.Mobile robots are widely used and provide convenience to people in places like restaurants and airports. However, there are situations in which their limited autonomy necessitates additional user interaction. While existing studies suggest using hand gestures to manipulate mobile robots, the practical application of such an approach is limited, as they cannot fully address the needs of the user needs and can just primarily be operated in teleoperation environments. In this work, a novel hand gesture operation method has been proposed that can operate a wheeled mobile robot through a manipulation method. In addition to implementing the suggested gesture recognition technique, representative hand gestures were selected by conducting experiments to identify natural gestures based on context for improving usability. A combination of the YOLOv8-based object recognition and pose recognition models was employed to achieve the hand gesture recognition function. Finally, a wheeled mobile robot was also designed to evaluate the hand gesture manipulation system for performing appropriate movements according to the gesture. The overall system was evaluated through experiments and interviews with participants about the workload and usability of the gesture manipulation system.