Asset Modification Verification via NLP Shift Report Analysis
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Facilities with continuous operations face challenges in identifying unauthorized modifications to assets, which can lead to accidents, operational deviations, and increased costs due to the difficulty in capturing such modifications in operation shift reports.
Innovation Solution
An Unauthorized Modification Identification System (UMIS) that utilizes a natural language processing model to analyze operation shift reports and correlation parameters to detect unauthorized modifications, initiating investigations or authorizing modifications based on correlation scores.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Productivity
If facilities run continuously for longer durations to increase production and reduce costs, then productivity and cost efficiency are improved, but the risk of unauthorized modifications to assets increases and becomes harder to detect
Solution Approach 1:
The system performs preliminary actions by capturing images of asset operational displays at the beginning of each operation shift and storing them as reference images. This preliminary capture enables later comparison to detect any unauthorized modifications that occur during continuous operation, resolving the contradiction between continuous operation and modification detection.
2Loss of information
If operation shift reports are manually maintained to track asset modifications, then information about modifications is captured, but unauthorized modifications can still be missed and investigation efficiency is reduced
Solution Approach 1:
The system replaces the manual mechanical process of maintaining operation shift reports with an automated image recognition system. The natural language processing model automatically analyzes images of operational displays, extracts modification information, and generates reports without human intervention. This substitution eliminates information capture gaps and reduces investigation time by providing automated, consistent analysis.
Solution Approach 2:
The system enables self-service by allowing the image analysis model to autonomously detect modifications, generate reports, and initiate investigations without requiring manual review. The automated system serves itself by continuously monitoring asset displays and independently identifying unauthorized changes, reducing both information loss and time loss.
3Measurement precision
If natural language processing models are used to analyze operation shift reports, then unauthorized modifications are detected more accurately, but system complexity increases
Solution Approach 1:
The system introduces an intermediary natural language processing model that acts as a mediator between the raw image data from asset displays and the modification detection requirement. This intermediary model automatically analyzes the images, extracts relevant information, and generates structured reports, thereby improving detection accuracy while managing system complexity through automated processing.
Data Source
Figure 1
Figure 2
Figure 3
AI summary
Techniques for identifying unauthorized modifications made to assets within a facility are described. In an example, an asset modification indication representative of a first modification made to a first asset within a facility is received, where the asset modification indication comprises a first asset identifier associated with the asset. An operation shift report for the facility is then received, where the operation shift report is indicative of modifications made to assets within the facility. The operation shift report is then analyzed to extract a first set of modification records, wherein each of the first set of modification records comprises an asset identifier corresponding to an asset and modifications made to the asset. Thereafter, it is determined that a modification record corresponding to the first asset is not present in the first set of modification records using at least one correlation parameter. Accordingly, an investigation with respect to the first modification is initiated.