AI Process Mining for RPA Workflow Extraction and Automation

Resolve Bottlenecks,
Find Innovative SolutionsGenerate Solutions

Solution Overview

Problem

Current robotic process automation (RPA) techniques struggle to effectively identify and automate repetitive tasks, as they rely on costly and time-consuming manual logging and review, which can be inaccurate.

Innovation Solution

An AI-based system that deploys listener applications on user computing systems to generate logs of user interactions, which are then analyzed by AI layers to identify potential RPA processes, automatically generating workflows and robots to automate these processes.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Measurement precision

If manual logging and review methods are used to identify RPA processes, then process identification can be achieved, but the process becomes costly and time-consuming

Engineering Contradiction:

Improveprocess identification accuracyVSAvoidtime to identify processes

Core Design Contradiction:

Measurement precisionVSLoss of time

Solution Approach 1:

The patent replaces the mechanical manual review process with an AI-based automated system that uses machine learning models to analyze user interaction data and identify RPA processes, thereby eliminating the time-consuming manual effort while maintaining or improving identification accuracy

Inventive Principle:

Principle #28Mechanics substitution (Replace mechanical system)

Solution Approach 2:

The system enables automated self-identification of RPA processes by having the AI model directly analyze logs and generate process identifications without requiring human reviewers, allowing the system to serve itself in the process discovery function

Inventive Principle:

Principle #25Self-service

2Measurement precision

If manual logging and review methods are used to identify RPA processes, then process identification can be achieved, but the cost increases significantly

Engineering Contradiction:

Improveprocess identification accuracyVSAvoidcost to identify processes

Core Design Contradiction:

Measurement precisionVSLoss of energy

Solution Approach 1:

The patent substitutes expensive manual expert review with an AI-based automated analysis system that processes user interaction logs, significantly reducing the cost of process identification while maintaining accurate identification of RPA processes

Inventive Principle:

Principle #28Mechanics substitution (Replace mechanical system)

Solution Approach 2:

The system uses cost-effective automated AI analysis instead of expensive manual expert resources, treating the process identification as an automated computational task rather than a resource-intensive human activity

Inventive Principle:

Principle #27Cheap short-living objects (Disposable)

3Productivity

If human reviewers analyze user actions to identify processes, then process identification can be performed, but accuracy decreases due to reviewer errors

Engineering Contradiction:

Improveprocess identification capabilityVSAvoidprocess identification accuracy

Core Design Contradiction:

ProductivityVSMeasurement precision

Solution Approach 1:

The patent replaces human reviewers with an AI-based automated analysis system that objectively processes user interaction data, eliminating human errors such as misidentification of applications or actions while maintaining high productivity in process discovery

Inventive Principle:

Principle #28Mechanics substitution (Replace mechanical system)

Solution Approach 2:

The system uses automated AI analysis that can consistently and accurately process user interaction logs without the variability and errors inherent in human review, providing reliable and repeatable process identification results

Inventive Principle:

Principle #23Feedback

4Adaptability or versatility

If comprehensive log data is collected from multiple user computing systems, then process discovery capability improves, but system complexity increases

Engineering Contradiction:

Improveprocess discovery capabilityVSAvoidsystem architecture complexity

Core Design Contradiction:

Adaptability or versatilityVSDevice complexity

Solution Approach 1:

The patent implements a universal AI-based analysis platform that can process log data from multiple different user computing systems and applications through a single standardized system, achieving high process discovery capability without proportionally increasing complexity

Inventive Principle:

Principle #6Universality (Multi-functionality)

Solution Approach 2:

The system uses an intermediary AI layer that standardizes and processes diverse log data from multiple sources, acting as a mediator between heterogeneous user systems and the process discovery functionality, thereby managing complexity while maintaining versatility

Inventive Principle:

Principle #24Intermediary (Mediator)

Data Source

PatentUS12337464B2Artificial intelligence-based process identification, extraction, and automation for robotic process automation

Publication Date: 2025.06.24 UIPATH INC

AI summary

Artificial intelligence (AI)-based process identification, extraction, and automation for robotic process automation (RPA) is disclosed. Listeners may be deployed to user computing systems to collect data pertaining to user actions. The data collected by the listeners may then be sent to one or more servers and be stored in a database. This data may be analyzed by AI layers to recognize patterns of user behavioral processes therein. These recognized processes may then be distilled into respective RPA workflows and deployed to automate the processes.