This study explores the use of computer vision combined with AI algorithms as a potential replacement for traditional PIR and microwave sensors commonly used in lighting control systems. By implementing an AI-based Intelligent Lighting System in a library, we successfully extended the detection range to more than 20 m, several times greater than the previous sensor systems. The system was able to switch lights on or off within two seconds of detecting the presence or absence of an object, meeting the requirements for efficient smart control and energy savings. During the research, it was found that signal transmission across the network was a critical factor in the system’s overall response time. In this study, the time from the PLC server to each lighting fixture accounted for approximately 40% of the total system response time. Additionally, significant variations in response time were observed during the power supply’s activation, largely due to transient electrical responses in the components used to drive the lighting fixtures.