Workflow Recommendation Engine for Intelligent RPA Automation
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional script automation in robotic process automation (RPA) is limited to sequential user action mimicking and lacks an intelligent automation experience, failing to provide a comprehensive solution for user activity automation.
Innovation Solution
A workflow recommendation assistant engine that uses image pattern matching to analyze user interface activities, identify matches with existing automations, and suggest or execute relevant robotic process automations (RPAs), leveraging machine learning and artificial intelligence for intelligent automation.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Extent of automation
If conventional script automation is used to automate user activity, then automation can be delivered through existing scripts in a managed framework, but the system is limited to one-to-one sequential user action mimicking and cannot provide an intelligent automation experience
Solution Approach 1:
The patent replaces conventional script-based mechanical automation with an AI-powered image recognition system. The workflow recommendation assistant engine uses machine learning models to analyze screenshots and identify user activities, substituting the rigid sequential script execution with intelligent, context-aware automation recommendations that can understand and adapt to varying user actions.
Solution Approach 2:
The system transforms the automation approach by changing from fixed script parameters to dynamic AI-generated recommendations. The image pattern matching engine continuously analyzes current screen states and generates adaptive workflow recommendations, allowing the automation system to respond to changing user activities rather than following predetermined sequential scripts.
2Extent of automation
If custom script development and maintenance is performed to achieve automation, then specific user activities can be automated, but the complexity and maintenance burden increases
Solution Approach 1:
The workflow recommendation assistant engine performs self-service by automatically analyzing user activities through image recognition and generating appropriate automation recommendations without requiring manual script development. The system autonomously captures screenshots, identifies user actions, matches them with available workflows, and presents recommendations, eliminating the need for users to write and maintain custom automation scripts.
Solution Approach 2:
The patent creates a universal automation system where a single AI-powered engine can handle diverse user activities across different applications and interfaces. Instead of requiring separate custom scripts for each automation task, the image pattern matching engine recognizes various user actions and recommends appropriate workflows from a shared repository, making the system multi-functional and reducing overall complexity.
3Ease of operation
If sequential user action mimicking is used for automation, then existing scripts can be leveraged for managed framework automation, but the system cannot provide intelligent automation experience
Solution Approach 1:
The patent replaces mechanical sequential action mimicking with AI-driven intelligent recognition. The workflow recommendation assistant engine uses machine learning to understand user intent and context from screenshots, generating intelligent automation recommendations rather than simply replicating sequential user actions. This substitution maintains ease of operation through automated analysis while adding intelligent adaptability.
Data Source
AI summary
Disclosed herein is a computing system. The computing system includes a memory and a processor. The memory stores processor executable instructions for a workflow recommendation assistant engine. The processor is coupled to the memory. The processor executes the workflow recommendation assistant engine to cause the computing device to analyze images of a user interface corresponding to user activity, execute a pattern matching of the images with respect to existing automations, and provide a prompt indicating that an existing automation matches the user activity.