Distributed Data Analytics Networks for Real-Time Process Monitoring
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current process control systems face challenges in real-time analytics and fault detection due to limited controller memory, bandwidth, and processor capability, leading to inaccurate and delayed analytics results, especially with large data sets, and inability to handle streaming data, which affects plant safety, reliability, and efficiency.
Innovation Solution
Embedding distributed data engines within process control devices to perform real-time analytics, using data analytics networks for localized data processing and streaming, and employing frequency analysis techniques for early fault detection, enabling real-time monitoring and optimization across the entire process plant.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If centralized data processing is used in process control systems, then data analytics can be performed, but controller memory, bandwidth, and processor capability are exceeded, leading to inaccurate and delayed results
Solution Approach 1:
The patent divides the centralized data processing architecture into distributed data engines deployed across multiple process control devices. Each data engine performs local data collection, filtering, and preliminary analytics, segmenting the overall analytics workload from the central controller. This segmentation enables accurate analytics on large datasets without overwhelming single-point controller resources.
Solution Approach 2:
The patent introduces a new architectural dimension by deploying data engines at the device level rather than only at the controller level. This spatial redistribution of processing capabilities across the control network enables parallel analytics operations, transforming the single-point processing bottleneck into a distributed processing network that scales with plant complexity.
2Reliability
If real-time analytics are implemented in process control systems, then plant safety and reliability improve, but limited controller memory and bandwidth prevent handling of large data sets
Solution Approach 1:
The patent extracts data processing functions from the central controller and places them in distributed data engines at the device level. This extraction removes the burden of processing large volumes of real-time data from the controller, enabling real-time analytics on extensive datasets without exceeding controller memory and bandwidth limitations, thereby maintaining plant safety and reliability.
Solution Approach 2:
Each process control device runs its own data engine that autonomously performs local data collection, filtering, and analytics. This self-service capability enables devices to process their own data in real-time without consuming central controller resources, allowing the system to handle large data volumes while maintaining real-time safety monitoring.
3Productivity
If distributed data engines are deployed across process control devices, then real-time analytics capability improves, but system complexity increases
Solution Approach 1:
The patent creates a universal data engine architecture that can be deployed across different types of process control devices (controllers, I/O cards, field devices). This universal design performs multiple functions including data collection, filtering, analytics, and communication, reducing the need for device-specific implementations and simplifying system integration despite the distributed nature of the deployment.
Solution Approach 2:
The distributed data engines implement feedback mechanisms where each engine reports its processing status, data quality metrics, and analytics results to neighboring engines and the central controller. This feedback system enables automated coordination and load balancing across the distributed network, managing system complexity through self-regulation rather than centralized control.
4Loss of time
If data is processed centrally, then system management is simplified, but analytics results are delayed and inaccurate for large data sets
Solution Approach 1:
The distributed data engines perform preliminary data processing, filtering, and analytics locally before transmitting results to the central controller. This preliminary action at the source reduces the volume and complexity of data requiring central processing, eliminating analytics delays for large datasets while maintaining simplified central system management through standardized data interfaces.
Data Source
AI summary
A system for monitoring and analyzing data in a distributed process control system is provided. The system includes a user interface having a set of user controls for selecting and configuring data blocks to create a data diagram representing a data model. The data blocks are associated with data operations, such as data analytics functions, and may be configured by the user for particular instances of general blocks. The data blocks are interconnected by wires conveying outputs or inputs of the blocks, which may also connect data sources to the data blocks. The data sources may include on-line data (i.e., data streams) or off-line data (i.e., stored data) from the process control system. Additional user controls may be used to evaluate the data diagram or convert the data diagram from an off-line to an on-line version.