Systems and methods for generating robotic process automation (RPA) via documentation

Patent metadata

Field	Value
Publication number	US-12554249-B1
Application number	US-202318103838-A
Country	US
Kind code	B1
Filing date	Jan 31, 2023
Priority date	Jan 31, 2022
Publication date	Feb 17, 2026
Grant date	Feb 17, 2026

Abstract

Documentation may be used to automatically create a set of scripts to be executed by a robotic process automation system. A machine learning engine may be used to identify patterns in a first documentation and to then derive a flowchart descriptive of a documented process, subprocess, or combination thereof. The flowchart is compiled into a script executable by the robotic process automation system.

First claim

The invention claimed is:

1 . A non-transitory computer-readable medium, comprising computer-executable instructions that, when executed by one or more processors, cause the one or more processors to:
identify, via a machine learning engine, patterns in a first documentation using a training set comprising a second documentation and one or more secondary flowcharts described in the second documentation, wherein the patterns are used to derive a flowchart descriptive of a documented process, of a documented subprocess, or a combination thereof; and compile the flowchart into a script, wherein the script is executable by a robotic process automation system.

2 . The non-transitory computer-readable medium of claim 1 , wherein the first documentation comprises:
text manuals describing the documented process, the documented subprocess, or the combination thereof; or audio describing the documented process, the documented subprocess, or the combination thereof; or video describing the documented process, the documented subprocess, or the combination thereof; or a combination thereof.

3 . The non-transitory computer-readable medium of claim 1 , wherein the computer-executable instructions cause the one or more processors to prepare the first documentation by identifying key words, phrases, logic elements, or a combination thereof, in the first documentation.

4 . The non-transitory computer-readable medium of claim 1 , comprising computer-executable instructions, that when executed by the one or more processors cause the one or more processors to:
identify the patterns in the first documentation, by:
identifying, in the first documentation, an enumerated list of actions; and
identifying, in the first documentation, objects that the enumerated list of actions are applied to; and
derive the flowchart from the enumerated list of actions and the objects that the enumerated list of actions are applied to.

5 . The non-transitory computer-readable medium of claim 4 , comprising computer-executable instructions, that when executed by the one or more processors cause the one or more processors to:
compile the flowchart into the script, by:
identifying script actions that correspond to the enumerated list of actions; and
inserting, into the script, the script actions, with an indication to apply the script actions to graphical user interface (GUI) objects corresponding to the objects that the enumerated list of actions are applied to in the first documentation.

6 . The non-transitory computer-readable medium of claim 1 , comprising computer-executable instructions, that when executed by the one or more processors cause the one or more processors to:
identify the patterns in the first documentation, by:
identifying, in the first documentation, illustrations representative of interactions with a graphical user interface (GUI); and
identifying, in the first documentation, objects that the interactions are applied to; and
derive the flowchart from the interactions and the objects.

7 . The non-transitory computer-readable medium of claim 6 , comprising computer-executable instructions, that when executed by the one or more processors cause the one or more processors to:
compile the flowchart into the script, by:
identifying script actions that correspond to the interactions; and
inserting, into the script, the script actions, with an indication to apply the script actions to GUI objects corresponding to the objects.

8 . A computer-implemented method, comprising:
identifying, via a machine learning engine, patterns in a first documentation; using a training set comprising a second documentation and one or more secondary flowcharts described in the second documentation to identify, via the machine learning engine, the patterns in documentation; deriving a flowchart descriptive of a documented process, of a documented subprocess, or a combination thereof; and compiling the flowchart into a script, wherein the script is executable by a robotic process automation system.

9 . The method of claim 8 , wherein the first documentation comprises:
text manuals describing the documented process, the documented subprocess, or the combination thereof; or audio describing the documented process, the documented subprocess, or the combination thereof; or video describing the documented process, the documented subprocess, or the combination thereof; or a combination thereof.

10 . The method of claim 8 , comprising preparing the first documentation by identifying key words, phrases, logic elements, or a combination thereof, in the first documentation.

11 . The method of claim 8 , comprising:
identifying the patterns in the first documentation, by:
identifying, in the first documentation, an enumerated list of actions;
identifying, in the first documentation, objects that the enumerated list of actions are applied to; and
deriving the flowchart from the enumerated list of actions and the objects that the enumerated list of actions are applied to.

12 . The method of claim 11 , comprising:
compiling the flowchart into the script, by:
identifying script actions that correspond to the enumerated list of actions; and
inserting, into the script, the script actions, with an indication to apply the script actions to graphical user interface (GUI) objects corresponding to the objects that the enumerated list of actions are applied to in the first documentation.

13 . The method of claim 8 , comprising:
identifying the patterns in the first documentation, by:
identifying, in the first documentation, illustrations representative of interactions with a graphical user interface (GUI);
identifying, in the first documentation, objects that the interactions are applied to; and
deriving the flowchart from the interactions and the objects.

14 . The method of claim 13 , comprising:
compiling the flowchart into the script, by:
identifying script actions that correspond to the interactions; and
inserting, into the script, the script actions, with an indication to apply the script actions to GUI objects corresponding to the objects.

15 . A robotic process automation (RPA) script generation system, comprising:
a natural language processor configured to receive a first documentation and parse the first documentation into computer-readable data; a machine learning engine configured to identify patterns in the computer-readable data using a training set comprising a second documentation and one or more secondary flowcharts described in the second documentation and derive a flowchart descriptive of a documented process, of a documented subprocess, or a combination thereof of the first documentation; and an RPA script generation engine configured to compile the flowchart into an RPA script, wherein the RPA script is executable by a robotic process automation system to automate subsequent playback of the RPA script in a graphical user interface (GUI).

16 . The RPA script generation system of claim 15 , wherein:
identifying the patterns in the first documentation, comprises:
identifying, in the first documentation, an enumerated list of actions; and
identifying, in the first documentation, action objects that the enumerated list of actions are applied to; or
identifying, in the first documentation, illustrations representative of interactions with the GUI; and
identifying, in the first documentation, illustration objects that the interactions are applied to; and
deriving the flowchart comprises:
deriving the flowchart from the enumerated list of ac

Assignees

• Usaa

Inventors

• Mangalagiri Pradeep R
• Bitter Dustin Bowen
• Meyer Gregory
• Russell Christopher
• Lopez Jeanie Graciela
• Lawrence Carol Lyn
• Czerlinsky Timothy Benjamin

Classifications

• G05B2219/50391

  Robot · CPC title

• G05B19/4155 Primary

  characterised by program execution, i.e. part program or machine function execution, e.g. selection of a program · CPC title