In this study, a fuzzy logic compensator incorporated in a Smith predictor is designed to control an induction motor using a frequency converter in a network control system, thus forming an intelligent controller. Through research on the state of artificial intelligence techniques applied to network control systems, controllers based on fuzzy logic are taken as the main axis to improve the response of network control systems. A commonly used technique is the Smith predictive control algorithm based on error modeling for time delay compensation of a networked control system. When the controlled plant is a nonlinear time-varying system, the Smith predictor identification model error also varies, resulting in poor control performance. For this problem, the fuzzy control method, which is based on the self-adjustment of the gain of the Smith predictor identification model to improve the robustness of the compensator, is introduced. The proposed compensator was implemented in a training kit to evaluate its performance in a real plant. The experimental and simulation results demonstrate that the fuzzy-Smith compensator has better performance in time delay compensation, with a shorter response time, small overshoots of less than 5%, and correction of modeling errors reaching a minimum error of 3%, compared to a conventional PID controller for a network control system.