Urban areas are experiencing a rapid increase in vehicle numbers, which is
outpacing the development of necessary traffic infrastructure. This imbalance has led to
heightened congestion, particularly during traffic disruptions, with significant
implications for economic development, public health, and environmental
sustainability. The rise in traffic congestion correlates with increased traffic accidents
and greenhouse gas emissions, extending travel times and adversely affecting the
quality of life in urban settings.
To address these challenges, contemporary societies are encouraged to implement
traffic management systems that aim to mitigate congestion and its associated negative
effects. These systems employ a range of applications and management tools aimed at
improving the effectiveness and security of transportation networks. By collecting and
analyzing data from multiple sources, traffic management systems can identify risks
that impede traffic flow and propose effective solutions.
The review categorizes Intelligent Transportation Systems (ITS) into four main groups:
traffic data collection solutions, traffic management solutions, congestion mitigation
strategies, and travel time estimation tools. Each category is examined in terms of its
foundational technologies, advantages, and limitations. The study emphasizes the role
of machine learning and computational intelligence in developing methodologies to
alleviate congestion and improve travel time predictions.
The paper also assesses the integration of artificial intelligence (AI) in traffic
management, focusing on research from the past decade. It identifies trends and
characteristics of AI applications, particularly in the context of explainable I (XAI)
within traffic management systems. The findings culminate in a conceptual framework
known as the DPP model (Descriptive, Predictive, and Prescriptive), which is
accompanied by a speculative application scenario for the year 2030. The study
highlights the need for further research to enhance user acceptance of AI systems in air
traffic management (ATM) and emphasizes the importance of developing effective
XAI methodologies.
Additionally, the paper discusses the Internet of Things (IoT) and its role in connecting
various networks, including ITS, Smart Grids, and Smart Cities. The interconnectivity
of these networks generates substantial traffic with diverse quality of service (QoS)
requirements, challenging traditional network management protocols. The emergence
of machine learning has significantly influenced decision-making processes related to
network protocols, with ongoing research exploring various ML algorithms to address
challenges in Fifth Generation (5G) and IoT heterogeneous networks, including
cybersecurity threats and transport layer issues. The analysis of these techniques
reveals their strengths, weaknesses, and practical applications in managing complex
network environments.
Keywords: Artificial intelligence (AI), Connected autonomous vehicles (CAVs), Fifth generation (5G), Intelligent traffic management system (ITMS), Internet of things (IoT), Maharashtra state road development corporation (MSRDC), Realtime management (RTM), Sustainable urban mobility (SUMs), Traffic management systems (TMS), Traffic safety management (TSM).
