Multitasking RPA Agent with Dynamic Job Switching
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing RPA systems operate in a sequential execution mode, leading to inefficiencies in processing speed, especially in web service monitoring where RPA needs to wait for result pages to load, significantly increasing overall job execution time.
Innovation Solution
A system and method for executing multitasking RPA that includes a scheduler, an RPA job queue, and an RPA agent with a multitasking manager. The multitasking manager instructs the RPA engine to perform jobs through multitasking by switching to other jobs when delays occur in executed RPA jobs, thereby improving processing speed.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Speed
If RPA operates in sequential execution mode, then system simplicity is maintained, but job execution speed deteriorates due to waiting periods
Solution Approach 1:
The patent segments the RPA execution process into multiple independent task units that can be executed concurrently. The job is divided into sub-tasks that are managed separately by the multitasking manager, allowing parallel execution of multiple RPA agents on different user PCs, thereby improving overall execution speed without significantly complicating the core RPA engine.
Solution Approach 2:
The patent implements dynamic task switching and allocation mechanisms. The multitasking manager dynamically assigns tasks to available RPA agents based on current system state, and RPA agents can dynamically switch between tasks. This dynamic allocation allows the system to adapt to varying workloads and maintain high execution speed while managing complexity through flexible resource utilization.
2Loss of time
If RPA waits for result page display, then accuracy of web service monitoring is ensured, but time consumption increases significantly
Solution Approach 1:
The patent implements preliminary actions by pre-loading and caching commonly accessed web service data and result pages before they are actually needed. The system anticipates future monitoring needs and prepares data in advance, reducing waiting time while ensuring that accurate data is available when required for monitoring comparisons.
Solution Approach 2:
The patent employs feedback mechanisms where the system continuously monitors web service results and compares them against expected outcomes. This real-time feedback allows the system to detect anomalies immediately without requiring extended waiting periods, maintaining monitoring accuracy while reducing the time each monitoring cycle takes to complete.
3Productivity
If multiple RPA jobs are executed sequentially, then resource consumption is minimized, but overall productivity deteriorates
Solution Approach 1:
The patent makes RPA agents universal by enabling them to handle multiple types of tasks and web service monitoring functions. A single RPA agent can be dynamically assigned to different monitoring tasks, performing multiple functions rather than requiring dedicated agents for each specific job. This multi-functionality increases productivity by allowing parallel execution while managing resource consumption through flexible agent utilization.
Solution Approach 2:
The patent implements self-service mechanisms where RPA agents autonomously manage their own task queues, resource allocation, and execution scheduling. Agents can independently pick up new tasks, manage their own state, and coordinate with the multitasking manager without requiring excessive system resources for centralized control, thereby improving productivity while maintaining efficient resource consumption.
Data Source
AI summary
An apparatus for executing multitasking RPA includes a scheduler configured to receive registration of a schedule for executing RPA jobs at a predetermined time from a user, an RPA job queue configured to receive registration of RPA jobs to be performed on a user PC so that the scheduler executes a job scenario at a predetermined time, each of the RPA jobs including scripts to execute the job scenario, and an RPA agent configured to receive a signal reporting start of a job from the RPA job queue, request a job to be performed from the RPA job queue, and receive a job of a corresponding type when the requested job of the corresponding type is in the RPA job queue, wherein the RPA agent includes an RPA engine and a multitasking manager.