Visual Log Generation for Process Mining Bottlenecks
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional process mining techniques face challenges in capturing comprehensive insights into operation processes due to limitations in event logs, which only record completed states and fail to capture interactions between workers and devices, making it difficult to identify bottlenecks and improve processes efficiently.
Innovation Solution
A process mining system generates visual logs from video streams of worker devices, extracting activity labels from images and assigning case identifiers to provide a holistic view of operation processes, including interactions and handovers, thereby enhancing the analysis of operation processes.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Loss of information
If conventional event logs are used to record process actions, then the logging system remains simple and lightweight, but the logs only capture completed states and miss worker-device interactions, reducing analysis comprehensiveness
Solution Approach 1:
The patent introduces visual logs as an intermediary data source between worker devices and the process mining system. Video streams capture worker-device interactions that conventional event logs miss, while the visual log generator acts as a mediator to process these videos into structured data that complements traditional event logs, thereby improving information completeness without directly modifying the simplicity of existing logging systems
Solution Approach 2:
The system merges conventional event logs with visual logs to create a comprehensive process mining data foundation. Event logs provide structured action data while visual logs capture interaction details, and their integration through the process mining system enables complete process analysis while maintaining the simplicity of each individual logging mechanism
2Loss of information
If human operators directly observe workers to determine actions, then comprehensive process insights can be obtained, but this becomes time-consuming and infeasible for thousands or millions of worker devices
Solution Approach 1:
The system enables self-service process mining by automatically capturing and analyzing worker-device interactions through video streams. The visual log generator autonomously processes video data to extract process information without requiring human observation, allowing the system to scale to millions of worker devices while maintaining comprehensive process insights
Solution Approach 2:
The patent replaces the mechanical system of human observation with an automated computer vision system. Video processing algorithms automatically detect and analyze worker-device interactions, substituting human operators with computational methods that can process data from countless devices simultaneously, eliminating time constraints while maintaining insight completeness
3Loss of information
If video streams are processed to extract activity labels, then detailed interaction information is captured, but the processing complexity and computational resources increase
Solution Approach 1:
The visual log generator extracts only the essential interaction information from video streams needed for process mining analysis. By selectively extracting activity labels and relevant interaction details while discarding redundant visual data, the system captures comprehensive interaction information while managing processing complexity through focused data extraction
Solution Approach 2:
The system segments video processing into distinct functional components: video capture, frame analysis, activity label extraction, and log generation. This segmentation allows each component to be optimized independently and enables parallel processing of multiple video streams, reducing overall processing complexity while maintaining detailed interaction capture
Data Source
AI summary
A process mining system performs process mining using visual logs generated from video streams of worker devices. Specifically, for a given worker device, the process mining system obtains a series of images capturing a screen of a worker device while the worker device processes one or more tasks related to an operation process. The process mining system determines activity labels for a plurality of images. An activity label for an image may indicate an activity performed on the worker device when the image was captured. The activity label is determined by extracting information from pixels of the image and inferring the activity of the worker device from the extracted information. The process mining system generates event logs from the visual logs of worker devices and uses the event logs for process mining.