This research addresses the challenge of precise trajectory tracking control in the KUKA YouBot robotic manipulator, focusing on an alternative computed torque control method. The study aims to overcome the inherent nonlinearities in robotic systems, taking into account both structured uncertainties (model errors) and unstructured uncertainties (friction). The proposed approach implements a cascaded PID controller, combining position and velocity controllers to enhance disturbance rejection. A critical aspect of the research is the accurate identification of dynamic model parameters, as those provided by manufacturers are often imprecise. The work includes a detailed analysis of the semantics used in rigid body algorithms, contributing to a more accurate construction of geometric relationships. To improve controller performance, the study incorporates friction compensation. The implementation is complemented by a safety control layer that continuously monitors joint states, protecting valuable system hardware. Experimental validation includes gravity compensation tasks and controller testing with analytical trajectories. Friction modeling and compensation experiments were conducted, and the control scheme was evaluated on the joints furthest from the base. The results demonstrate that, despite inaccuracies in dynamic model parameters, the controller achieves precise trajectory tracking with acceptable errors in the manipulator’s joints. This achievement is significant given the complexity of the system and the challenges posed by uncertainties.