An information generation method is performed by an information generation device which generates information for a learning model that infers whether a mobile object is movable in a predetermined region. The information generation method includes: obtaining at least (i) first information and (ii) second information when the mobile object moves in a first region, the first information being obtained from a sensor provided in the mobile object, the second information relating to movement of the mobile object; inferring whether the mobile object is movable in the first region according to the second information; and generating fourth information for a learning model, the fourth information associating the first information, the second information, and third information with one another, the third information indicating an inference result which is obtained in the inferring.

A method for training a control policy for a robot device includes acquiring a reference state of an environment of the robot device and a reference observation of the environment for the reference state. The method also includes generating, for each of a plurality of errors of an estimation of a pose of the robot device, an observation that is disturbed with respect to the reference observation according to the error of the pose estimation and a training data element comprising the generated observation as a training input. The method further includes training the control policy using the generated training data elements.

A system for managing flight parameters of aircraft is parametrized with overarching flight cost objectives. For a first flight, parameters of a cost function for the first flight are determined relative, respectively, to different cost factors. Flight parameters are optimized so as to minimize the cost function. Avionics of an aircraft carrying out the first flight are programmed with the flight parameters. On detecting an event in flight requiring the flight parameters to be revised, the parameters of the cost function are recalculated, as well as the flight parameters, and the avionics are reprogrammed accordingly. The method is repeated for at least a second flight, taking into account the effective contribution of the first flight to the overarching objectives, and so on. Thus, an airline can carry out overarching multi-objective optimization on its flights.

An unmanned aerial vehicle (UAV) or drone executes a neural network to assist with detecting and responding to attacks. The neural network may monitor, in real time, the data stream from a plurality of onboard sensors during navigation and may communicate with a high-altitude pseudosatellite (HAPS) platform. For example, if the neural network detects a cyber-attack but determines that it does not interfere with external communications, it may shift navigation control of the drone to the HAPS.

A method for guiding an autonomous aircraft, the aircraft includes an automatic pilot, a plurality of sensors and an imaging unit, the aircraft being configured to fly over a geographic zone comprising overflight prohibited zones, the guidance method can advantageously comprise a phase of real flight of the autonomous aircraft by using a given guidance law, comprising the following steps: determining a current state of the autonomous aircraft; determining an optimum action to be executed by using a neural network receiving the current state; determining a plurality of control instructions compatible with the guidance law based on the optimum action to be executed; transmitting to the automatic pilot the plurality of control instructions, which provides a new state of the autonomous aircraft.

Systems, methods, and autonomous vehicles may obtain sensor data associated with an environment surrounding an autonomous vehicle; provide the sensor data to a plurality of plugins; independently determine, with each plugin, based on the sensor data, whether to request a remote guidance session for the autonomous vehicle, each plugin of the plurality of plugins including a different model that is applied by that plugin to the sensor data to determine whether to request the remote guidance session; receive, from at least one plugin, a request to initiate the remote guidance session; and in response to receiving the request to initiate the remote guidance session, communicate with a computing device external to the autonomous vehicle to establish the remote guidance session.

A method for constructing episodic memory model based on rat brain visual pathway and entorhinal-hippocampal structure mainly applied to environment cognition and navigation of an intelligent mobile robot to complete tasks of environment cognition map construction and target-oriented navigation is provided. The image information of the environment, the head-direction angle and speed of the robot are collected, and then the head-direction angle and speed of the robot are input into the entorhinal-hippocampal CA3 neural computational model to obtain the robot's precise position. The visual information is input into the computational model of the visual pathway to obtain the scene information in the current vision of the robot. The above two kinds of information are fused and stored in a cognitive node with the topological relationship. Utilizing scenario information to correct the path integration errors during the exploration process of the robot, thereby constructing the episodic cognitive map representing the environment.

Systems and methods for enhanced MHV operation using an automation processing system for bystander detection and bystander pose estimation to control operation of the MHV.

Examples relate to a method for generating an Unmanned Aerial Vehicle (UAV) controller model for controlling an UAV, a system including an UAV, a wind generator, a motion-tracking system and a control module, and to an UAV. The method for training the UAV controller model includes providing a wind generator control signal to a wind generator, to cause the wind generator to emit a wind current towards the UAV. The method includes operating the UAV using the UAV controller model. A flight of the UAV is influenced by the wind generated by the wind generator. The method includes monitoring the flight of the UAV using a motion-tracking system to determine motion-tracking data. The method includes training the UAV controller model using a machine-learning algorithm based on the motion-tracking data.

Aspects of embodiments to systems and methods for navigating a mobile platform using an imaging device on the platform, from a point of origin towards a target located in a scene, and without requiring a Global Navigation Satellite system (GNSS), by employing the following steps: acquiring, by the imaging device, an image of the scene comprising the target; determining, based on analysis of the image, a direction vector pointing from the mobile platform to the target; advancing the mobile platform in accordance with the direction vector to a new position; and generating, by a distance sensing device, an output as a result of attempting to determine, with the distance sensing device, a distance between the mobile platform and the target. The mobile platform advanced towards the target until the output produced by the distance sensing device is descriptive of a distance which meets a low-distance criterion.