Process Monitoring KPI Filtering for APC Asset Availability
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current process monitoring systems for APC controllers and PWOs lack accuracy in identifying and categorizing data associated with non-availability of controllers, CVs, MVs, and DVs, leading to inaccurate KPI calculations and suboptimal process control.
Innovation Solution
Implementing specially-configured models to evaluate process data, automatically identify and categorize data as 'not required', and exclude it from KPI calculations, providing real-time insights into APC and PWO utilization.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If process data is collected without filtering out maintenance and off mode conditions, then data collection is simple and fast, but KPI calculation accuracy deteriorates
Solution Approach 1:
The system performs preliminary evaluation of process data to identify and categorize data associated with maintenance modes and off modes before KPI calculations are performed. This preliminary filtering action ensures that only relevant operational data is used in subsequent calculations, thereby improving accuracy without adding complexity to the core KPI computation logic.
Solution Approach 2:
The patent introduces an intermediary evaluation layer between raw data collection and KPI calculation. This intermediary component assesses each data sample to determine its operational mode status, acting as a mediator that separates relevant from irrelevant data. This approach maintains simple data collection while ensuring accurate KPI computation through the intermediary filtering mechanism.
2Measurement precision
If all process data samples are used for KPI calculations, then calculation speed is fast, but measurement accuracy deteriorates due to inclusion of irrelevant data
Solution Approach 1:
The system extracts and removes data samples associated with maintenance modes and off modes from the overall process data set before performing KPI calculations. By taking out these irrelevant data portions, the system ensures that only operational data contributes to performance metrics, significantly improving measurement precision while the automated extraction process minimizes additional processing time.
Solution Approach 2:
The patent applies partial action by selectively processing only the necessary portion of data samples - specifically evaluating each sample's operational mode status and including only relevant samples in KPI calculations. This partial processing approach avoids the time cost of processing all data while ensuring accuracy by excluding irrelevant maintenance and off mode data.
3Measurement precision
If manual evaluation of each data sample is performed to categorize asset conditions, then classification accuracy is high, but processing speed deteriorates
Solution Approach 1:
The system implements self-service automation where the evaluation process automatically assesses each data sample's operational mode status without requiring manual intervention. The automated evaluation uses predefined criteria to categorize data samples as operational, maintenance, or off mode, achieving high classification accuracy while maintaining high processing throughput through computer-based automated decision-making.
4Reliability
If process data associated with maintenance and off modes is included in KPI calculations, then data utilization is maximized, but reliability of performance metrics deteriorates
Solution Approach 1:
The system performs preliminary categorization of each data sample to identify its operational mode status before inclusion in KPI calculations. This preliminary action separates reliable operational data from unreliable maintenance and off mode data, ensuring that only high-quality information contributes to performance metrics. The categorization process itself preserves information about asset availability by tracking which samples were excluded and why.
Solution Approach 2:
The patent implements feedback mechanisms that track and report the proportion of excluded data samples associated with maintenance and off modes. This feedback provides valuable information about asset availability and operational patterns while ensuring that KPI calculations are based only on reliable operational data. The feedback loop maintains reliability by continuously monitoring data quality and adjusting evaluations accordingly.
Data Source
AI summary
Embodiments of the present disclosure provide improved process monitoring. Process data associated with a controller may be received. The process data may comprise a plurality of data samples. One or more process data evaluation iterations may be performed to generate one or more process data subsets by removing one or more portions of the process data that is associated with at least one asset condition of one or more asset conditions. One or more asset performance scores may be generated based on the one or more process data subsets. A user interface comprising the one or more asset performance scores may be caused to be rendered on a user device.