This study explores the potential application of virtual sensing techniques based on current frequency variations in stepper motors that drive a robot in motion. The goal is to analyze current behaviour under varying speed conditions on each motor of the robot arm used for testing, providing insights into operational performance. The methodology involves generating position and speed trajectories using third-degree polynomials and acquiring data through encoders and current sensors via STM32 Nucleo boards. Data acquired align with theoretical predictions, showing that the current amplitude remains within the driver-defined limits, while the frequency changes in relation to speed. The results obtained using a digital twin highlight virtual sensing’s potential to enhance robot control and diagnostics, setting a starting point for more advanced applications such as indirect measurements of forces applied on the end effector and robot’s movement reconstruction based only on current data.