Abstract: In this paper, a rule of real-time motion planning for visual servo robots due to the limited actuator power is presented. The rule employs a fuzzy logic adaptive algorithm to modify the motion of robots on-line based on the potential of actuators such as a DC motor according to its mechanical character and performing condition. Time-optimal motion is achieved, disregarding the influence of singularity. A simulation of 2-DOF planar manipulator is also presented in this paper.