretical studies. Zhang et al. [52] emphasized that economic analysis as core content of project feasibility studies directly affects decision-making quality through its accuracy, yet urban river-crossing corridor construction projects involve numerous economic impact factors including direct construction costs, indirect social costs, and long-term economic benefits. The quantitative assessment of these factors presents significant practical difficulties, with different stakeholders possessing varying perceptions and evaluation standards for cost–benefit considerations.

Regional variability and technological adaptability issues compound implementation challenges. Liu et al. [53] studied the spatial impacts of trans-Yangtze highway fixed links and demonstrated that different regions exhibit significant variations in traffic characteristics, travel habits, and network structures, necessitating substantial customization of technical solutions. These regional adaptation requirements not only increase technical development costs but also impose higher demands on the professional capabilities of technical personnel and organizational change management processes.

3.

Implementation Failure Patterns and Risk Factors

The available literature reveals several recurring patterns that contribute to implementation difficulties, though comprehensive failure documentation remains limited due to publication bias toward successful deployments. Technical risks include algorithm performance degradation in real-world conditions compared to controlled testing environments, sensor network reliability issues under adverse weather conditions, and communication system vulnerabilities that can compromise data integrity. Economic risks encompass cost overruns during deployment phases, longer-than-expected payback periods, and maintenance costs that exceed initial projections. Organizational risks involve inter-departmental coordination difficulties, staff resistance to technological change, and inadequate training programs that limit system utilization effectiveness.

6.2.2. Technological Evolution Barriers and Future Implementation Challenges

Despite the tremendous potential demonstrated by data-driven methods at both theoretical and practical levels, numerous technical challenges and implementation barriers persist in actual applications. These challenges primarily concentrate on data quality control, system integration optimization, standardization framework establishment, and technological adaptability enhancement, requiring gradual resolution through continuous technological innovation and practical exploration.

The complexity of economic evaluation and cost–benefit analysis further intensifies implementation difficulties. Different stakeholders possess varying perceptions and evaluation standards for cost–benefit considerations, increasing the complexity of technical solution selection and implementation. These regional adaptability requirements not only increase technical development costs but also impose higher demands on the professional capabilities of technical personnel.

The existence of these technical challenges and implementation barriers indicates that while data-driven methods possess obvious technological advantages, their widespread application still requires resolution of a series of fundamental issues. Only through comprehensive measures including continuous technological innovation, standardization framework improvement, and talent development can the complete transformation from empirical judgment to data-driven methods be truly achieved.

6.3. Future Development Trends and Technological Evolution Directions

Based on analysis of current technological development status and practical requirements, data-driven methods for traffic reorganization and management during urban river-crossing corridor construction are advancing toward more intelligent, integrated, and refined directions. Future technological evolution will focus on system performance enhancement, application domain expansion, and standardization framework improvement.

The continuous improvement of intelligence levels represents the primary trend for future development. The application scope of work zone intelligent transportation systems will further expand, with technology maturity and practicality significantly enhanced. Livingston [54] demonstrated remarkable effectiveness in reducing congestion through traffic management contract incentive measures in a case study of work zone intelligent transportation systems for Arizona State Route 68 reconstruction project. This case validated the application value and development prospects of intelligent transportation systems in complex work zone environments through detailed personnel interviews and field investigations. In the future, similar intelligent technologies will be applied in more river-crossing corridor construction projects, forming more mature and comprehensive technical systems through continuous technological optimization and experience accumulation.

The deep application of digital twin technology will become an important direction for technological evolution. The bridge group digital twin system based on machine vision fusion monitoring proposed by Dan et al. [55] achieves interconnection of bridge groups in regional transportation infrastructure networks through measuring traffic loads. This system establishes comprehensive bridge traffic load monitoring systems based on Weigh-in-Motion (WIM) and multi-source heterogeneous machine vision information fusion in physical space while employing lightweight sensors for structural response information collection. This technical architecture provides important references for constructing more complete and precise digital twin systems, demonstrating the development trend of digital twin technology transformation from concept verification to engineering applications.

Technological innovations in bridge inspection and monitoring fields are equally noteworthy. Hagen and Andersen [56] demonstrated how to combine physics-based methods with machine learning technologies to facilitate damage detection and diagnosis in their research on reinforced concrete bridge damage detection and virtual inspection applications. This multi-technology fusion approach not only enhances detection accuracy but also provides technical support for automated detection and intelligent maintenance. Comprehensive enhancement of system integration capabilities will advance the breadth and depth of technological applications. Gao et al. [57] developed bridge digital twin systems for actual bridge operation and maintenance through integrating GIS and BIM technologies. This system adopts three-layer architecture, combining WebGIS, WebBIM, and graph algorithms to establish Common Data Environments (CDE) for addressing cross-platform compatibility issues, achieving multiple functions including real-time monitoring, drone inspection, maintenance planning, traffic diversion, and logistics optimization.

The deep integration of artificial intelligence technologies will further enhance system intelligence levels. The AI agent-based Intelligent Urban Digital Twin (I-UDT) concept proposed by Choi and Yoon [58] achieves efficient integration of distributed urban-scale data and building feature extraction through integrating GPT technology within urban digital twins. This AI-driven technological mode points the direction for intelligent development of urban traffic management, indicating that artificial intelligence will play increasingly important roles in future traffic management.

The comprehensive development of intelligent mobility will drive traffic management evolution toward higher levels. Goumiri et al. [59] studied smart mobility in intelligent cities, analyzing emerging challenges, latest advances, and future directions. This research provided an overview of intelligent mobility, discussing major challenges related to its key building blocks, including parking and traffic management, traffic routing, and emission and road safety impacts. This comprehensive development trend indicates that future traffic management will no longer be limited to single technologies or single scenarios but will develop toward systematized and integrated directions.

The deep application of ETC data will provide new technical pathways for urban traffic management. Wang et al. [60] conducted a comprehensive review of urban traffic state sensing and analysis based on ETC data, pointing out challenges of insufficient sensing capabilities and inadequate operational state assessment in urban traffic management. The rapid expansion of ETC systems from highways to urban roads provides new opportunities for addressing these problems, with vast amounts of “dormant” ETC data containing rich traffic information urgently requiring deep mining and effective utilization. This finding provides new data sources and technical means for data-driven traffic management.

The integrated application of emerging technologies will open broader development spaces. Mahomed and Saha [61] studied unleashing the potential of 5G smart cities, focusing on real-time digital twin integration. The advent of 5G technology is transforming smart city creation by providing unparalleled speed, extremely low latency, and extensive device connectivity. These developments enable seamless integration of Internet of Things devices, real-time monitoring systems, and advanced urban applications, providing powerful technical support for traffic management systems to develop toward more real-time and precise directions.

Table 2. Comparative Analysis of Typical River-Crossing Corridor Construction Traffic Management Cases at Home and Abroad.

Case Name	Country	Project Characteristics	Main Technical Methods	Key Innovation Points	Application Effects	Literature Reference
Cross-Hangzhou Bay Channel Prediction System	China	Travel time prediction for important corridors in urban agglomerations	GA-LSTM neural network integration framework	Genetic algorithm optimized road segment division, LSTM captures spatiotemporal dependencies	Significantly improved prediction accuracy compared to traditional methods	[46]
Bolshoy Smolensky Bridge Construction Project	Russia	Socioeconomic impact assessment of urban transportation infrastructure	Comprehensive impact assessment methodology system	Multi-dimensional socioeconomic benefit quantitative analysis	Identified benefits including reduced travel time and enhanced regional investment attractiveness	[47]
I-495 Bridge Emergency Repair Project	United States	59-day emergency closure for repair, crossing Christina River	Multi-modal multi-attribute tradeoff decision analysis	Rapid emergency response mechanisms, innovative construction methods for accelerated recovery	Temporary measures effectively alleviated congestion, bus and rail ridership increased significantly	[48]
Sejong City Bridge Construction Project	Republic of Korea	Traffic dispersion analysis for new city bridge construction	Agent-based urban management model	Individual-level movement behavior modeling, representative simulation of entire city population	Model achieved statistical validation consistency with real data, providing support for city-level management	[49]
Construction Traffic Management Optimization Project	Multiple European Countries	Coordinated management of multiple parallel construction projects	Integration of traditional traffic simulation and transportation planning	Rigorous construction transport planning, peak-avoidance construction strategies	Significantly alleviated traffic congestion through peak time avoidance	[50]
Interstate 68 Highway Reconstruction Project	United States	Intelligent transportation systems work zone application	Traffic management contract incentive mechanisms	Deep integration of contract incentives with intelligent transportation systems	Work zone congestion significantly reduced, demonstrating ITS application prospects	[54]

demonstrates the application development trajectory of data-driven traffic management methods, with successful practices of these typical cases providing important references for traffic management during urban river-crossing corridor construction: valuable experiences and practices exist in emergency response capability development, advanced modeling technology applications, multi-project coordination mechanism construction, and intelligent system integration.

The analysis of international case studies reveals important patterns regarding both successes and limitations of data-driven traffic management implementations. While the documented cases demonstrate significant achievements, several common challenges emerge across different contexts. The I-495 bridge emergency repair project [48] successfully managed a 59-day closure through innovative multimodal approaches yet highlighted the substantial coordination challenges and resource requirements for emergency response. Similarly, the Sejong City bridge construction project [49] demonstrated effective agent-based modeling applications but also revealed the complexity of scaling individual behavior models to city-level management systems.

Economic factors frequently constrain the scope of implementation, with organizations often scaling back initial ambitious plans due to budget limitations or uncertain return on investment timelines. The European construction traffic management studies [50] emphasized how rigorous planning can significantly alleviate congestion but also noted the substantial coordination requirements among multiple simultaneous construction projects that strain available resources and management capabilities.

Organizational resistance to technological change represents a recurring theme across multiple case studies, though specific failure rates are not systematically documented in the available literature. The success of implementations appears strongly correlated with institutional readiness, staff training programs, and cross-departmental coordination mechanisms, suggesting that technological capabilities alone are insufficient to ensure successful deployment of data-driven traffic management systems.

6.4. Comprehensive Analysis of Data-Driven Traffic Management Technologies

To better understand the application characteristics and effects of different data-driven technologies in traffic management during urban river-crossing corridor construction, this section conducts comprehensive analysis of major technology development trends and application prospects based on the aforementioned research findings.

The application of digital twin technology in transportation infrastructure management will continue to deepen. Wu et al. [62] conducted a comprehensive review of digital twin technology in transportation infrastructure, covering current applications, challenges, and future directions. This review first summarized fundamental concepts and architectures involved in transportation infrastructure digital twin systems, then categorized transportation infrastructure digital twin systems from lifecycle perspectives, providing systematic frameworks for understanding application patterns and development trends of digital twin technology. Research indicates that digital twin technology is transforming from concept verification stages to practical application stages, with its value in enhancing management precision, optimizing decision quality, and reducing operation and maintenance costs becoming increasingly prominent.

The innovative development of bridge inspection and monitoring technologies will provide stronger support for river-crossing corridor management. Gkoumas et al. [63] studied new technologies for bridge inspection and monitoring from European Union research and innovation project perspectives, pointing out that European transportation infrastructure including bridges requires optimized maintenance plans and appropriate monitoring throughout their lifecycles to ensure safety and serviceability. Compared to existing and established inspection and monitoring methods, deployment of new technologies can help achieve these goals while providing numerous advantages. These technological innovations not only enhance detection accuracy and efficiency but also provide technical foundations for preventive maintenance and intelligent management, holding important significance for ensuring traffic safety during river-crossing corridor construction.

The application of big data technology in intelligent transportation systems will further expand. Ahmad Jan et al. [64] studied the significance of big data in intelligent transportation systems, analyzing current trends, challenges, and future directions. Intelligent transportation systems generate massive amounts of big data through sensing and non-sensing platforms, with this data supporting both batch processing and stream processing that are crucial for reliable road operations and connected vehicles. Research indicates that big data technology possesses enormous potential in traffic state monitoring, travel behavior analysis, and congestion prediction, while simultaneously facing challenges including data quality control, privacy protection, and computational efficiency.

System reliability-oriented optimization strategies will become important development directions. Chen et al. [65] studied optimized bridge maintenance strategies, employing system reliability-based approaches to enhance road network performance. This research proposed innovative models based on system reliability principles, developing optimized bridge maintenance strategies aimed at improving overall reliability of bridge-dominated road networks through considering quantified impacts of bridge component deterioration. This systematic optimization approach provides new theoretical guidance and methodological support for traffic management during river-crossing corridor construction.

The integrated application of Internet of Things and BIM technologies will drive precise development of traffic load prediction. Lou et al. [66] studied traffic load prediction for bridge construction based on IoT and BIM. With the development of urban transportation infrastructure, bridge construction often leads to traffic congestion and safety hazards. Traditional traffic load prediction cannot address dynamic changes in traffic during construction periods. This research proposed traffic load prediction and dynamic optimization methods based on Building Information Modeling (BIM) and Internet of Things (IoT) integration, providing new technical pathways for addressing traffic management problems in complex construction environments.

Traffic dynamic modeling technology will play an important role in specific scenario applications. Nagatani [67] studied dynamic models for traffic concentration and congestion near bridges, investigating traffic flows and congestion near bridges when many vehicles flow into roads connecting bridges during morning peak hours. This research proposed mathematical dynamic models to simulate traffic concentration and congestion near bridges, with dynamic models described by differential equations on decorated one-dimensional lattices with multiple inflows and outflows. This refined modeling approach for specific traffic scenarios provides valuable technical tools for traffic analysis during river-crossing corridor construction.

The application of deep learning in traffic flow analysis will continue to deepen. Pamuła and Żochowska [68] studied traffic flow-based OD matrix estimation and prediction in uncongested urban road networks based on deep learning. This research proposed new methods for OD matrix prediction based on traffic data using deep learning. Input values for developed models were determined based on road network structural data, trip origins and destinations, and traffic intensity data for road network segments recorded by video sensing devices. The application of this deep learning approach provides more precise and intelligent technical means for traffic flow analysis and prediction. Future technical systems will pay more attention to synergistic effects among different technologies, addressing complex traffic management problems through multi-technology integration. Simultaneously, technological applications will emphasize practicality and economic efficiency while ensuring technological advancement, enhancing engineering operability and investment benefits.

7. Conclusions

7.1. Primary Research Conclusions

This research conducted comprehensive examination of the complete transformation process from empirical judgment to data-driven approaches in traffic reorganization and management during urban river-crossing corridor construction. Research results indicate that this transformation represents not merely an update in technical means but a profound revolution in management philosophy and decision-making modes, possessing important theoretical value and practical significance.

The technical advantages of data-driven methods have been fully validated. Based on analysis from the preceding chapters, data-driven methods demonstrate obvious advantages in traffic prediction accuracy, management decision-making scientific rigor, and system operational efficiency. Particularly, traffic impact assessment frameworks based on Bridge Information Modeling and multi-source data fusion demonstrate substantial improvements in prediction reliability, marking a fundamental shift from reactive to predictive traffic management paradigms. Similarly, intelligent transportation systems show significant operational efficiency gains in work zone management, validating the transformation from experience-based to evidence-based decision-making processes. These achievements collectively indicate that data-driven methods have matured from experimental concepts to practical implementation tools, establishing new benchmarks for scientific rigor in urban traffic management.

The systematic nature and completeness of technical systems constitute core characteristics of data-driven methods. Through analysis of different technology categories, research found that data-driven methods represent not single technology applications but formation of complete technical chains encompassing “data sensing—intelligent analysis—predictive warning—decision support.” This technical architecture covers the entire process from data acquisition to decision implementation, achieving systematization and scientification of traffic management. Particularly in key technical components including multi-source data fusion, artificial intelligence algorithms, and digital twin simulation, important breakthroughs have been achieved, providing solid foundations for constructing complete technical systems.

The diversity and effectiveness of application practices validate the practical value of data-driven methods. From emergency response in the Delaware I-495 bridge repair project to agent-based modeling traffic dispersion analysis in Sejong City, Republic of Korea, and systematic planning of construction traffic management in Europe, different types of engineering practices have all demonstrated the effectiveness of data-driven methods. These successful cases not only validate method feasibility at technical levels but also demonstrate broad prospects for widespread application at management mode levels.

7.2. Theoretical Contributions and Application Value

The theoretical contributions of this research are primarily manifested in constructing complete technological evolution analysis frameworks and establishing systematic method classification systems, providing important support for the development of data-driven traffic management theory.

Regarding technological evolution analysis, while previous surveys have made valuable contributions by examining individual technological advances or specific applications, this research contributes to the literature by constructing a more comprehensive framework that traces the complete paradigm transformation from empirical judgment to data-driven approaches. Building upon existing reviews that often treat technological development as sequential improvements, our analysis reveals the complex, interconnected nature of this transformation, identifying how sociotechnical factors, institutional barriers, and implementation contexts shape technological adoption. This approach advances knowledge by demonstrating that successful paradigm shifts require understanding not only technological capabilities, but also the organizational and contextual factors that enable or constrain transformation.

Regarding method classification systems, while existing literature has provided valuable categorizations based on technical characteristics, this survey contributes by developing an integrated classification framework that better bridges technical capabilities with implementation requirements and contextual constraints. Rather than simply cataloging technologies, our framework helps reveal the interdependencies between different approaches and establishes decision pathways for technology selection based on project characteristics, resource availability, and organizational readiness. This represents a contribution toward more prescriptive knowledge that can better support practitioners in making informed implementation decisions, complementing the descriptive approaches found in existing technology surveys.

The practical guidance significance of this research provides concrete decision-making frameworks for different professional stakeholders. For policymakers, our analysis offers standardization roadmaps for data-driven traffic system deployment, including regulatory framework development and inter-agency coordination protocols. Urban planners can utilize our classification system to select appropriate technologies based on project characteristics: large-scale projects with mature data infrastructure should prioritize advanced AI and digital twin technologies, while medium-scale projects with limited resources can achieve significant benefits through intelligent transportation systems and traffic simulation approaches. For engineers, our implementation pathway analysis provides step-by-step guidance for technology integration, risk assessment frameworks, and performance evaluation metrics. Additionally, our budget-consideration framework helps organizations determine optimal technology combinations based on available resources, with cost-effective entry points for organizations with limited initial investment capacity and scalable upgrade pathways for long-term development.

7.3. Overall Conclusions

Based on the analysis results of this research and current technology development trends, future development of data-driven traffic management methods will exhibit overall characteristics of technological integration, application intelligence, and standardization unification, encountering important development opportunities across multiple aspects.

Data-driven methods demonstrate obvious advantages in technological advancement, management scientific rigor, and application practicality, providing new technical pathways and methodological means for addressing complex urban traffic management problems. Although current challenges persist in aspects including data quality, system integration, and standardization, these problems will be gradually resolved as related technologies continue improving and engineering practices accumulate. Future technological development will advance toward integration, intelligence, and standardization directions, providing support for constructing more comprehensive traffic management systems.

From broader perspectives, the development of data-driven traffic management methods represents not only technological progress but also embodies innovation in urban governance philosophy and transformation of management modes. Through scientific, precise, and intelligent management approaches, urban development needs and citizen travel requirements can be better served, contributing important strength to building smart cities and sustainable development cities. This technological development trend will have profound impacts on future urban traffic management and even entire urban governance systems, possessing important theoretical significance and application value.

8. Discussion

8.1. Emerging Technology Integration and System Evolution

Technology integration will become the primary trend for future development. With the rapid development of emerging technologies such as 5G communications, edge computing, and Internet of Things, data-driven traffic management will increasingly adopt multi-technology integration solutions. This integration is manifested not only in mutual complementation at technical levels but also in collaborative optimization at system levels. Future traffic management systems will pay more attention to organic combination among different technologies, addressing complex problems that cannot be handled by single technologies through technological integration. Specifically, digital twins and AI integration will fundamentally transform construction-phase traffic management through three key mechanisms. First, real-time digital replicas of construction zones will enable predictive scenario modeling, allowing traffic managers to simulate and optimize multiple traffic management strategies before implementation. Second, AI-powered adaptive control systems will automatically adjust traffic signals, variable message signs, and route guidance based on real-time construction progress and traffic conditions, eliminating the lag time inherent in traditional manual adjustments. Third, integrated platforms will enable seamless coordination between construction scheduling systems and traffic management systems, optimizing both construction efficiency and traffic flow simultaneously.

Application intelligence will represent an important direction for technological development. Artificial intelligence technologies, particularly machine learning and deep learning applications in traffic management, will become more in-depth and widespread. Future systems will possess stronger self-learning capabilities, adaptive abilities, and autonomous decision-making capacities, enabling intelligent management and control in complex and variable traffic environments. Simultaneously, combined applications of artificial intelligence with other technologies will generate new technological breakthroughs and application modes.

Institutional reforms, particularly standardized data-sharing frameworks, will be essential for wider adoption of data-driven traffic management. Key reforms include establishing mandatory open data policies for transportation agencies, creating unified data exchange protocols that enable real-time information sharing across jurisdictions, and developing privacy-preserving data sharing mechanisms that balance transparency with security concerns. Additionally, regulatory frameworks must evolve to accommodate dynamic traffic management approaches, including flexible permitting processes for adaptive construction scheduling and updated liability frameworks for AI-driven traffic control decisions. These institutional changes will require coordination between transportation departments, technology vendors, and regulatory bodies to ensure seamless implementation.

Sustainability goals, particularly carbon neutrality and resilience objectives, will fundamentally reshape next-generation traffic systems through multiple pathways. Carbon neutrality targets will drive the integration of emission optimization algorithms into traffic management systems, enabling real-time adjustment of traffic flows to minimize vehicular emissions while maintaining operational efficiency. Smart traffic systems will incorporate electric vehicle charging infrastructure coordination, optimizing both traffic flow and energy grid stability. Resilience considerations will mandate the development of adaptive traffic management systems capable of rapid response to climate-related disruptions, including flood-resistant sensor networks and emergency evacuation optimization algorithms. Furthermore, circular economy principles will influence system design, promoting the reuse of traffic data for urban planning purposes and the development of modular, upgradeable infrastructure that minimizes waste throughout its lifecycle.

8.2. Research Limitations and Future Research Directions

While this survey provides comprehensive analysis of data-driven approaches in traffic management during urban river-crossing corridor construction, several limitations and future research opportunities should be acknowledged.

The geographic and cultural diversity of case studies, though spanning multiple countries, may not fully represent the complete spectrum of global practices. Future research could benefit from including more cases from developing countries and regions with different transportation infrastructure characteristics. Additionally, the long-term performance evaluation of implemented data-driven systems requires further investigation, as most case studies focus on short-term implementation effects rather than sustained operational outcomes.

The rapid evolution of emerging technologies such as artificial intelligence, 5G communication, and autonomous vehicles presents both opportunities and challenges for future research. The integration of these technologies with existing traffic management systems requires comprehensive investigation of technical compatibility, implementation costs, and societal acceptance. Furthermore, privacy and cybersecurity concerns associated with data-driven traffic management systems warrant dedicated research attention.

Future research should also focus on developing standardized evaluation frameworks for comparing different data-driven approaches across various contexts. The establishment of common performance metrics and assessment methodologies would facilitate better understanding of technology effectiveness and support evidence-based decision-making in technology adoption.

The socioeconomic implications of paradigm shifts in traffic management deserve deeper investigation. Research examining the impacts on employment, skill requirements, and organizational structures within transportation agencies would provide valuable insights for policy development and implementation planning. Additionally, the environmental and sustainability aspects of data-driven traffic management systems require more comprehensive life-cycle assessments to support sustainable urban development goals.