Next, the design and construction of an upper limb movement emulation system through non-invasive electromyography and artificial neural networks is described in detail. Implementing electromyographic signals captured by electrodes located in the user's arm and refined by an EMG signal conditioning system that in turn is responsible for sending obtained bio-potentials to an array of control devices, Arduino class, which possess a program of neural networks previously trained in Visual Studio 2012, using extracted and defined movement patterns thanks to the use of a digital EMG oscilloscope. In this system, depending on the levels and pattern model, the control of the prosthesis activates each one of the actuators (motor-reducers) thanks to previously defined movement sequences.