Industrial Data Quality Indicators for Reliable Automation Analytics
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Industrial automation systems face challenges in effectively managing and analyzing data quality from industrial devices, lacking comprehensive methods to assess and enhance the reliability and context of the data, which hampers predictive maintenance and operational efficiency.
Innovation Solution
A system comprising a processor that retrieves industrial data from an asset model, determines a quality indicator based on timestamp and control data, and generates an output representing data quality, utilizing machine learning for data analysis and enhancement.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Quantity of substance
If industrial data is collected from multiple devices without quality assessment, then data quantity increases, but data reliability deteriorates
Solution Approach 1:
The system performs preliminary quality assessment of industrial data before it is used for predictive maintenance or analytics. The quality assessment component evaluates data quality indicators such as completeness, validity, and timeliness before the data enters the analysis pipeline, preventing low-quality data from degrading subsequent processing results.
Solution Approach 2:
The system implements feedback mechanisms where quality assessment results are fed back into the data collection and processing pipeline. This feedback enables continuous improvement of data quality by identifying sources of poor quality data and adjusting collection parameters or filtering criteria accordingly.
2Loss of information
If comprehensive data analysis is performed without quality indicators, then analytical depth increases, but decision accuracy deteriorates
Solution Approach 1:
Quality indicators are calculated and attached to data points before comprehensive analysis begins. This preliminary quality tagging enables the analysis system to weight, filter, or prioritize data based on quality metrics, ensuring that deeper analytical insights are derived from high-quality data sources.
Solution Approach 2:
The system applies different quality assessment criteria and weighting to different data sources, devices, or data types based on their specific characteristics and reliability profiles. This localized quality management allows comprehensive analysis while maintaining decision accuracy by treating different data sources according to their individual quality characteristics.
3Reliability
If data quality assessment is implemented, then data reliability improves, but system complexity increases
Solution Approach 1:
The quality assessment system is divided into modular components that can be independently configured and deployed. Different quality indicators (completeness, validity, timeliness) are assessed separately and can be enabled or disabled based on specific needs, allowing incremental implementation that manages complexity while improving reliability.
4Loss of information
If timestamp data is collected concurrently with control data, then data context improves, but data processing complexity increases
Solution Approach 1:
Timestamp data and control data are merged into a unified data structure with associated quality indicators. This integration maintains the contextual relationship between time-stamped events and control actions while enabling centralized quality assessment that manages processing complexity through standardized handling of combined data elements.
Data Source
AI summary
Industrial automation data quality and analysis (e.g., using a computerized tool) is enabled. For example, a system can comprise: a memory that stores executable components, and a processor, operatively coupled to the memory, that executes the executable components, the executable components comprising: a device interface component that retrieves industrial data from an industrial device represented in an industrial asset model, wherein the industrial data comprises control data applicable to the industrial device and timestamp data generated by the industrial device concurrently with the control data, a quality component that, based on the control data and the timestamp data, determines a quality indicator applicable to the control data, and an output component that, based on the quality indicator, generates an output representative of a quality of the control data.