Sensor Tiles with Adaptive Weighting for Process Installation Reliability
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing sensor arrangements in process plants rely on central evaluation units for communication between sensors, which can lead to inefficiencies and failures if the central unit malfunctions, and do not effectively account for the position and status of sensor tiles in determining overall process information.
Innovation Solution
A sensor arrangement with multiple sensor tiles that communicate directly with each other, using MEMS components and simple, inexpensive sensors, where data exchange and weighting of values from each tile consider position, function, and accuracy to develop a swarm intelligence capable of functioning without a central control unit, allowing for adaptive weighting and self-management.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Device complexity
If a central evaluation unit is used for sensor communication, then the system structure is simplified and easier to manage, but the system becomes vulnerable to failures and less reliable when the central unit malfunctions
Solution Approach 1:
The patent divides the centralized sensor evaluation system into multiple independent sensor tiles, each with its own control and evaluation unit. This segmentation eliminates the single point of failure in centralized systems while distributing functionality across multiple autonomous units that can operate independently.
Solution Approach 2:
Each sensor tile is equipped with autonomous control and evaluation capabilities, allowing it to independently process and evaluate sensor data without relying on a central evaluation unit. This self-service capability ensures continuous operation even when other tiles fail.
2Device complexity
If sensor tiles communicate via a central evaluation unit, then the communication structure is simple, but complete process information cannot be obtained and communication efficiency is reduced
Solution Approach 1:
The patent implements a mesh network topology where sensor tiles communicate directly with each other in addition to potential central coordination. This merging of direct peer-to-peer communication paths with optional central coordination enables complete information exchange while maintaining communication efficiency.
Solution Approach 2:
Sensor tiles exchange data directly with neighboring tiles, creating multiple feedback loops for information verification and completion. This direct feedback mechanism ensures that complete process information is obtained through cross-validation between adjacent sensors.
3Ease of manufacture
If all sensor tiles use identical communication protocols, then the system is easier to implement, but adaptive weighting based on position and status cannot be achieved
Solution Approach 1:
The patent implements dynamic weighting mechanisms where each sensor tile's contribution to overall process information is adjusted based on its position, operational status, and data quality. This dynamic adaptability allows the system to optimize information fusion in real-time while maintaining a standardized communication protocol foundation.
Solution Approach 2:
Different sensor tiles are assigned different weighting factors based on their local characteristics such as position in the process flow, operational reliability, and measurement quality. This local quality differentiation enables adaptive information fusion while using uniform communication protocols.
4Measurement precision
If expensive complex sensors are used, then measurement precision is improved, but cost and device complexity increase
Solution Approach 1:
The patent combines multiple simple sensor measurements from adjacent tiles to achieve the measurement precision previously requiring complex single sensors. By fusing data from multiple independent simple sensors, the system achieves equivalent or superior measurement accuracy while reducing individual sensor complexity and cost.
Solution Approach 2:
Instead of using expensive complex sensors, the patent uses multiple copies of simple, inexpensive sensor tiles. The redundancy and data fusion from multiple copies compensate for individual sensor simplicity, achieving high measurement precision through quantity and coordination rather than individual complexity.
Data Source
Figure 1~2
AI summary
The invention relates to a sensor arrangement on a process installation comprising at least two sensor tiles, wherein each sensor tile comprises a support which can be arranged on the process installation and a plurality of sensors arranged on the support for determining a physical or chemical variable of a measuring medium found in the process installation, a process characteristic of the measuring medium and/or a state of the process installation, wherein a first of the sensor tiles comprises a control and/or evaluation unit having at least one transmit and receive module for data exchange with a first control and/or evaluation unit of a second sensor tile, wherein the first control and/or evaluation unit of the first sensor tile and/or a second control and/or evaluation unit allocated to the sensor arrangement is designed to determine the physical or chemical variable of the measuring medium, the process characteristic of the measuring medium and/or a state of the process installation by weighting the values determined by each sensor tile, wherein weighting is carried out as a function of the measured value variations of at least one sensor of a sensor tile, the position of the sensor tile in the process installation and/or the function of the sensor tile.