Aircraft maintenance event prediction using higher-level and lower-level system information

Patent metadata

Field	Value
Publication number	US-2019102957-A1
Application number	US-201715721533-A
Country	US
Kind code	A1
Filing date	Sep 29, 2017
Priority date	Sep 29, 2017
Publication date	Apr 4, 2019
Grant date	—

Abstract

A method and apparatus for maintaining an aircraft. Real-time event information indicating faults in systems on the aircraft and aircraft condition monitoring system data indicating conditions of the systems on the aircraft are stored during a plurality of legs of flights of the aircraft. A feature table comprising the real-time event information and the aircraft condition monitoring system data is built. Feature vectors are extracted from the feature table. A machine learning algorithm is applied to the extracted feature vectors to generate a predicted maintenance event message that identifies a predicted maintenance event. The predicted maintenance event message is used to perform a maintenance operation on the aircraft.

First claim

What is claimed is:

1 . A method of maintaining an aircraft, comprising:
storing higher-level condition information indicating conditions of systems on the aircraft during a plurality of legs of flights of the aircraft; storing lower-level condition information indicating the conditions of the systems on the aircraft during the plurality of legs of flights of the aircraft; using the higher-level condition information and the lower level condition information to generate a predicted maintenance event message that identifies a predicted maintenance event; and using the predicted maintenance event message to perform a maintenance operation on the aircraft.

2 . The method of claim 1 , wherein:
storing the higher-level condition information comprises storing the higher-level condition information generated during a plurality of legs of flights of a plurality of aircraft including the aircraft; and storing the lower-level condition information comprises storing the lower-level condition information generated during the plurality of legs of flights of the plurality of aircraft including the aircraft.

3 . The method of claim 1 , wherein:
the higher-level condition information comprises real-time event information indicating faults in the systems; and the lower-level condition information comprises aircraft condition monitoring system data.

4 . The method of claim 1 , wherein using the higher-level condition information and the lower-level condition information to generate the predicted maintenance event message comprises:
extracting relevant lower-level condition information from the lower-level condition information; and aligning the relevant lower-level condition information with the higher-level condition information to provide matched lower-level condition information.

5 . The method of claim 1 , wherein using the higher-level condition information and the lower-level condition information to generate the predicted maintenance event message comprises identifying informative features in the lower-level condition information that are informative for predicting the predicted maintenance event.

6 . The method of claim 1 , wherein using the higher-level condition information and the lower-level condition information to generate the predicted maintenance event message comprises building a feature table comprising the higher-level condition information and the lower-level condition information.

7 . The method of claim 6 further comprising extracting feature vectors from the feature table.

8 . The method of claim 7 , wherein extracting feature vectors from the feature table comprises applying a sliding window to the feature table to generate a set of past events and a set of future events relative to a plurality of times.

9 . The method of claim 7 further comprising applying a machine learning algorithm to the extracted feature vectors to generate the predicted maintenance event message.

10 . A method of maintaining an aircraft, comprising:
storing real-time event information indicating faults in systems on the aircraft during a plurality of legs of flights of the aircraft; storing aircraft condition monitoring system data indicating conditions of the systems on the aircraft during the plurality of legs of flights of the aircraft; building a feature table comprising the real-time event information and the aircraft condition monitoring system data; extracting feature vectors from the feature table; applying a machine learning algorithm to the extracted feature vectors to generate a predicted maintenance event message that identifies a predicted maintenance event; and using the predicted maintenance event message to perform a maintenance operation on the aircraft.

11 . The method of claim 10 , wherein:
storing the real-time event information comprises storing the real-time event information generated during a plurality of legs of flights of a plurality of aircraft including the aircraft; and storing the aircraft condition monitoring system data comprises storing the aircraft condition monitoring system data generated during the plurality of legs of flights of the plurality of aircraft including the aircraft.

12 . An apparatus, comprising:
a condition information storer configured to store higher-level condition information indicating conditions of systems on an aircraft during a plurality of legs of flights of the aircraft and lower-level condition information indicating the conditions of the systems on the aircraft during the plurality of legs of flights of the aircraft; a predicted maintenance event predictor configured to use the higher-level condition information and the lower level condition information to generate a predicted maintenance event message that identifies a predicted maintenance event; and a predicted maintenance event message deliverer configured to send the predicted maintenance event message to maintenance personnel for use by the maintenance personnel to perform a maintenance operation on the aircraft.

13 . The apparatus of claim 12 , wherein the condition information storer is configured to:
store the higher-level condition information generated during a plurality of legs of flights of a plurality of aircraft including the aircraft; and store the lower-level condition information generated during the plurality of legs of flights of the plurality of aircraft including the aircraft.

14 . The apparatus of claim 12 , wherein:
the higher-level condition information comprises real-time event information indicating faults in the systems; and the lower-level condition information comprises aircraft condition monitoring system data.

15 . The apparatus of claim 12 , wherein using the higher-level condition information and the lower-level condition information to generate the predicted maintenance event message comprises:
extracting relevant lower-level condition information from the lower-level condition information; and aligning the relevant lower-level condition information with the higher-level condition information to provide matched lower-level condition information.

16 . The apparatus of claim 12 , wherein using the higher-level condition information and the lower-level condition information to generate the predicted maintenance event message comprises identifying informative features in the lower-level condition information that are informative for predicting the predicted maintenance event.

17 . The apparatus of claim 12 , wherein using the higher-level condition information and the lower-level condition information to generate the predicted maintenance event message comprises building a feature table comprising the higher-level condition information and the lower-level condition information.

18 . The apparatus of claim 17 further comprising extracting feature vectors from the feature table.

19 . The apparatus of claim 18 , wherein extracting feature vectors from the feature table comprises applying a sliding window to the feature table to generate a set of past events and a set of future events relative to a plurality of times.

20 . The apparatus of claim 18 further comprising applying a machine learning algorithm to the extracted feature vectors to generate the predicted maintenance event message.

Assignees

• Boeing Co

Inventors

• Huber David J
• Waagen Alex N
• Jiang Qin
• Lu Tsai-Ching

Classifications

• B64D2045/0085

  Devices for aircraft health monitoring, e.g. monitoring flutter or vibration · CPC title

• B64F5/60

  Testing or inspecting aircraft components or systems · CPC title

• B64D45/00

  Aircraft indicators or protectors not otherwise provided for · CPC title

• B64F5/40

  Maintaining or repairing aircraft · CPC title

• G06Q10/20

  Administration of product repair or maintenance · CPC title