mobility.

The rapid growth of urban vehicle numbers and insufficient transportation infrastructure pose significant challenges for developing smart cities. The overwhelming presence of cars in urban settings contributes to increased air and noise pollution, negatively impacting environmental quality and public health. This congestion diminishes fuel efficiency and exacerbates delays, particularly in areas where speed regulation systems are inadequate [1]. The resultant pollution and traffic disturbances necessitate advanced traffic management strategies extending beyond traditional signal control mechanisms, which are now integrated into comprehensive smart city frameworks to enhance traffic safety and efficiency [2]. Collecting accurate traffic data is essential for effective traffic management. City authorities have long employed traditional methods like speed detectors, radar, and cameras to monitor traffic flow [3]. However, the Internet of Things (IoT) presents a transformative approach to data collection, enabling real-time monitoring and analysis of traffic conditions in smart city environments [4-6]. Effective detection of traffic congestion remains a significant challenge for traffic control systems. Major thoroughfares frequently experience bottlenecks in many urban areas due to limited public transportation options and inadequate traffic management initiatives. Without proactive measures, these issues will likely escalate [7]. Various strategies have been proposed to alleviate traffic congestion in smart cities. Some approaches focus on identifying optimal routes for individual vehicles, which, while potentially beneficial, can inadvertently shift congestion to other areas [8]. Others have introduced machine learning techniques, including decision trees, support vector machines, and clustering algorithms, to enhance traffic prediction capabilities [9]. Nonetheless, many existing models fail to address the complex dynamics of urban traffic effectively. A robust system that accurately identifies and manages traffic conditions could improve congestion management and overall urban mobility [10]. This paper proposes an innovative IoT-based predictive analytics framework to enhance smart city traffic management. This approach aims to optimize traffic flow and minimize waiting times at intersections by leveraging microprocessor technology and advanced data analytics. The subsequent sections of this paper will explore relevant literature in traffic prediction and management (Section 2), detail the predictive analytics methodology employed (Section 3), present the findings of our research (Section 4), and conclude with insights and recommendations for future studies (Section 5).

Recent advancements in IoT-based traffic management have seen several innovative approaches to improve urban mobility. Yao et al. [11] introduced a traffic classification mechanism for the IoT within smart city frameworks. Their system integrates extraction, collection, and classification functionalities while eliminating the need for manual traffic selection methods. Notably, the inclusion of capsule networks enhanced the accuracy of traffic classification. Although the system demonstrated high performance, it faced challenges with measurement costs and extended training times. In another study, Bai et al. [12] focused on predicting traffic congestion in smart cities through a relative congestive tensor. They developed a congestion matrix emphasizing regional traffic networks and utilized a Long Short-Term Memory (LSTM) approach for forecasting congestion across various road nodes. Their experiments indicated significant improvements over baseline models, although they noted issues with spatial sparsity and the need for better separation within the road network. Mondal and Rehena [13] explored traffic congestion classification using an Artificial Neural Network (ANN). Their Intelligent Traffic Management System (ITMS) dynamically adjusts traffic signals and recommends alternative routes based on real-time congestion data. While the model performed effectively across different road segments, it encountered challenges related to storage overhead and lengthy training periods. Joo et al. [14] proposed a reinforcement learning-based approach for managing traffic signals in smart cities, aiming to optimize the flow of vehicles at intersections. Their system used Q-learning to adaptively manage traffic signals, resulting in shorter queue times and reduced waiting periods. However, the model's accuracy in signal operation could still be improved. Rego et al. [15] presented a framework for managing emergencies in urban areas through a networked control system. This design utilized IoT devices, including traffic cameras and lighting systems, to enhance the routing of emergency service vehicles. Their algorithm effectively reduced emergency response delays by 33%, although it faced limitations in applicability to diverse real-world smart city scenarios. Further developments in traffic management highlight the need for robust analytics and modeling, especially with the increasing integration of IoT and artificial intelligence [16]. Intelligent traffic signal management systems (IS) have expanded to include features like public transit prioritization, emergency vehicle prioritization, adaptive signal control, and environmentally conscious operational modes [17], [18]. Despite the complexities associated with these new systems, managing traffic flow efficiently remains a critical concern, with many approaches focusing on optimizing 60-second intervals for signal changes [19].

Traffic management has become a significant challenge in urban areas, largely due to the continuous increase in vehicle numbers, which leads to recurring congestion issues. Integrating the IoT within smart cities presents an excellent opportunity to address these traffic challenges by developing ITMSs. This paper introduces an effective Optimized Weight Elman Neural Network (OWENN) algorithm for traffic prediction and signal control, utilizing an Intel 80,286 microprocessor tailored for smart city applications. The proposed system, illustrated in Fig. 1, encompasses three main phases: 1) IoT data collection, 2) traffic prediction and feature selection, and 3) traffic signal control. Each of these phases will be explained in the following subsections.

The implementation of IoT-based predictive analytics for traffic management has demonstrated significant improvements in traffic control efficiency. The system could predict traffic patterns and adjust signal timings by collecting real-time data from IoT sensors deployed at intersections and other key locations. The results indicate a substantial reduction in traffic congestion, especially during peak hours, as the predictive model could foresee traffic build-ups and manage signal phases to prevent bottlenecks.

Implementing IoT-based predictive analytics for traffic management presents a transformative approach to improving urban mobility. By leveraging real-time data from IoT sensors and applying predictive models, this system has demonstrated the potential to significantly reduce traffic congestion, optimize signal timings, and enhance overall traffic flow. The results show that predictive analytics can dynamically adjust to traffic conditions, reducing waiting times at intersections and minimizing environmental impacts by lowering fuel consumption and emissions. Despite some challenges, such as the reliance on sensor coverage and the need for continuous model updates, the system's scalability and integration into smart city infrastructures offer promising prospects. As cities grow and traffic patterns evolve, this approach can be a foundation for more intelligent, responsive, and sustainable traffic management systems. With further advancements in IoT technology and data integration, the system can be expanded to manage larger areas and incorporate additional factors, leading to more efficient and eco-friendly urban transportation networks.

The author solely conducted the research and prepared the manuscript and has approved its final version.

Consent for publication has been obtained from the author.

This article does not include experiments involving humans or animals.

This research received no specific grant from any funding agency in the public, commercial, or not-for-profit sectors.

All data are included in the text.

The author declare that they have no known competing financial interests or personal relationships that could have influenced the work reported in this paper.