Industrial Process Subsystem Synchronization Using Predictive Set Points
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
In large-scale industrial automation systems, such as industrial mining operations, disparate process subsystems often run asynchronously, leading to inefficiencies and the need for real-time and historical data synchronization to ensure optimal performance, which is challenging due to their distributed nature.
Innovation Solution
A computing system utilizes machine learning models to generate process duration predictions and determine updated set points for interrelated subsystems, synchronizing them by aggregating and characterizing time-series data streams, and providing predictive and preventative maintenance insights to improve system integration and operational efficiency.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Adaptability or versatility
If process subsystems operate independently in distributed locations, then system flexibility and deployment ease are improved, but synchronization difficulty and system stability deteriorate
Solution Approach 1:
The system divides the industrial automation process into multiple independent process subsystems that can operate autonomously at different locations. Each subsystem maintains its own operational independence while being part of the larger distributed system, enabling flexible deployment without compromising individual subsystem functionality.
Solution Approach 2:
A feedback mechanism continuously monitors process duration across subsystems and dynamically adjusts set points to maintain synchronization. The system receives time-series data from all subsystems, processes this information through machine learning models, and provides corrective feedback to keep subsystems coordinated despite their distributed nature.
2Reliability
If real-time data synchronization is implemented across distributed subsystems, then system coordination is improved, but data processing complexity and computational requirements worsen
Solution Approach 1:
The system performs preliminary actions by continuously collecting and storing time-series data from all subsystems before synchronization is needed. Machine learning models are pre-trained on historical data to predict process durations, so when synchronization is required, the system can quickly retrieve and use pre-computed predictions rather than performing complex real-time analysis.
Solution Approach 2:
A central computing system acts as an intermediary that receives time-series data from multiple distributed subsystems, processes this data through machine learning models to generate process duration predictions, and then distributes updated set points back to the appropriate subsystems. This intermediary layer simplifies the synchronization complexity by centralizing the data processing functions.
3Productivity
If machine learning models are used for dynamic process duration prediction, then operational efficiency is improved, but computational resource requirements and system complexity worsen
Solution Approach 1:
The machine learning models operate autonomously to generate process duration predictions without requiring manual intervention. The system self-adjusts by continuously learning from incoming time-series data and automatically updating its predictions, enabling operational efficiency improvements while minimizing the need for complex external computational resources.
Solution Approach 2:
The system dynamically changes operational parameters by adjusting set points based on machine learning predictions. Instead of using fixed parameters, the system continuously adapts parameters like process duration and set point values based on real-time data analysis, improving operational efficiency while the complexity is managed through automated parameter adjustment rather than manual configuration.
Data Source
AI summary
Techniques to facilitate synchronization of industrial assets in an industrial automation environment are disclosed herein. In at least one implementation, a computing system receives time-series industrial process data associated with a plurality of process subsystems of an industrial automation process. The time-series industrial process data is fed into a machine learning model associated with the industrial automation process to dynamically generate a process duration prediction for a first one of the process subsystems and responsively determine an updated set point for a second one of the process subsystems based on the process duration prediction for the first one of the process subsystems. The updated set point for the second one of the process subsystems is provided to an industrial controller associated with the second one of the process subsystems.