Method and system using a neural network to predict a time of occupancy of a runway by an aircraft

A method for predicting a time of occupancy of a runway of an aerodrome by an aircraft when the aircraft lands, said method comprising a sequence of the following steps:

• a generation step, implemented by a generation unit, for generating at least one current value of a parameter relating to a movement of an aircraft;
• a selection step, implemented by a selection unit, for allowing a crew member of the aircraft to select an exit from the runway used for landing, said exit having an exit type and an exit distance;
• a data processing step, implemented by an avionics computer, for computing a value of a time of occupancy of the runway, wherein the avionics computer, to compute said value of the time of occupancy of the runway:
• utilizes at least one data from:
• an initial ground speed generated in the generation step, and the exit type and the exit distance of the exit selected in the selection step;
• and
• implements a prediction model comprising a neural network trained on landing data originating from previous landings;
• a comparison step, implemented by the avionics computer, of comparing the computed value of the time of occupancy of the runway with a threshold;
• a transmission step, implemented by a transmission unit, for transmitting at least said value of the time of occupancy of the runway to a user system;
• and
• an execution step, implemented by the user system, of executing an avoidance task on at least one of the aircraft or a subsequent aircraft when the time of occupancy of the runway exceeds the threshold as determined in the comparison step.