Hrl Laboratories, LLC

The Boeing Company

Gm Global Technolgoy Operations, Inc.

Tsai-Ching Lu

Franz David Betz

David L Allen

Kang-Yu Ni

Jiejun Xu

Charles E Martin

Alice A Murphy

Rashmi N Sundareswara

Hankyu Moon

David J Huber

Aruna Jammalamadaka

Steve C Slaughter

Qin Jiang

Ryan F Compton

Alex N Waagen

Christopher R Wezdenko

Nigel D Stepp

Matthew S Keegan

Michael D Howard

Samuel D Johnson

David A Jurgens

Derek Samuel Fok

Swarup S Medasani

Dmitriy V Korchev

Gavin D Holland

Alexander Norman Waagen

Yuri Owechko

Yang Chen

Heiko Hoffmann

Karim El Defrawy

Shahriar Alam

Craig A Lee

John E Harrison

Victor Ardulov

Stephen C Slaughter

Douglas Peter Knapp

John L Ross

Deepak Khosla

David W Payton

Yilu Zhang

Mutasim Abdurrahman Salman

Praveen K Pilly

Aleksey Nogin

Ryan M Uhlenbrock

Vincent De Sapio

John Anthony Cafeo

Kirby Joe Keller

Dnyanesh Rajpathak

Chong Ding

Hyun (Tiffany) J Kim

Alexei Kopylov

Chongwon Cho

Alejandro Nijamkin

Eric P Tressler

Dana M Warmsley

Richard J Patrick

Shane M Roach

Daniel K Xie

Nikolay Pavlovich Laptev

Laura Cruz-Albrecht

Patrick Park

Ian A Maxon

Alexey Rudenko

Sofia S V Apreleva

Pejman Mahboubi

Veronika Strnadova

Veronika Stmadova

Sofia Apreleva

Krishna Bathina

Hyun J Kim

A method, apparatus, system and computer program product for managing a platform. A computer system generates a training dataset comprising historical metric values from historical sensor information for a set of metrics for a part and historical maintenance events for the part. The computer system trains a machine learning model using the training dataset. The computer system determines maintenance thresholds for a metric in the set of metrics for performing maintenance on the part using the machine learning model trained with the training dataset. The computer system selects a maintenance threshold from the maintenance thresholds meeting an objective, wherein the maintenance threshold is used to determine when a maintenance action is needed for the part.

Vehicle health management using a counterfactual machine learning model

An apparatus includes memory to store computer-readable program code for a knowledge-based system including an inference engine and a knowledge base, and processing circuitry configured to access the memory, and execute the code. The apparatus is caused to at least receive a time series of measurements of operating conditions of a machine recorded during an operation of the machine. The apparatus is also caused to cluster the time series of measurements into one or more respective clusters and identify a pattern across the clusters. The apparatus is also caused to define a current state of the machine that includes the pattern across the clusters, access and search a knowledge base for a historical case describing a respective solution to a historical problem state similar to the current state, the respective solution including a repair action, and generate an output display indicating the repair action to address the current state.

Automated prediction of repair based on sensor data

A method, apparatus, system, and computer program product for managing a platform. Sensor information for a platform health of the platform is received from a sensor system for the platform. The sensor information for the platform health of the platform is sent by a computer system into a machine learning model trained using historical sensor information indicating a historical platform health and historical context information corresponding to the historical sensor information in which the historical context information is for a set of operating conditions. A remaining useful life of a component in the platform is received by the computer system from the machine learning model.

Aircraft maintenance of line replaceable units

A method, apparatus, system, and computer program product for managing a platform. A computer system generates a training dataset comprising historical metric values from historical sensor information for a set of metrics for a part and historical maintenance events for the part. The computer system trains a machine learning model using the training dataset. The computer system determines different maintenance thresholds for maintenance parameters for a metric in the set of metrics for performing maintenance on the part using the machine learning model trained with the training dataset. The computer system selects maintenance thresholds for the maintenance parameters from the maintenance thresholds meeting an objective to form a maintenance plan. The maintenance plan is used to determine when a maintenance action is needed for the part.

Vehicle health management using a counterfactual machine learning model to optimize maintenance parameters

Methods and systems are provided for inferred information propagation for aircraft prognostics. The method includes receiving, by a processor, an original time-series of data points for a component as an input; preprocessing the input to divide the original time-series of data into subsets of data by applying a time-window over the original time-series of data points; and computing, by the processor, a Mutual Information (MI) value for each pair of variables within each subset of data. The method also includes constructing, by the processor, a sequence of relationship graphs using the computed MI values; clustering, by the processor, each relationship graph; and analyzing, by the processor, the time-ordered sequence of clustered relationship graphs to identify features in the component.

Methods and systems for inferred information propagation for aircraft prognostics

A method includes obtaining a plurality of data sets, where each data set of the plurality of data sets includes multivariate time series data for a respective sample period of a plurality of sample periods. The method also includes, for each data set of the plurality of data sets, determining recurrence data indicative of recurrent states in the data set and determining, based on the recurrence data, determinism values of a determinism metric and laminarity values of a laminarity metric. The method further includes determining that a particular data set of the plurality of data sets includes data representing an anomalous state based on a determinism-laminarity curve representing the particular data set, where the determinism-laminarity curve is based on the determinism values of the particular data set and the laminarity values of the particular data set. The method also includes generating output data indicating the anomalous state.

Detection of anomalous states in multivariate data

A method includes receiving input data including a plurality of feature vectors and labeling each feature vector based on a temporal proximity of the feature vector to occurrence of a fault. Feature vectors that are within a threshold temporal proximity to the occurrence of the fault are labeled with a first label value and other feature vectors are labeled with a second label value. The method includes determining, for each feature vector of a subset, a probability that the label associated with the feature vector is correct. The subset includes feature vectors having labels that indicate the first label value. The method includes reassigning labels of one or more feature vectors of the subset having a probability that fails to satisfy a probability threshold and, after reassigning the labels, training an aircraft fault prediction classifier using supervised training data including the plurality of feature vectors and the labels.

System and method for generating an aircraft fault prediction classifier

A method, system, and computer program product for predicting abnormal operation of at least one component of a machine is provided. Real time monitoring data from an operating machine is received and monitoring features that are informative of likely abnormal operation are extracted and/or calculated. The monitoring features are applied to a prediction matrix that outputs probabilities of abnormal operation within one or more prediction time horizons. If the output probabilities exceed a threshold probability, then an alert can be output. Maintenance can be automatically scheduled in response to the alert.

Monitored machine performance as a maintenance predictor

Described is a system for producing indicators and warnings of adversarial activities. The system receives multiple networks of transactional data from different sources. Each node of a network of transactional data represents an entity, and each edge represents a relation between entities. A worldview graph is generated by merging the multiple networks of transactional data. Suspicious subgraph regions related to an adversarial activity are identified in the worldview graph through activity detection. The suspicious subgraph regions are used to generate and transmit an alert of the adversarial activity.

Computational framework for modeling adversarial activities

Described is a system for learning and predicting key phrases. The system learns based on a dataset of historical forecasting questions, their associated time-series data for a quantity of interest, and associated keyword sets. The system learns the optimal policy of actions to take given the associated keyword sets and the optimal set of keywords which are predictive of the quantity of interest. Given a new forecasting question, the system extracts an initial keyword set from a new forecasting question, which are perturbed to generate an optimal predictive key-phrase set. Key-phrase time-series data are extracted for the optimal predictive key-phrase set, which are used to generate a forecast of future values for a value of interest. The forecast can be used for a variety of purposes, such as advertising online.

System and method for learning contextually aware predictive key phrases

A method for generating human-machine hybrid predictions of answers to forecasting problems includes: parsing text of an individual forecasting problem to identify keywords; generating machine models based on the keywords; scraping data sources based on the keywords to collect scraped data relevant to the individual forecasting problem; providing the scraped data to the machine models; receiving machine predictions of answers to the individual forecasting problem from the machine models based on the scraped data; providing, by the computer system via a user interface, the scraped data to human participants; receiving, by the computer system via the user interface, human predictions of answers to the individual forecasting problem from the human participants; aggregating the machine predictions with the human predictions to generate aggregated predictions; and generating and outputting a hybrid prediction based on the aggregated predictions.

System and method for human-machine hybrid prediction of events

An apparatus for detecting a fault state of an aircraft is provided. The apparatus accesses a training set of flight data for the aircraft. The training set includes observations of the flight data, each observation of the flight data includes measurements of properties selected and transformed into a set of features. The apparatus builds a generative adversarial network including a generative model and a discriminative model using the training set and the set of features, and builds an anomaly detection model to predict the fault state of the aircraft. The anomaly detection model is trained using the training set of flight data, simulated flight data generated by the generative model, and a subset of features from the set of features. The apparatus deploys the anomaly detection model to predict the fault state of the aircraft using additional observations of the flight data.

Detecting fault states of an aircraft

A method includes obtaining sensor data captured by a sensor of an aircraft during a power up event. The sensor data includes multiple parameter values, each corresponding to a sample period. The method further includes determining a set of delta values, each indicating a difference between parameter values for consecutive sample periods of the sensor data. The method further includes determining a set of quantized delta values by assigning the delta values to quantization bins based on magnitudes of the delta values. The method further includes determining a normalized count of delta values for each quantization bin. The method further includes comparing the normalized counts of delta values to anomaly detection thresholds. The method further includes generating, based on the comparisons, output indicating whether the sensor data is indicative of an operational anomaly.

Anomaly prediction and detection for aircraft equipment

Systems and methods for fault prediction through a Bayesian framework are provided. Fault prediction for a valve system may be provided by generating a Bayesian framework by collecting a plurality of historical parameters related to opening and closing of a valve across a plurality of operational legs; generating a plurality of historical feature metrics based on the plurality of historical parameters; in response to detecting a fault, defining a prefault state corresponding to the historical feature metrics; monitoring a plurality of operational parameters related to opening and closing of the valve during a given operational phase of an operational leg; generating a plurality of operational feature metrics based on the plurality of operational parameters monitored during the given operational phase; and in response to determining, using the generated Bayesian framework, that the operational feature metrics indicate the prefault state of the subsystem, generating a notification.

Fault prediction in valve systems through Bayesian framework

A method for computing a human-machine hybrid ensemble prediction includes: receiving an individual forecasting question (IFP); classifying the IFP into one of a plurality of canonical question topics; identifying machine models associated with the canonical question topic; for each of the machine models: receiving, from one of a plurality of human participants: a first task input including a selection of sets of training data; a second task input including selections of portions of the selected sets of training data; and a third task input including model parameters to configure the machine model; training the machine model in accordance with the first, second, and third task inputs; and computing a machine model forecast based on the trained machine model; computing an aggregated forecast from machine model forecasts computed by the machine models; and sending an alert in response to determining that the aggregated forecast satisfies a threshold condition.

Systems and methods for forecast alerts with programmable human-machine hybrid ensemble learning

Growing community of inventors

Thousand Oaks, CA, United States of America

Tsai-Ching Lu