Abstract: A device for managing air traffic, in an airspace includes a reference aircraft and at least one other aircraft, the device receiving a three-dimensional representation of the airspace at a time when an air conflict is detected between the reference aircraft and the at least one other aircraft, the device comprising an airspace-encoding unit configured to determine a reduced-dimension representation of the airspace by applying a recurrent autoencoder to the three-dimensional representation of the airspace at the air-conflict detection time; a decision-assisting unit configured to determine a conflict-resolution action to be implemented by the reference aircraft, the decision-assisting unit implementing a deep-reinforcement-learning algorithm to determine the action on the basis of the reduced-dimension representation of the airspace, of information relating to the reference aircraft and/or the at least one other aircraft, and of a geometry corresponding to the air conflict.

Abstract: In the field of air traffic control, a method is provided to determine a processing complexity of an ATC situation. For this purpose, the method includes grouping parameters of the paths by pairs of paths in a matrix, applying to this matrix a transformation aiming to concentrate the energy, then calculating the complexity index of the ATC situation as a function of the concentration level of the energy per component.

Abstract: A computer-implemented method is provided for training a supervised machine learning engine able to predict characteristics of aircraft trajectories from parameters of an aircraft, and environment parameters of the aircraft trajectory. A system able to train the supervised machine learning engine, a system for using the engine, and a computer-implemented method for using the engine are provided. The methods and systems provided are particularly useful for air traffic flow management applications.

Abstract: A device for managing air traffic, in an airspace includes a reference aircraft and at least one other aircraft, the device receiving a three-dimensional representation of the airspace at a time when an air conflict is detected between the reference aircraft and the at least one other aircraft, the device comprising an airspace-encoding unit configured to determine a reduced-dimension representation of the airspace by applying a recurrent autoencoder to the three-dimensional representation of the airspace at the air-conflict detection time; a decision-assisting unit configured to determine a conflict-resolution action to be implemented by the reference aircraft, the decision-assisting unit implementing a deep-reinforcement-learning algorithm to determine the action on the basis of the reduced-dimension representation of the airspace, of information relating to the reference aircraft and/or the at least one other aircraft, and of a geometry corresponding to the air conflict.

Abstract: A computer-implemented method includes receiving a trajectory request from an unmanned aerial vehicle, the request comprising: an initial point; a final point; at least one manoeuvrability parameter of the unmanned aerial vehicle; computing a plurality of optimized 4D trajectories between the initial point and the final point, complying with the at least one manoeuvrability parameter, and avoiding obstacles in an airspace, each 4D trajectory being associated with a performance score; a flight simulator simulating the plurality of 4D trajectories in order of decreasing performance score, until a 4D trajectory is considered to be flyable by the flight simulator; sending the trajectory considered to be flyable by the flight simulator to the unmanned aerial vehicle.

Abstract: The embodiments of the invention provide a device for predicting the value of a variable intended to be used by a computer-implemented control system, the variable depending on multiple parameters, the parameters comprising a non-explicit parameter. Advantageously, the prediction device comprises a first neural network-based predictor configured so as to compute an estimate of the non-explicit parameter and a second neural network-based predictor configured so as to compute an estimate of the value of the variable from the estimate of the non-explicit parameter, the two predictors receiving an input dataset, each neural network being associated with a set of weights.

Abstract: A method for determining the trajectory of at least one mobile element from position data, includes an initial step consisting in classifying a set of positions relating to at least one detected mobile element by applying a first data classification algorithm to the set of positions, which provides an initial trajectory relating to each detected mobile element. The method comprises the following steps, implemented on each current observation window: classifying each new position detected in at least one trajectory by applying a second data classification algorithm; identifying, for each detected mobile element, the positions relating to the detected mobile element; determining an intermediate complete trajectory for each detected mobile element; determining a final complete trajectory for each detected mobile element.