Abstract: The present invention refers to an intelligent controller for predicting and automatically compensating for variations in cycle time, in machines for forming articles of glass or other materials, by predicting estimated cycle periods of the machine based on present and past measurements, so as to effect compensations without mechanical failures and/or without having to repeat cycles, even in cases when the information signals from the machine are lost; and which comprises: at least one forming cycle sensor for the whole of the machine; a signal conditioner which conditions the signals detected by the forming cycle sensor of the machine; at least one real-time data processing unit that calculates the initial forming cycle time and the times of operation of the forming mechanisms and compares the present and past cycle times in order to predict the next start-up of the forming cycle of whatever section of the machine, and/or determine the cycle in cases where the detected signals are lost, and supplies said pre