Edge Prediction Models for Reduced Industrial Cloud Data Transmission
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Data transmission from industrial automation devices to cloud systems is costly and limited by bandwidth, and is prone to interruptions, necessitating improved data reduction techniques.
Innovation Solution
Implementing a process data model on both edge devices and cloud systems to generate estimated process data, comparing these with real data for deviations, and transmitting only when deviations exceed a threshold, thereby reducing data transmission.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If real process data are continuously transmitted from edge devices to cloud systems, then data accuracy and system reliability are improved, but data transmission costs and bandwidth requirements increase significantly
Solution Approach 1:
The patent creates a digital twin (process data model) that replicates the behavior and state of the physical process on the edge device. This virtual copy generates estimated process data locally, eliminating the need to continuously transmit actual sensor data to the cloud. The digital twin model is trained offline using historical data and then deployed to run autonomously at the edge, providing accurate process estimates without continuous data transmission.
Solution Approach 2:
The process data model is trained in advance using historical process data before deployment. This preliminary training phase allows the model to learn patterns and relationships in the data, enabling it to accurately predict current process states without needing real-time data transmission. The model is prepared beforehand to handle edge cases and anomalies, ensuring reliable operation during actual deployment.
2Speed
If data transmission rate is increased to meet bandwidth requirements, then real-time monitoring capability is improved, but transmission costs and network load increase
Solution Approach 1:
The patent extracts only the essential information needed for monitoring by using a digital twin to predict process states. Instead of transmitting all raw sensor data at high rates, the system extracts key process variables through the model's estimation capabilities. This selective information extraction maintains monitoring effectiveness while dramatically reducing transmission volume and associated costs.
Solution Approach 2:
The edge device performs self-service by running the process data model locally to generate its own process estimates. This autonomous operation at the edge eliminates dependence on continuous cloud communication for basic monitoring functions. The device serves its own information needs through local computation, reserving cloud transmission only for model updates and anomaly reporting.
3Quantity of substance
If data transmission is reduced to lower costs, then transmission expenses are decreased, but system response time to anomalies may increase
Solution Approach 1:
The system implements feedback by continuously comparing actual sensor measurements with model-predicted values at the edge device. When deviations exceed a threshold, indicating potential anomalies or model drift, the system immediately triggers data transmission to the cloud. This feedback mechanism ensures rapid response to critical events while maintaining reduced transmission volumes during normal operation.
Solution Approach 2:
The digital twin model proactively identifies and corrects deviations from expected behavior before they become critical failures. By continuously simulating process behavior and comparing with actual measurements, the system prevents anomalies from developing. This preliminary protective action reduces the need for emergency data transmission and cloud intervention, maintaining both cost efficiency and rapid response capability.
Data Source
AI summary
A method for providing process data of a device in an industrial automation environment to a computer system. In one embodiment, the method includes the following steps: executing a process data model on the device for generating estimated process data; determining that the estimated process data deviates from the real process data by more than a threshold value; and only if the estimated process data deviates from the real process data by more than the threshold value: transmitting information representing the real process data from the device to the computer system.