Artificial intelligence layer-based process extraction for robotic process automation

Patent metadata

Field	Value
Publication number	US-11836626-B2
Application number	US-202218051822-A
Country	US
Kind code	B2
Filing date	Nov 1, 2022
Priority date	Oct 15, 2019
Publication date	Dec 5, 2023
Grant date	Dec 5, 2023

Abstract

Artificial intelligence (AI) layer-based process extraction for robotic process automation (RPA) is disclosed. Data collected by RPA robots and/or other sources may be analyzed to identify patterns that can be used to suggest or automatically generate RPA workflows. These AI layers may be used to recognize patterns of user or business system processes contained therein. Each AI layer may “sense” different characteristics in the data and be used individually or in concert with other AI layers to suggest RPA workflows.

First claim

The invention claimed is:

1. A computer-implemented method for performing artificial intelligence (AI)-based process extraction for robotic process automation (RPA) using at least one AI layer, comprising:
running the data collected by listeners through the at least one AI layer, by the computing system, to process the collected data and identify one or more RPA processes in the collected data that exceed a confidence threshold; and
generating an RPA workflow for the identified RPA process that exceeds the confidence threshold, by the computing system.

2. The computer-implemented method of claim 1 , further comprising:
generating an automation for an RPA robot configured to implement the generated RPA workflow for the identified RPA process, by the computing system.

3. The computer-implemented method of claim 1 , wherein the data collected by the listeners comprises user interactions with respective user computing systems, audio, video, light, heat, motion, acceleration, radiation, or any combination thereof.

4. The computer-implemented method of claim 1 , wherein the at least one AI layer comprises a sequence extraction layer, a clustering detection later, a visual component detection layer, a text recognition layer, an audio-to-text translation layer, or any combination thereof.

5. The computer-implemented method of claim 1 , wherein each AI layer of the at least one AI layer has an associated modifier based on an estimated accuracy of the respective AI layer.

6. The computer-implemented method of claim 1 , wherein
the at least one AI layer comprises a plurality of AI layers, and
the RPA workflow is only generated when a collective confidence threshold for the plurality of AI layers has been exceeded.

7. The computer-implemented method of claim 1 , wherein the at least one AI layer is configured to perform statistical modeling and utilize deep learning techniques to identify the one or more RPA processes in the collected data that exceed the confidence threshold.

8. The computer-implemented method of claim 1 , further comprising:
identifying a similar existing RPA process to the identified RPA process that exceeds the confidence threshold, by the computing system;
determining that the existing RPA process works less optimally than the identified RPA process, by the computing system; and
replacing the existing RPA process with the identified RPA process by replacing an automation for the existing RPA process with an automation for the identified RPA process, by the computing system.

9. The computer-implemented method of claim 8 , wherein similarity between the existing process and the identified RPA process is determined by entropy, minimization of a process detection objective function, or a combination thereof.

10. The computer-implemented method of claim 1 , wherein
the at least one AI layer comprises a plurality of AI layers, and
the collected data is run through the plurality of AI layers in series.

11. The computer-implemented method of claim 1 , wherein
the at least one AI layer comprises a plurality of AI layers, and
the collected data is run through the plurality of AI layers in parallel.

12. The computer-implemented method of claim 1 , wherein
the at least one AI layer comprises a plurality of AI layers, and
the collected data is fed through a combination of both series and parallel AI layers of the plurality of AI layers.

13. A non-transitory computer-readable medium storing a computer program, the computer program configured to cause at least one processor to:
run the data collected by a plurality of listeners through at least one artificial intelligence (AI) layer to process the collected data and identify one or more robotic process automation (RPA) processes in the collected data that exceed a confidence threshold; and
generate one or more respective RPA workflows for the one or more identified RPA processes that exceed the confidence threshold.

14. The non-transitory computer-readable medium of claim 13 , wherein the computer program is further configured to cause the at least one processor to:
generate one or more respective automations that are configured to execute the one or more respective generated RPA workflows for the one or more identified RPA processes.

15. The non-transitory computer-readable medium of claim 14 , wherein the computer program is further configured to cause the at least one processor to:
identify a similar existing RPA process to an identified RPA process of the one or more identified RPA processes that exceeds the confidence threshold;
determine that the existing RPA process works less optimally than the identified RPA process; and
replace the existing RPA process with the identified RPA process by replacing an existing automation for the existing RPA process with an automation for the identified RPA process.

16. The non-transitory computer-readable medium of claim 15 , wherein similarity between the existing RPA process and the identified RPA process is determined by entropy, minimization of a process detection objective function, or a combination thereof.

17. The non-transitory computer-readable medium of claim 13 , wherein the data collected by the plurality of listeners comprises user interactions with respective user computing systems, audio, video, light, heat, motion, acceleration, radiation, or any combination thereof.

18. The non-transitory computer-readable medium of claim 13 , wherein
the at least one AI layer comprises a plurality of AI layers, and
the collected data is run through the plurality of AI layers in series or the collected data is run through the plurality of AI layers in parallel.

19. The non-transitory computer-readable medium of claim 13 , wherein
the at least one AI layer comprises a plurality of AI layers, and
the collected data is fed through a combination of both series and parallel AI layers of the plurality of AI layers.

20. An apparatus, comprising:
memory storing computer program instructions for performing artificial intelligence (AI)-based process extraction for robotic process automation (RPA) using at least one AI layer; and
at least one processor communicably coupled to the memory and configured to execute the computer program instructions, wherein the computer program instructions are configured to cause the at least one processor to:
run the data collected by a plurality of listeners through the at least one AI layer to process the collected data and identify one or more RPA processes in the collected data that exceed a confidence threshold, and
generate an RPA workflow for an identified RPA process that exceeds the confidence threshold.

21. The apparatus of claim 20 , wherein the computer program instructions are further configured to cause the at least one processor to:
generate an automation that is configured to implement the generated RPA workflow for the identified RPA process;
identify a similar existing RPA process to the identified RPA process that exceeds the confidence threshold;
determine that the existing RPA process works less optimally than the identified RPA process; and
replace the existing RPA process with the identified RPA process by replacing an automation for the existing RPA process with the generated automation.

Assignees

• Uipath Inc

Inventors

• Singh Prabhdeep
• Berg Christian

Classifications

• G06N3/0442

  characterised by memory or gating, e.g. long short-term memory [LSTM] or gated recurrent units [GRU] · CPC title

• G06N3/0464

  Convolutional networks [CNN, ConvNet] · CPC title

• G06Q10/06315

  Needs-based resource requirements planning or analysis · CPC title

• G06Q10/06311 Primary

  Scheduling, planning or task assignment for a person or group · CPC title

• G06N20/00

  Machine learning · CPC title