Urban traffic congestion poses a persistent challenge in modern cities, leading to prolonged delays, increased fuel consumption, and higher emissions. This paper introduces an intelligent and adaptive traffic light control system that leverages real-time vehicle detection through the YOLOv8 model, combined with two green light adjustment strategies: a dynamic threshold controller based on a PLC and a fuzzy logic controller. The PLC includes a fallback mechanism with a fixed cycle mode to ensure operation during camera failures, software malfunctions, or communication loss. A manual emergency stop button is also integrated for safety. The system was validated through simulations using Pygame and through a real-world hardware setup. The results confirm the practicality of merging AI-based detection with industrial control systems for smarter traffic management.