Distributed Process Control for Real-Time Cloud Analytics
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current cloud-based process control systems for industrial processes are not real-time capable due to latency issues in data acquisition, transmission, and analysis, which can lead to unacceptable delays and insecure communication, especially in critical industrial settings.
Innovation Solution
A distributed system comprising a local automation unit and a plant-external computing unit, where the local unit performs initial process control calculations and sends data to the external unit for more complex calculations, ensuring results are available within a predetermined sampling time to maintain real-time deterministic control, even if the cloud-based system experiences delays.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Adaptability or versatility
If cloud-based analytics are used for complex process control calculations, then analytical capability and control optimization are improved, but real-time responsiveness and deterministic control are worsened due to data transmission latency
Solution Approach 1:
The system segments control calculations into two parts: complex analytical calculations are performed in cloud-based analytics units, while time-critical control decisions are made by local automation units. This segmentation allows each component to handle tasks appropriate to its capabilities and location, resolving the contradiction between cloud-based analytical power and real-time responsiveness.
Solution Approach 2:
The local automation unit performs preliminary control actions based on locally available data and simple control logic before cloud-based analytics results are available. This preliminary action ensures that time-critical control decisions are not delayed by cloud communication latency, while still benefiting from cloud-based optimization when results are received.
2Productivity
If data is transmitted to external computing units for analysis, then comprehensive data processing is improved, but communication reliability and system security are worsened
Solution Approach 1:
A data transmission module acts as an intermediary between the local automation unit and cloud-based analytics, managing data exchange and implementing security measures. This intermediary layer protects the local control system while enabling comprehensive data processing in the cloud, resolving the contradiction between data processing capability and communication reliability.
Solution Approach 2:
The system implements different quality levels for different parts of the control architecture: local automation units maintain high reliability for time-critical control, while cloud-based analytics provide high processing capability for non-time-critical optimization. This local quality differentiation allows each component to excel at its specific function.
3Manufacturing precision
If complex control algorithms are implemented in higher-level systems, then control precision and optimization are improved, but system complexity and integration requirements are worsened
Solution Approach 1:
The control system is segmented into multiple levels with different algorithmic complexities: simple control algorithms run locally in automation units, while complex control algorithms run in higher-level control systems and cloud-based analytics. This segmentation allows complex algorithms to improve control precision without requiring every local unit to have equivalent computational complexity.
Solution Approach 2:
The control architecture implements a universal hierarchical structure where local automation units, higher-level control systems, and cloud-based analytics all operate within a standardized framework. This multi-functional hierarchical design allows complex algorithms to be deployed at appropriate levels without increasing overall system integration complexity, as each level follows the same architectural patterns.
Data Source
Figure 1
Figure 2
AI summary
Plant-side automation unit (2), wherein the plant-side automation unit (2) serves to read process input variables M, wherein the plant-side automation unit (2) serves to send the process input variables M to a plant-external computing unit for processing, wherein the plant-side automation unit serves to receive results (11) of a process control algorithm (7) from the plant-external computing unit, wherein the plant-side automation unit serves to check whether the results of the process control algorithm are available within a time t <tfreq oder t=tfreq empfangen wurden, wobei tfreq einer vorgegebenen Abtastzeit bzw. einer festgelegten Zykluszeit entspricht.