Urban traffic congestion is a major challenge in modern cities, leading to increased travel time, fuel consumption, and environmental pollution. Conventional traffic signals rely on fixed timing cycles that cannot adapt to changing traffic conditions. This paper presents APEX (Adaptive Predictive EXchange Traffic Network), a modular, computer vision-based traffic control system designed to retrofit existing signals with adaptive capabilities. A 360-degree camera monitors all lanes, and a scheduling algorithm dynamically prioritizes traffic while ensuring fairness through a waiting-time adjustment parameter. A predictive component estimates near-future traffic flow to optimize signal timing. The design emphasizes low cost, rapid installation, and minimal infrastructure modification, providing a practical upgrade path for cities seeking efficient traffic management. Simulation results demonstrate significant reductions in vehicle delay, queue length, and carbon emissions compared to conventional fixed-time traffic signals.
KarthikAramana (Mon,) studied this question.