Deployment of an intelligent transportation system (ITS) for traffic flow optimization and decision support

Enhanced traffic management systems (e.g., signal control, event monitoring, emergency rescue, and electronic toll collection, dynamic navigation, parking information services, etc.)

Geographic Information System (GIS)-based decision support system for data management and analysis

Integration of advanced information technology, data communication, electronic sensors, control systems, and computing technologies for real-time traffic management

Use of panel data for statistical analyses to evaluate transportation system performance

Inequitable access to essential services for rural populations, including healthcare and groceries

Rising operational costs for demand-responsive transport systems

Scarce resources and difficulty balancing multiple objectives, such as cost, coverage, and equity

GIS-based decision support system for data management and analysis

Redesigned bus routes optimized for equity and cost

Applied a heuristic procedure for route optimization balancing cost and equity

GIS for mapping population density and major points of interest across rural regions

Data envelopment analysis (DEA) for multi-objective optimization and efficiency assessment

Multi-objective utility analysis (MOUA) and conjoint analysis for equity evaluation

Dijkstra algorithm for path optimization

Increasing vehicular density leading to congestion

Failure of classical approaches to ensure smooth and efficient traffic flow due to rising urban traffic demands

Adverse weather conditions affecting road traffic and safety

Ineffectiveness of conventional traffic management solutions in managing increasing vehicular density

Dynamic congestion prediction using artificial neural networks (ANNs)

Real-time adjustments to signal timings based on congestion levels

Weather-integrated traffic management to enhance road safety and reduce delays

Back Propagation Neural Networks (BPNN) for predicting traffic delays with input-output mapping

Levenberg–Marquardt algorithm for training multilayer perceptron networks with dataset division

Smart traffic control system architecture

Non-Linear Autoregressive with External Input (NARX) model for time-series traffic data analysis and simulation using MATLAB R2017a

IoT-based data acquisition from sensors deployed at intersections for real-time traffic monitoring and control

Traffic congestion caused by insufficient real-time data processing

Challenges in integrating heterogeneous IoT data sources for transportation

High computational demands for managing large-scale data

Real-time data analysis for traffic flow prediction and congestion management

Dynamic decision-making using real-time event detection and response systems

Development of a smart transportation system architecture using big data analytics for real-time processing

Utilization of Apache Hadoop for large-scale data storage and processing

MapReduce programming for parallel processing mechanism to manage large datasets efficiently

IoT integration using traffic sensors, cameras, and GPS data aggregation for real-time monitoring

Noise reduction in data using statistical methods like Kalman filters

Need for integration of real-time data to reduce uncertainty and improve response

Severe traffic congestion due to rapid urbanization and economic growth

Inefficiency in traditional public transport scheduling due to static timetables

Passenger dissatisfaction caused by long waiting times and irregular bus services

Unpredictable traffic conditions impacting resource allocation and vehicle routing

Real-time passenger flow analysis and vehicle control optimization

Dynamic dispatching of buses based on real-time traffic and passenger data

Transport flow prediction using periodic pattern mining

Decision support system with evolutionary algorithms for scheduling optimization

IoT-based framework which includes perception, network, and application layers

Sensors for GPS tracking, automatic passenger counting, and real-time traffic monitoring

Communication protocols like GSM, 4G, Zigbee, and Wi-Fi

Dynamic scheduling and optimization applications

Cloud computing for data processing and analysis

Optimization algorithms for vehicle dispatching

Meta-heuristic algorithms for complex scheduling problems

Real-time handling of dynamic and volatile traffic data streams

Lack of real-time machine learning algorithms to predict non-recurrent traffic incidents across entire road networks

Predicting traffic flow under diverse, evolving conditions

Deep learning approach for real-time traffic flow prediction and impact propagation estimation across arterial road networks

Deep reinforcement learning model to optimize traffic control actions using real-time data

Social media data integration model to capture commuter sentiment and emotions during non-recurrent traffic events

Bluetooth traffic monitoring system for trajectory data collection at arterial road junctions

Online learning techniques to manage high-volume and high-velocity data with data-driven processing triggers

Deep Neural Networks (DNNs) for forecasting traffic flow and identifying critical road segments

Deep Reinforcement Learning (DRL) for adaptive and intelligent traffic control systems

Long short-term memory (LSTM) architecture for capturing temporal traffic patterns

Sentiment analysis of social media data, extending traditional methods to extract deeper emotions

Integration of cloud computing, network control systems, and traffic management to provide technical support for traffic control

Edge computing to provide localized control for critical nodes in the ITS

Development of an intelligent transportation cyber-physical cloud control system for real-time data management and control

Traffic redistribution strategies to alleviate congestion in overburdened areas

Utilization of cloud computing for scalable traffic flow prediction, optimization, and scheduling

Cyber-physical systems (CPS) for acquisition, transmission, and computation of traffic data in real time for the optimization, decision making, scheduling, planning, prediction, and control of the system

Traffic big data analytics for advanced data mining techniques for accurate and efficient traffic data processing

Deep Belief Network-Support Vector Regression (DBN-SVR) for large datasets

Back Propagation Bilateral Extreme Learning Machine (BP-BELM) for small-scale predictions

Edge computing for real-time, localized control

Cloud systems for large-scale computation and data storage

Severe congestion at key locations and during peak times

Insufficient predictive tools for identifying long-term traffic patterns

Data imbalance affecting accurate predictions of extreme congestion levels

Predictive analytics for classifying jam severity and supporting real-time traffic management

Identification of traffic hotspots and peak times for infrastructure planning

Enhanced visualization for dynamic traffic analysis

Navigation app data captured at millisecond intervals, providing raw traffic datasets for analysis

Hadoop big data system for storing and processing large-scale traffic datasets using distributed parallel computing

Hive ecosystem with HiveQL for data schema creation, cleaning, summarization, and sample dataset generation

Multiclass Decision Forest for classification of jam levels

Interactive visuals to illustrate traffic jams, time-based trends, and segmented information for better analysis

High variability in freight travel time due to stochastic events (e.g., congestion, weather)

Lack of robust short-term predictive models for real-time logistics

Difficulties in processing large-scale, heterogeneous IoT data in real-time

Data-driven predictive analytics combining real-time traffic data with stochastic modeling to enhance travel time reliability

Adaptive logistics management by integrating predictive analytics into routing and scheduling

Enhanced response speed and operational efficiency in Logistics 4.0

Traffic IoT integrating big data sources for logistics and transportation system improvements

Traffic sensors, GPS, and vehicle detectors

5G-enabled Vehicle-to-Everything (V2X) for seamless connectivity and data exchange in transportation systems

Ensemble machine learning techniques, such as bagging and boosting, to enhance model performance by combining multiple predictors

Lack of integration of multimodal traffic data with situational context for improved mobility management

Insufficient context-aware analysis to capture situational dynamics (e.g., large-scale events, traffic restrictions, weather, and urban planning changes) on mobility patterns

Context-aware and multimodal traffic analysis principles for integrating urban data sources

Spatiotemporal and contextual modeling for emerging traffic trends and system optimization

Predictive traffic analysis using deep learning models such as recurrent neural networks and graph neural networks (GNNs) for short-term and long-term forecasting

Integration of automated fare collection systems to track cross-carrier passenger flows and infer comprehensive origin-destination matrices

Urban data fusion techniques to enhance data consistency and completeness

Machine learning and analytics for predictive, prescriptive, and diagnostic analysis of multimodal traffic data

Spatiotemporal pattern mining and relational data mining techniques for understanding user-specific mobility patterns

Deployment of reinforcement learning and deep neural networks for optimizing public transport operations

Scenario-based simulation approaches to support decision-making in traffic management

Traffic dashboards and GIS-based multimodal data representation

Processing and analyzing large volumes of smart city data for timely decision-making

Traffic congestion and crowded parking due to high vehicle density and limited space

Inefficiency in traffic light operations

Real-time traffic forecasting for flow optimization

Automated traffic management mechanism to prevent congestion and optimize urban mobility

Decision trees for fault detection and rapid decision-making

Parking occupancy prediction to assist in space allocation

Multi-agent system (MAS) architecture designed for urban traffic management processes

K-nearest neighbor (KNN) for occupancy rates and Random Tree for traffic flow

Monitoring agents equipped with expert systems using the Jess rule engine for anomaly detection

Implementation within the Java Agent Development Framework (JADE) for inter-agent communication

Data mining tool for forecasting, classification, data pre-processing, and visualization

Traffic congestion due to increased urbanization and vehicle density

Inefficient traffic management caused by the inability to process real-time data effectively

Delays in routing decisions for commuters and emergency services

Real-time traffic analysis to identify congested routes and redistribute traffic

Traffic flow analysis using periodic pattern mining

Smart routing for commuters and emergency services using weighted graphs

Data-driven decision-making for urban planners to design better traffic systems

CPS, which integrates IoT sensors and big data tools for real-time data collection and analysis

Graph algorithms that utilized Dijkstra’s algorithm and maximum spanning tree techniques for traffic route optimization

Big data tools using Apache Spark 2.3.1 GraphX and Hadoop 2.6.5 for big data processing

Adaptive green light scheduling based on real-time traffic inflow data

Smart navigation system to guide vehicles along time-optimized routes and evenly distribute traffic across available paths

Real-time situation-aware traffic management to enhance road safety and improve overall traffic flow

IoT for acquiring vehicle mobility data through onboard sensors, infrastructure-based sensors, and vehicle-to-infrastructure communication via Vehicular Ad-hoc Networks (VANETs)

Edge computing for low-latency, location-aware, and real-time processing of mobility data

Cloud computing for data storage and large-scale processing of traffic data

The reverse edge layer facilitates time-efficient smart navigation, ensuring the optimal distribution

Safety risks in mixed traffic environments with autonomous and manual vehicles

Poor real-time performance and low accuracy in intention recognition systems

Limited data processing capabilities in traditional ITS frameworks

5G, edge computing, and AI-based deep learning traffic safety solution for mixed traffic

Use of normalized driving trajectory and natural-driving datasets

Decision-level fusion of historical and natural-driving data for higher recognition accuracy

5G technology for data collection and edge processing without centralized servers

TensorFlow framework for building long short-term memory networks

Moving average algorithm for smoothing driving trajectory data

LSTM network to handle data dependence and time sequence issues

Edge computing to support decentralized data processing, reducing the load on centralized servers

Traffic congestion impacts air quality and urban livability

Lack of tools for integrating traffic and pollution data in real-time

Sensor data anomalies reduce reliability for traffic modeling

Trafair Traffic Dashboard to analyze and visualize real-time traffic sensor data and traffic flow simulations

Integration of traffic modeling with air quality impact analysis to provide actionable insights for city administrations

Scenario-based simulations to evaluate sustainable urban policies

Microscopic traffic simulations and route generation from sensor data

Data cleaning using speed-flow correlation filters and anomaly detection with Seasonal-Trend Decomposition using Loess and Interquartile Range analysis

PostgreSQL database with PostGIS and Timescale extensions for efficient spatial and temporal data management

Anomaly detection algorithms applied to sensor data

Balancing mobility with health restrictions during a pandemic

Computing safe and fast routes considering changing alert levels and curfews

Integrating diverse data sources for decision-making

Real-time computation of routes based on health data and mobility restrictions

Dynamic alert-level-based traffic control for urban areas

Modular architecture allowing easy adaptation to other emergencies

Complex Event Processing (CEP) to detect real-time events and health risks using tools like MEdit4CEP

Fuzzy Inference System implemented to facilitate traffic restriction decisions

Colored Petri Net (CPN) models to mapping for route simulations, integrating alert levels and time restrictions

Deployment of the Pandemic intelligent transportation system to support authorities in making mobility decisions and ensuring efficient routing for drivers

High variability and unpredictability of travel times due to traffic congestion, accidents, and roadwork

Inadequate performance of traditional predictive models under dynamic traffic conditions

Growing emphasis on green transportation necessitates efficient logistics and transportation planning

Bi-directional time processing for enhanced predictive accuracy

Real-time vehicle routing optimization using predicted travel times

Integration of IoT data for continuous updates and system improvement

Bi-Directional Isometric-Gated Recurrent Unit (BDIGRU) for chronological and retrospective information processing

IoT-based terminal technology for collecting vehicle operation and road condition data

Machine learning-based gradient descent optimization for real-time analytics

Integration of traffic data processing with enterprise resource management for cost-effective logistics optimization

Inability of traditional traffic simulations to adapt to real-time changes

Ineffective testing of control strategies like variable speed limits in offline models

Difficulty in developing control strategies that perform consistently across varying traffic states in dynamic and stochastic environments such as motorways

Development of a run-time synchronized digital twin model of the Geneva motorway using real-time traffic data from motorway counters

Safe and efficient testing of traffic management strategies using parallel Digital Twin Instances (DTIs)

Predictive analytics for early detection of traffic anomalies and optimization of traffic flows

Leveraging the open data and the microscopic traffic emulator to enhance simulation capabilities

Simulation of Urban Mobility (SUMO) for modeling and simulating synchronized digital replicas of real motorway traffic

Dynamic Flow Calibration mechanism for real-time adjustment of traffic flow distributions

Real-time data collection from motorway traffic counters aggregated every minute and accessed via the Open Data Platform Mobility Switzerland (ODPMS)

Use of a central repository for storing and accessing high-resolution traffic data for real-time simulations

Increased bus travel times due to mixed traffic conditions

Delays caused by general traffic, ridership levels, and weather factors

Inefficiencies in calculating passenger time savings with traditional methods

Exclusive bus lanes to mitigate traffic interference and reduce travel times

Utilizing sensor and management system data with high spatial and temporal detail to enhance analysis

Enhanced methodologies for calculating passenger time savings using big data

Alighting prediction algorithms to improve ridership data accuracy

Bluetooth sensors for traffic flow and travel times

Inductive loops for road occupancy and flow rate

Statistical methods tailored to fit specific case studies for improved accuracy

Managing heterogeneous IoT traffic from interconnected devices in transportation networks

Effective network monitoring and management for Internet of Vehicles (IoV) systems

Preventing traffic congestion by utilizing ITS data for proactive forecasting

Security threats such as intrusions and attacks on vehicular networks

Development of an intelligent intrusion detection system (IDS) for IoV-based vehicular networks using tree-based machine learning and ensemble learning techniques

Stacked ensemble learning approach for improved accuracy in intrusion detection

Feature selection to optimize network monitoring and reduce computational costs

Enhancing accuracy and attack detection capability

Application of machine learning-based IDS for IoV within vehicular ad hoc networks

Use of tree-based learning techniques such as Decision Tree, Random Forest, Extra Tree, and XGBoost for classification and training

Stacking methodology to enhance classifier robustness and reliability by combining multiple learners

Traffic flow prediction models for real-time signal optimization and dynamic routing

Ensemble learning methods to improve the accuracy of traffic flow predictions

Optimizing traffic signal timing and resource allocation through precise forecasting

Random Forest, Support Vector Regression (SVR), and ARIMA for traffic forecasting

Gradient boosting as ensemble learning techniques

Heat maps to visualize traffic flow distribution

Data-driven decision-making integrated into smart city systems

Low network efficiency due to underutilized stops and poorly connected routes

Limited integration between subway and bus systems in traditional network planning

Inefficiencies in passenger flow management during peak hours

Subway–bus double-layer network model to optimize public transportation routes and connections

Structural adjustments to add/remove stops based on demand and efficiency metrics

Simulation-driven decision-making to identify areas for network improvement

Passenger flow and travel time data derived from IC card swipes

Weighted edge connections based on passenger flow and travel time

L-space modeling for connectivity analysis and C-space modeling for transfer optimization

Gephi for visualization, network efficiency calculations, and complex network analysis

Cyber–physical traffic control for real-time signal timing optimization

Data-driven decision support using traffic simulations and predictive models

Interactive dashboards for exploring traffic, safety, and energy metrics

Integration of diverse sensors, including Radar Detection Sensors and CCTV cameras, to provide traffic insights

Cloud-based infrastructure for large-scale data processing and management

Edge computing for distributed data processing at the source to reduce latency and enhance real-time responsiveness

Simulation tools for traffic dynamics and energy consumption modeling

Modular design ensuring system scalability and interoperability with IoT services

Complexities of freight operations in route planning, resource allocation, and fleet management

Need for real-time data sharing and decision-making

High environmental impact due to emissions and inefficiencies

Technology-driven fleet and route optimization

Real-time monitoring for cargo and emission control

Enhanced decision-making with integrated data systems

IoT sensors, GPS, mobile and wireless communication technologies

Cloud computing for data analytics and storage

Fuzzy DEMATEL analysis to evaluate interdependencies among capabilities

Roadmap development to categorizes capabilities into base (core) and triggered (dependent)

Integration of ICT to enhance supply chain management and facilitate strategic decision-making

Delays at signalized intersections due to suboptimal traffic signal timing

Persistent traffic congestion negatively impacts fuel consumption, emissions, and public health

Lack of proactive traffic management tools that adapt to real-time demand

Digital twin based adaptive traffic signal control (ATSC) framework for dynamic signal phase optimization

Real-time traffic demand prediction and simulation for proactive traffic management

Parallel simulations to assess various ATSC algorithms and trade-offs in delay reduction

Simulation of real-world traffic flow, including road networks, vehicles, and traffic signals for digital twin

Parallel simulations, real-time data integration, and dynamic algorithm selection for the traffic simulation based on the digital twin

Data aggregation, synchronization, and visualization

High fatality rates due to recurring violations, such as drunk driving and running red lights

Lack of integration between violation data and accident prevention strategies

Insufficient use of advanced sensor technologies in accident hotspots

Installation of sensor-equipped cameras for real-time monitoring and enforcement

Big data analytics to identify accident-prone areas and common violation patterns

Enhanced enforcement through predictive models based on historical trends

Hadoop distributed file system for splitting and storing large datasets across distributed locations

Spark for faster in-memory data processing and trend analysis

Tableau for dynamic dashboards and interactive trend analyses

LiDAR, CCD cameras, and radar for traffic monitoring