Dual Computing Architecture for Field Devices
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
The high complexity and energy requirements of using artificial neural networks in field devices for process automation have made their technical feasibility unattainable, limiting their application in determining process measurement variables efficiently.
Innovation Solution
A field device is designed with a dual computing architecture, comprising a deterministic computing device and an AI module, allowing tasks to be distributed between them, optimizing performance and energy consumption by assigning computing operations based on energy efficiency, and utilizing an AI module for probability-based and classification-based operations.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Adaptability or versatility
If an artificial neural network is used in field devices for computing operations, then the capability to perform complex computing operations is improved, but the energy consumption and device complexity increase significantly
Solution Approach 1:
The patent divides the computing arrangement into two separate computing devices: a first computing device for deterministic computing operations and a second computing device with AI modules for probability-based computing operations. This segmentation allows the system to leverage AI capabilities when needed while avoiding continuous AI processing that would consume excessive energy, thus resolving the contradiction between enhanced computing capability and reduced energy consumption.
2Adaptability or versatility
If an artificial neural network is used in field devices, then the capability to perform complex computing operations is improved, but the device complexity increases
Solution Approach 1:
The computing arrangement is segmented into two distinct computing devices with specialized functions. The first computing device handles deterministic operations while the second computing device handles AI-based probability operations. This segmentation allows the system to incorporate AI capabilities without requiring the entire device to be redesigned for AI processing, thereby managing device complexity while maintaining enhanced computing capability.
Solution Approach 2:
The patent implements a universal computing architecture that can handle both deterministic and probability-based computing operations. The second computing device with AI modules serves multiple purposes: it can perform neural network computations, support various AI algorithms, and integrate with the first computing device for hybrid operations. This multi-functionality approach allows the system to achieve high adaptability without proportionally increasing device complexity.
3Use of energy by moving object
If computing operations are distributed between two computing devices, then energy efficiency is improved, but the device complexity increases
Solution Approach 1:
The computing tasks are segmented and assigned to appropriate computing devices based on their nature. Deterministic operations are handled by the first computing device while probability-based operations are handled by the second computing device with AI modules. This segmentation enables energy-efficient processing by matching task requirements with device capabilities, and the modular architecture manages complexity through clear functional separation.
Data Source
AI summary
A field device is provided for detecting a process measurement variable, which is configured as a level measuring device for detecting a fill level of a medium, the field device including: a sensor arrangement to detect a measurement signal correlating with the process measurement variable; and a computing arrangement to determine a measurement value of the process measurement variable based on the measurement signal, the computing arrangement including at least one first computing device and at least one second computing device, the first computing device being configured to perform at least one deterministic computing operation, and the second computing device including at least one artificial intelligence module, and being configured to perform at least one probability-based and/or classification-based computing operation. A method of operating the field device is also provided.