Industrial AI Model Graphs for Explainable Process Control
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional enterprise systems face challenges in optimizing operations due to complex mathematical programming, dependency on intricate metadata models, and the overwhelming volume of data generated in industrial processes, which can lead to unachievable objectives and operational burdens.
Innovation Solution
A system comprising a processor and memory that executes computer-executable components, including a configuration component to construct graphical representations of industrial processes as models, a visualization component to present these models on a human-machine interface, and an AI component to monitor and adjust the models in response to operational changes, thereby optimizing process operations.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Adaptability or versatility
If conventional enterprise systems use complex mathematical programming and metadata models to optimize operations, then they can handle sophisticated optimization requirements, but the system complexity and difficulty of operation increase significantly
Solution Approach 1:
The patent introduces an enterprise model as an intermediary layer between sensors and higher-level systems. This enterprise model comprises multiple devices/models at different levels (enterprise level, facility level, asset level), where each level processes and filters data before passing it upward. The enterprise model acts as a mediator that transforms raw sensor data into meaningful information for higher-level systems, reducing the complexity burden on individual components while maintaining comprehensive optimization capability across the entire enterprise.
2Loss of information
If conventional systems process large volumes of sensor data through higher-level devices, then they can capture comprehensive process information, but the operational burden on higher-level devices increases significantly
Solution Approach 1:
The patent segments the data processing function across multiple hierarchical levels. Sensors at the asset level generate raw data, which is then processed by device models at the facility level, and finally aggregated by the enterprise model. Each segment handles a specific portion of the data processing task, transforming raw sensor data into progressively more refined information. This segmentation distributes the operational burden across the hierarchy rather than concentrating it at the higher level, while maintaining complete information flow through the system.
3Loss of information
If conventional enterprise systems use highly complex metadata models, then they can represent detailed process information, but the models become non-explainable and difficult to interpret
Solution Approach 1:
The patent applies local quality by creating explainable views at each hierarchical level. The enterprise model provides a high-level explanatory view for enterprise-wide optimization, while individual device models provide detailed explanatory views for their specific functions. Each level maintains explainability appropriate to its scope, with the enterprise model showing aggregate trends and the device models showing detailed operational parameters. This allows the system to represent detailed process information while maintaining explainability at each level through appropriately scoped models.
Data Source
Figure 1
Figure 2
Figure 3A
AI summary
Various systems and methods are presented regarding monitoring and controlling operation of a process. A visual representation of the process can be created based on a supermodel comprising models (representing one or more devices) and nodes (representing respective device variables and constraints). Further, the process can be represented by levels, wherein devices at each level can be self-aware and have onboard artificial intelligence, such that a device at any level can auto-configure itself in accordance with a requirement placed upon it. Field-level devices (IFLDs) can be smart devices which auto-configure based upon a requirement from a higher-level device. Accordingly, system awareness can be incorporated across all levels of the process enabling overall and device-specific optimization of the process. IFLDs can auto-configure to collect and transmit data in accordance with an instruction from a higher-level device, leading to efficient data collection, reduced data bandwidth/processing, and expedited system optimization.