Streaming Data Analytics for Smart Traffic Signal Optimization

Abstract

Urban road networks are becoming increasingly congested due to rising vehicle populations, inefficient traffic management strategies, and inconsistent signal timings. Traditional fixed-time traffic signal systems fail to adapt to real-time conditions, resulting in long queues, increased travel times, and higher carbon emissions. With advancements in Internet of Things (IoT) sensors, vehicular communication systems, and cloud-edge computing, streaming data analytics has emerged as a promising solution for intelligent traffic signal optimization. This study proposes a real-time traffic signal optimization framework leveraging streaming data analytics to dynamically adjust signal timings based on live vehicular flow, queue lengths, and predicted congestion levels.

The proposed system ingests high-velocity data from roadside sensors, GPS-enabled vehicles, and surveillance cameras, processes it in real time using distributed stream processing engines, and applies adaptive control algorithms to optimize green-light intervals. Simulation experiments conducted on a realistic traffic network model in SUMO (Simulation of Urban Mobility) demonstrate a 35% reduction in average vehicle waiting time, a 28% improvement in throughput, and a 21% reduction in CO₂ emissions compared to static timing strategies. This work highlights the importance of low-latency analytics pipelines, predictive congestion modeling, and machine learning-driven decision-making for next-generation smart cities.