Nowadays, electric motors are an integral part of most modern electromechanical systems that are used in industry. It follows that industrial processes are becoming more dependent on their efficiency. If faults in electric motors are not rectified, they can lead to malfunctions and accidents, as well as production downtime. Symmetry of a three-phase system means that the voltage and current in the three phase conductors are equal to each other, with a period of 120°. Asymmetry occurs if one of these conditions or both conditions are violated at the same time. In most cases, asymmetry is caused by loads. Predictive diagnostics is the most effective way to identify motor faults while the motor is in operation and prevent the likelihood of failure. Predictive diagnostics can identify problems that could lead to major failures, thus reducing production downtime and maintenance costs. The paper discusses the control and diagnosis of electric motors using prediction techniques. In particular, the use of neural network models and predictive control to improve accuracy and reliability is investigated. The main objective of this research is to develop a neural network controller based on predictive model predictive control (MPC), which will improve the quality of the control and diagnostics system of electric motors, ensuring their stability and preventing possible malfunctions.