This paper proposes an initial fuzzy control methodology for guiding robotic movements based on Fuzzy Cognitive Maps (FCMs). The model integrates fuzzification, causal reasoning via an adjacency matrix, and defuzzification to determine real-time motor commands in an Arduino-based mobile robot. Linguistic Terms were defined for distance and speed. The influence weights were assigned and tuned through experimental test. The methodology was validated through physical experiments in which sensor readings and control outputs showed coherent behavior, demonstrating the model's ability to adapt to obstacle proximity and respond with appropriate movement adjustments. In physical trials, the FCM controller performed better than a baseline PID in deviation rate (76.32% vs 42.11%) and deviation time (1.53 ± 0.82 s vs 3.71 ± 1.62 s), while RMSE remained similar, indicating comparable path-tracking accuracy. These results support FCMs as an interpretable and responsive alternative for embedded robotic control.