EP-4736151-A2 - TRAINING OF ARTIFICIAL INTELLIGENCE BATTLEGROUND GUIDANCE SYSTEM

Abstract

A method of training an artificial intelligence (AI) system, comprising: (a) displaying an initial version of a battlefield environment visualization showing at least one asset and two or more threats; (b) receiving from a user a threat level for each of the two or more threats; (c) displaying a subsequent version of the battlefield environment visualization in which the two or more threats have a subsequent geospatial relationship with respect to the at least one asset; (d) receiving from the user an updated threat level for at least one of the two or more threats for the subsequent version of the battlefield environment visualization; and (e) reiterating steps c and d at least once to create a training data set of updated threat levels for changing subsequent geospatial relationships of threats relative to the at least one asset.

Inventors

• SNYDER, GLENN THOMAS

Assignees

• Red Six Aerospace Inc.

Dates

Publication Date

20260506

Application Date

20240627

Claims (20)

1. 1. A method for training an artificial intelligence (Al) battle guidance system, comprising: (a) displaying an initial version of a battlefield environment visualization showing at least one asset and two or more threats to said at least one asset, wherein said two or more threats have an initial geospatial relationship with respect to said at least one asset; (b) receiving from a user a threat level for each of said two or more threats; (c) displaying a subsequent version of said battlefield environment visualization in which said two or more threats have a subsequent geospatial relationship with respect to said at least one asset, said subsequent geospatial relationship being different from a previous geospatial relationship of said two or more threats relative to said at least one asset of a previous battlefield environment visualization; (d) receiving from said user an updated threat level for at least one of said two or more threats for said subsequent version of said battlefield environment visualization; and (e) reiterating steps c and d at least once to create a training data set of updated threat levels for changing subsequent geospatial relationships of threats relative to said at least one asset.
2. 2. The method of training Al battle guidance system of claim 1, further comprising: (f) training an Al system using said training data set.
3. 3. The method of training Al battle guidance system of claim 3, further comprising: (g) running simulations using said Al system after step (f), said simulation displaying a simulated battlefield environment visualization showing at least one simulated asset and two or more simulated threats to said at least one simulated asset, wherein said two or more simulated threats have an initial simulated geospatial relationship with respect to said at least one simulated asset; (h) receiving from a user a threat level for each of said two or more simulated threats; (i) displaying a subsequent version of said simulated battlefield environment visualization in which said two or more simulated threats have a subsequent simulated geospatial relationship with respect to said at least one simulated asset, said subsequent simulated geospatial relationship being different from a previous simulated geospatial relationship of said two or more simulated threats relative to said at least one simulated asset of a previous simulated battlefield environment visualization; (j) receiving from said user an updated threat level for at least one of said two or more simulated threats for said subsequent version of said simulated battlefield environment visualization; (k) reiterating steps i and j at least once to create a second training data set of updated threat levels for changing subsequent simulated geospatial relationships of simulated threats relative to said at least one simulated asset; and (l) training said Al system with said second training data set.
4. 4. The method of training Al battle guidance system of claim 1, wherein said threat level is a priority relative to other threats.
5. 5. The method of training Al battle guidance system of claim 1, wherein said threat level is a weighted threat level.
6. 6. The method of training Al battle guidance system of claim 1, wherein said at least one asset is an aircraft.
7. 7. The method of training Al battle guidance system of claim 1, wherein said aircraft is a real aircraft.
8. 8. The method of training Al battle guidance system of claim 1, wherein said at least one asset comprises a real aircraft and a virtual aircraft.
9. 9. The method of training Al battle guidance system of claim 1, wherein said threats comprises at least one of an enemy aircraft, antiaircraft projectiles, an enemy antiaircraft installation, terrain or water, electromagnetic interference, a jamming signal, an enemy radar signal, an enemy missile lock signal, a low fuel condition of said at least one asset, an operational problem with said at least one asset, or a weather condition.
10. 10. The method of training Al battle guidance system of claim 1, wherein said initial version of a battlefield environment visualization and subsequent versions of said battlefield environment visualizations are based on the operation of a real aircraft in response to virtual threats.
11. 11. The method of training Al battle guidance system of claim 1, wherein said at least one asset comprises a primary asset and at least one secondary asset.
12. 12. The method of training Al battle guidance system of claim 11, wherein said primary asset and said at least one secondary asset operate in coordination with each other.
13. 13. The method of training Al battle guidance system of claim 12, wherein said training data set is with respect to operating said primary asset in coordination with said at least one secondary asset.
14. 14. The method of training Al battle guidance system of claim 11, wherein said initial geospatial relationship and said subsequent geospatial relationships are with respect to each of said primary asset and said at least one secondary asset.
15. 15. The method of training Al battle guidance system of claim 1, wherein said user uses a user interface to identify threats on a display and assign a threat level to each of said threats.
16. 16. The method of training Al battle guidance system of claim 1, further comprising: (f) outputting said training data set for use by an Al system.
17. 17. A trained Al system of claim 2.
18. 18. The trained Al system of claim 17, wherein said trained Al system is trained to control an asset.
19. 19. The trained Al system of claim 18, wherein said asset is a real aircraft.
20. 20. The trained Al system of claim 17, wherein said trained Al system is trained to provide guidance in real time to a pilot controlling a real aircraft.

Description

TRAINING OF ARTIFICIAL INTELLIGENCE BATTLEGROUND GUIDANCE SYSTEM REFERENCE TO RELATED APPLICATIONS [0001] This application is based on US Provisional Application No. 63/523,729, filed June 27, 2023, which is hereby incorporated by reference in its entirety. FIELD OF INVENTION [0002] The present invention relates, generally, to an artificial intelligence (Al) battleground guidance system, and, more specific, to training and Al battleground guidance system. BACKGROUND OF INVENTION [0003] A battleground environment involves complex enemy, friendly, and non- combatant interactions which are charged with emotion and difficult to operate in. Applicant recognizes that artificial intelligence (Al) will likely play a crucial role in helping personnel navigate these complex interactions. Therefore, Applicant has identified a need for Al-based battleground guidance systems. The present invention fulfills this need among others. SUMMARY OF INVENTION [0004] The following presents a simplified summary of the invention in order to provide a basic understanding of some aspects of the invention. This summary is not an extensive overview of the invention. It is not intended to identify key/critical elements of the invention or to delineate the scope of the invention. Its sole purpose is to present some concepts of the invention in a simplified form as a prelude to the more detailed description that is presented later. [0005] Applicant recognizes that its ability to dynamically map real and virtual objects as described herein can be used, not only to train pilots, but also to train Al to mimic expert pilots either in controlling real assets or providing guidance for pilots in real time or in training. [0006] One aspect of the present invention is a method for training an artificial intelligence (Al) battle guidance system. In one embodiment, the method comprises (a) displaying an initial version of a battlefield environment visualization showing at least one asset and two or more threats to the at least one asset, wherein the two or more threats have an initial geospatial relationship with respect to the at least one asset; (b) receiving from a user a threat level for each of the two or more threats; (c) displaying a subsequent version of the battlefield environment visualization in which the two or more threats have a subsequent geospatial relationship with respect to the at least one asset, the subsequent geospatial relationship being different from a previous geospatial relationship of the two or more threats relative to the at least one asset of a previous battlefield environment visualization; (d) receiving from the user an updated threat level for at least one of the two or more threats for the subsequent version of the battlefield environment visualization; and (e) reiterating steps c and d at least once to create a training data set of updated threat levels for changing subsequent geospatial relationships of threats relative to the at least one asset. [0007] Another aspect of the present invention is a non-transitory computer-readable storage medium for effecting the method described above. In one embodiment, the computer- readable storage medium including instructions that when executed by a computer, cause the computer to: (a) display an initial version of a battlefield environment visualization showing at least one asset and two or more threats to the at least one asset, wherein the two or more threats have an initial geospatial relationship with respect to the at least one asset; (b) receive from a user a threat level for each of the two or more threats; (c) display a subsequent version of the battlefield environment visualization in which the two or more threats have a subsequent geospatial relationship with respect to the at least one asset, the subsequent geospatial relationship being different from a previous geospatial relationship of the two or more threats relative to the at least one asset of a previous battlefield environment visualization; (d) receive from the user an updated threat level for at least one of the two or more threats for the subsequent version of the battlefield environment visualization; and (e) reiterate steps c and d at least once to create a training data set of updated threat levels for changing subsequent geospatial relationships of threats relative to the at least one asset. [0008] Yet another aspect of the invention is a user interface for creating a training data set for training an Al system. In one embodiment, the user interface comprises: (a) a display; (b) a user input device; and (c) a processor configured to perform the following steps: (1) display an initial version of a battlefield environment visualization showing at least one asset and two or more threats to the at least one asset, wherein the two or more threats have an initial geospatial relationship with respect to the at least one asset; (2) through the user input device, receive from a user a threat level for e