This paper deals with the implementation of a embedded position control system using a robust nonlinear predictive controller, which is employed to control simultaneously angular positions of the base and arm of an angular manipulator of 2DOF (2 Degrees of Freedom). The design of such control system requires the derivation of the dynamic nonlinear model of the manipulator, as well as the determination of the corresponding predictive control law. Intensive simulation studies permitted to find out the initial values of the tuning parameters of the predictive controller. A CRIO (Compact Reconfigurable Input/Output) device was used to embed the control system. Good performance of the predictive control system was verified via experimentation. © 2013 IEEE.