Facility Camera Monitoring for Automatic Work Flow Identification
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing technologies for monitoring the running state of facilities with multiple machines are inefficient as they require manual input of shutdown causes and cannot automatically identify work unrelated to machine operation, such as cleaning or parts replacement, leading to inaccurate and incomplete data.
Innovation Solution
A computer system equipped with cameras to generate facility operation data, a data obtaining module, a work identification module, and an output module that processes images to automatically identify work performed and associate it with relevant machines, providing precise monitoring of facility operations without user burden.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Extent of automation
If manual input method is used to record shutdown causes and work activities, then the system can collect facility operation data, but the user burden increases and input accuracy decreases
Solution Approach 1:
The patent replaces the manual mechanical input system with an automated optical recognition system. Cameras capture images of work activities, and image recognition algorithms automatically identify and classify the work being performed, eliminating the need for manual data entry while maintaining high accuracy through automated analysis
Solution Approach 2:
The system performs self-identification of work activities by automatically capturing images and analyzing them to determine what work is being performed. The facility monitors itself without requiring human intervention for data collection, automatically generating work activity records from captured images
2Adaptability or versatility
If only machine operation state is monitored, then the monitoring system is simple, but work activities unrelated to machine operation cannot be identified
Solution Approach 1:
The monitoring system is designed to perform multiple functions: it monitors both machine operation states and various work activities (cleaning, inspection, parts replacement, etc.). The same camera system and image recognition platform handle diverse work types, making the system versatile without requiring separate specialized systems for each work category
Solution Approach 2:
The patent segments work activities into distinct categories (cleaning, inspection, parts replacement, etc.) and uses image recognition to identify each type. By dividing the monitoring task into recognizable work categories, the system can comprehensively track all facility activities while maintaining manageable system complexity through structured classification
3Ease of operation
If manual data entry is required for shutdown causes, then data collection is possible, but the burden on users increases and errors occur
Solution Approach 1:
The patent replaces manual data entry operations with automated image capture and recognition systems. Cameras continuously capture work activities, and automated algorithms extract and record the necessary information, eliminating manual typing and form-filling operations while preventing human errors in data entry
Solution Approach 2:
The system creates visual copies of work activities through image capture and uses these copies to automatically generate accurate records. Instead of relying on manual transcription, the system copies the actual visual evidence of work being performed and automatically converts it into structured data, ensuring accuracy
Data Source
AI summary
A facility that includes machines, includes a camera generating facility operation data, which includes an image of any given space. A computer system comprises: a data obtaining module obtaining the facility operation data; a work identification module identifying work that is performed in the facility, based on the image included in the facility operation data; and an output module outputting time-series information indicating a flow of the work. The work identification module identifies an object included in the image; identifies the work based on information about the identified object; and generates work analysis data, which associates the identified work, a period in which the identified work has been performed, and machines that is related to the identified work with one another. The output module outputs the time-series information or statistical information by using the work analysis data.