Abstract: The present application describes an apparatus having a processor configured to receive a plurality of sensor measurements for each sensor of a plurality of sensors of the system. The processor may be configured to compare the plurality of sensor measurements from each sensor to a respective threshold value, determine, based on the comparisons, a condition of the system having a degraded state and one or more conditions of the system having a normal state, and select at least one of the one or more conditions of the system having a normal state. The processor may be configured to input the condition having degraded state and the at least one condition having a normal state into a diagnostic model. Further, the processor may be configured to isolate, using the diagnosis model, a failed or degraded component of the system.

Abstract: A method is provided for avoiding a conflict during a flight of an aircraft that includes a defined route of travel. The method includes receiving observations of states of the aircraft and a nearby obstacle in an environment of the aircraft as the aircraft travels the defined route. The method includes applying the states to a reinforcement learning framework to predict states of the aircraft to avoid a conflict between the aircraft and the nearby obstacle. The reinforcement learning framework determines maneuvers of the aircraft to avoid the conflict, using a policy trained using a surrogate model of the environment in which movements of the aircraft and the nearby obstacle are simulated, and determines the predicted states of the aircraft from the maneuvers. A collision avoidance trajectory is generated from the predicted states of the aircraft, and output for guidance, navigation or control of the aircraft.

Abstract: A method for identifying aircraft faults, comprising: receiving a dataset comprising a plurality of low priority messages and a plurality of high priority messages, each low priority message identifying a minor aircraft fault and each high priority message identifying a major aircraft fault; for each low priority message, generating an embedding vector which maps the low priority message in an embedding space; for each high priority message, generating an embedding vector which maps the high priority message in the embedding space; providing, to a machine learning unit, the embedding vector for each low priority message of the plurality of low priority messages and the embedding vector for each high priority message of the plurality of high priority messages; and obtaining, from the machine learning unit, a probability of a target high priority message occurring based on each low priority message of the plurality of low priority messages.

Abstract: A method is provided for optimizing a flight of an aircraft with at least one segment of formation flight. The method includes accessing flight plans for flights of a fleet, and transforming the flight plans into values of a set of features that describe segments of the flights. The values are applied to a machine learning model trained to predict the segment(s) during which the aircraft is within a region that includes at least one second aircraft of the fleet that is thereby capable of serving as a leading aircraft in the segment(s) of formation flight in which the aircraft is a trailing aircraft. A notification is sent to the aircraft of the segment(s) and the second aircraft capable of serving as the leading aircraft. And a second notification is sent to the second aircraft of the segment(s) and the aircraft capable of serving as the trailing aircraft.

Abstract: A method of optimizing flights of a fleet of aircraft is provided. The method includes accessing flight plans for flights of a fleet of aircraft through an air transportation network, and applying the flight plans to a reinforcement learning model configured to determine maneuvers for each aircraft on a respective flight that achieves a respective maximum cumulative value of an operational efficiency metric across the flights of the fleet of aircraft, one or more of the maneuvers constituting a deviation from a respective flight plan. A comparison of respective maximum cumulative values of the operational efficiency metric is performed for the aircraft of the fleet of aircraft, one of the aircraft is selected based on the comparison, and a notification of the deviation from the respective flight plan is sent to the one of the aircraft.

Abstract: A method of architecting machine learning pipelines is provided. Example implementations of the method include causing an apparatus to generate a graphical user interface (GUI) from which a computing platform is accessible to architect machine learning pipelines. In example implementations, the method includes for a machine learning pipeline for a phase in the machine learning lifecycle: building software components that are separate, distinct and encapsulate respective processes executable to implement the phase in the machine learning lifecycle, the software components including ports that are communication endpoints of the software components. The method further includes interconnecting the software components with connections attached to the ports and thereby forming a network of interconnected software components that embodies the machine learning pipeline.

Abstract: Example implementations relate to autonomous airport runway navigation. An example system includes a first sensor and a second sensor coupled to an aircraft at a first location and a second location, respectively, and a computing system configured to receive sensor data from one or both of the first sensor and the second sensor to detect airport markings positioned proximate a runway. The computing system is further configured to identify a centerline of the runway based on the airport markings and receive sensor data from both of the first sensor and the second sensor to determine a lateral displacement that represents a distance between a reference point of the aircraft and the centerline of the runway. The computing system is further configured to control instructions that indicate adjustments for aligning the reference point of the aircraft with the centerline of the runway during subsequent navigation of the aircraft.

Abstract: A method for identifying aircraft faults, comprising: receiving a dataset comprising a plurality of low priority messages and a plurality of high priority messages, each low priority message identifying a minor aircraft fault and each high priority message identifying a major aircraft fault; for each low priority message, generating an embedding vector which maps the low priority message in an embedding space; for each high priority message, generating an embedding vector which maps the high priority message in the embedding space; providing, to a machine learning unit, the embedding vector for each low priority message of the plurality of low priority messages and the embedding vector for each high priority message of the plurality of high priority messages; and obtaining, from the machine learning unit, a probability of a target high priority message occurring based on each low priority message of the plurality of low priority messages.

Abstract: A method for identifying aircraft faults, comprising: receiving aircraft health dataset comprising plurality of maintenance identifiers which each identify aircraft fault; storing diagnostics database storing plurality of part identifiers which each identify part of aircraft which is possible cause of generation of at least one maintenance identifier; generating graph of plurality of maintenance identifiers and plurality of edges in which maintenance identifiers are connected to one another by edge if maintenance identifiers are identified by common part identifier in diagnostics database; extracting clique from graph, clique comprising plurality of maintenance identifiers and respective plurality of edges of graph; determining intersection between at least two edges of clique; identifying candidate part identifier which is common to intersecting edges of clique, candidate part identifier identifying part of aircraft which is possible cause of generation of at least some of maintenance identifiers of clique;

Abstract: Example implementations relate to autonomous airport runway navigation. An example system includes a first sensor and a second sensor coupled to an aircraft at a first location and a second location, respectively, and a computing system configured to receive sensor data from one or both of the first sensor and the second sensor to detect airport markings positioned proximate a runway. The computing system is further configured to identify a centerline of the runway based on the airport markings and receive sensor data from both of the first sensor and the second sensor to determine a lateral displacement that represents a distance between a reference point of the aircraft and the centerline of the runway. The computing system is further configured to control instructions that indicate adjustments for aligning the reference point of the aircraft with the centerline of the runway during subsequent navigation of the aircraft.

Abstract: Methods and systems are provided for processing natural language for machine learning analytical systems. The method includes receiving, at a processor, an input including text representing one or more observed parameters of an environment. The inputted text is in a natural language format. The processor parses the input and extracts the one or more parameters. A function is defined representing a domain of the one or more observed parameters based upon the one or more extracted parameters.

Abstract: Technologies are described herein for detecting and recovering from a fire event within an aircraft. The technologies receive sensor data from a number of sensors associated with an aircraft. A determination is made as to whether the sensor data exceeds predefined thresholds indicating the fire event within the aircraft. In response to determining that the sensor data exceeds the predefined thresholds indicating the fire event, the technologies determine a location of the fire event within the aircraft based on the sensor data and depower components of the aircraft associated with the fire event. The technologies then initiate a fire suppressant mechanism within the aircraft directed to the location of the fire event.

Abstract: A method is disclosed for controlling at least one remotely operated unmanned object. The method may involve defining a plurality of body movements of an operator that correspond to a plurality of operating commands for the unmanned object. Body movements of the operator may be sensed to generate the operating commands. Wireless signals may be transmitted to the unmanned object that correspond to the operating commands that control operation of the unmanned object.

Abstract: Technologies are described herein for detecting and recovering from a fire event within an aircraft. The technologies receive sensor data from a number of sensors associated with an aircraft. A determination is made as to whether the sensor data exceeds predefined thresholds indicating the fire event within the aircraft. In response to determining that the sensor data exceeds the predefined thresholds indicating the fire event, the technologies determine a location of the fire event within the aircraft based on the sensor data and depower components of the aircraft associated with the fire event. The technologies then initiate a fire suppressant mechanism within the aircraft directed to the location of the fire event.

Abstract: A method is disclosed for controlling at least one remotely operated unmanned object. The method may involve defining a plurality of body movements of an operator that correspond to a plurality of operating commands for the unmanned object. Body movements of the operator may be sensed to generate the operating commands. Wireless signals may be transmitted to the unmanned object that correspond to the operating commands that control operation of the unmanned object.