Big Data and I2X Communication Infrastructure for Traffic Optimization and Accident Prevention on Automated Roads

Abstract

This paper presents a scalable big data infrastructure designed to support traffic optimization and accident prevention in automated and connected road environments. The proposed system integrates real-time data acquisition from heterogeneous sources, including multichannel roadside camera gantries and IoT-enabled vehicle telemetry. The architecture is built upon technologies such as Apache Kafka, Apache Beam, and MongoDB, enabling high-throughput data ingestion, processing, and storage. To validate its performance, two experimental use-cases were developed: one for large-scale image ingestion and another for vehicle telemetry data streaming. The system successfully handled over 48 GB of image data and more than 3.4 million telemetry messages under real-time constraints. Results show that applying data compression techniques—such as resolution reduction and transmission throttling—reduced image upload durations by up to 77%, improving ingestion efficiency without compromising system robustness. These findings demonstrate the feasibility of deploying the proposed infrastructure as a foundational layer for future intelligent traffic anagement systems.