Network digital twin of airline operations

Patent metadata

Field	Value
Publication number	US-12456080-B2
Application number	US-202117504732-A
Country	US
Kind code	B2
Filing date	Oct 19, 2021
Priority date	Oct 19, 2021
Publication date	Oct 28, 2025
Grant date	Oct 28, 2025

Abstract

Systems and methods of planning and scheduling resources of an airline include using machine learning models to generate custom behavioral characteristics for a plurality of operational components of the airline based on operations data. The custom behavioral characteristics are used to define inputs and outcomes in simulations of the operational components of the airline. The simulation outputs performance indicators and a feedback loop captures previously ignored interactions between the operational components of the airline based on the performance indicators and custom behavioral characteristics of inputs and outcomes corresponding to the plurality of operational components of the airline.

First claim

What is claimed is:

1. A method of identifying actions to improve a performance indicator related to airline operations in a system comprising a plurality of operational components, the method comprising:
receiving operations data corresponding to the plurality of operational components, wherein the operations data includes data corresponding to resources, and rules governing the usage of one or more of the resources;
training one or more machine learning models, using the received operations data, to generate custom behavioral characteristics corresponding to inputs or outcomes for each of the plurality of operational components, each custom behavioral characteristic corresponding to a variable used in simulating airline operations;
simulating operations of the airline using the generated one or more custom behavioral characteristics to determine values for the inputs of the plurality of operational components, wherein the simulated operations comprise a first scenario including one or more performance indicators corresponding to the first scenario;
generating a feedback loop where the one or more performance indicators are used as inputs to train the one or more machine learning models;
training the one or more machine learning models to identify interactions between the plurality of operational components based on the feedback loop;
refining the one or more performance indicators using the trained one or more machine learning models; and
displaying, using a user interface, the refined one or more performance indicators.

2. The method of claim 1 , wherein the operational components include a flight scheduling component, an aircraft route component, a crew scheduling component, a passenger itinerary component, a passenger welfare and compensation component, and a financial component.

3. The method of claim 1 , further comprising:
using the operations data and the one or more machine learning models to predict a disruption, wherein the disruption includes:
a violation of one or more of the rules, or
a performance indicator failing to meet a given threshold.

4. The method of claim 3 , further comprising
generate one or more alternative scenarios, wherein each of the one or more alternative scenarios includes a trade-off wherein at least one performance indicator is improved in comparison to the first scenario;
simulating the operations data, including the one or more performance indicators for each of the one or more alternative scenarios;
receiving selection of a scenario among the first scenario and the one or more alternative scenarios; and
scheduling usage of the resources corresponding to the plurality of operational components based on the selected scenario.

5. The method of claim 1 , wherein the one or more machine learning models include one or more of quantitative models, statistical models, simulation models, machine learning models, or artificial intelligence models.

6. A non-transitory computer-readable medium storing instructions that, when executed by a processor, cause the processor to perform steps comprising:
receiving operations data corresponding to a plurality of operational components, wherein the operations data includes data corresponding to resources, and rules governing the usage of one or more of the resources;
training one or more machine learning models, using the received operations data, to generate custom behavioral characteristics corresponding to inputs or outcomes for each of the plurality of operational components, each custom behavioral characteristic corresponding to a variable used in simulating airline operations;
simulating operations of the airline using the generated one or more custom behavioral characteristics to determine values for the inputs of the plurality of operational components, wherein the simulated operations comprise a first scenario including one or more performance indicators corresponding to the first scenario;
generating a feedback loop where the one or more performance indicators are used as inputs to train the one or more machine learning models;
training the one or more machine learning models to identify interactions between the plurality of operational components based on the feedback loop;
refining the one or more performance indicators using the trained one or more machine learning models; and
displaying, using a user interface, the refined one or more performance indicators.

7. The non-transitory computer-readable medium of claim 6 , storing further instructions that, when executed by a processor, cause the processor to further perform steps comprising:
receiving operations data from operational components corresponding to a flight scheduling component, an aircraft route component, a crew scheduling component, a passenger itinerary component, a passenger welfare and compensation component, and a financial component.

8. The non-transitory computer-readable medium of claim 6 , storing further instructions that, when executed by a processor, cause the processor to further perform steps comprising:
using the operations data and the one or more machine learning models to predict a disruption, wherein the disruption includes:
a violation of one or more of the rules, or
a performance indicator failing to meet a given threshold.

9. The non-transitory computer-readable medium of claim 8 , storing further instructions that, when executed by a processor, cause the processor to further perform steps comprising:
generating one or more alternative scenarios, wherein each of the one or more alternative scenarios includes a trade-off wherein at least one performance indicator is improved in comparison to the first scenario;
simulating the operations data, including the one or more performance indicators for each of the one or more alternative scenarios;
receiving selection of a scenario among the first scenario and the one or more alternative scenarios; and
scheduling usage of the resources corresponding to the plurality of operational components based on the selected scenario.

10. The non-transitory computer-readable medium of claim 6 , storing further instructions that, when executed by a processor, cause the processor to further perform steps comprising:
using one or more of quantitative models, statistical models, simulation models, machine learning models, or artificial Intelligence models to generate the custom behavioral characteristics corresponding to inputs or outcomes for each of the plurality of operational components.

11. A system for identifying actions to improve a performance indicator related to airline operations in a system comprising a plurality of operational components, the system comprising a processor configured to:
receive operations data corresponding to the plurality of operational components, wherein the operations data includes data corresponding to resources, and rules governing the usage of one or more of the resources;
train one or more machine learning models, using the received operations data, to generate custom behavioral characteristics corresponding to inputs or outcomes for each of the plurality of operational components, each custom behavioral characteristic corresponding to a variable used in simulating airline operations;
simulate operations of the airline using the generated one or more custom behavioral characteristics to determine values for the inputs of the plurality of operational components, wherein the simulated operations comprise a first scenario including one or more performance indicators corresponding to the first scenario;
generate a feedback loop where the one or more performance indicators are used as inputs to train the one or more machine learning

Assignees

• Ge Aviation Systems Llc

Inventors

• Arguello Michael
• Ephraim Charles P
• Dunsdon Jonathan Mark
• Davis Stephen Jonathan
• Schanche Kristin M

Classifications

• G06N20/00

  Machine learning · CPC title

• G06N3/08

  Learning methods · CPC title

• G06N5/01

  Dynamic search techniques; Heuristics; Dynamic trees; Branch-and-bound · CPC title

• G06N7/01

  Probabilistic graphical models, e.g. probabilistic networks · CPC title

• G06N20/10

  using kernel methods, e.g. support vector machines [SVM] · CPC title