On-Board Data Processor for Multi-Gas Sensing
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Metal oxide semiconductor (MOS) gas sensors face limitations in selectivity and power consumption, particularly when processing multi-gas sensing data, which hinders their adoption in applications like industrial safety, asset monitoring, and wearable systems due to high power consumption and slow communication with external computing systems.
Innovation Solution
Incorporating an on-board, low-power data processor that uses multivariable gas classification and quantitation models, leveraging ambient and contextual data to select less complex models from a library, reducing computational complexity and power consumption by performing dielectric excitation responses analysis at fewer operating temperatures.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If measurements are processed by an external computing system, then processing power and accuracy are improved, but power consumption and communication reliability deteriorate
Solution Approach 1:
The patent segments the data processing function by implementing a low-power processor integrated within the gas sensor device itself, separate from the external computing system. This on-board processor handles preliminary data analysis and model selection locally, reducing the data transmission burden and power consumption associated with continuous external processing while maintaining accuracy through selective external verification when needed.
Solution Approach 2:
The patent introduces an intermediary low-power processor between the gas sensor and external computing system. This intermediary component performs essential data processing, model selection, and preliminary analysis locally, acting as a bridge that reduces communication overhead and power consumption while preserving measurement precision through coordinated operation with external systems.
2Measurement precision
If complex multivariate analysis models are used, then gas resolution accuracy is improved, but computational complexity and power consumption increase
Solution Approach 1:
The patent implements dynamic model selection where the processor adapts its computational complexity based on the specific gas analysis requirements. The system selects from multiple pre-trained models of varying complexity, using simpler models for common gas types and reserving complex multivariate analysis for cases requiring higher precision, thus balancing accuracy needs with computational resource consumption.
Solution Approach 2:
The patent changes the parameter of model complexity dynamically by selecting different analysis depth levels based on the sensing task. The system adjusts computational parameters such as the number of variables considered, model depth, and processing granularity to match the specific gas resolution requirements, reducing unnecessary computational complexity while maintaining sufficient accuracy.
3Measurement precision
If more operating temperatures are used for analysis, then gas classification accuracy is improved, but power consumption increases
Solution Approach 1:
The patent applies partial action by using a limited subset of operating temperatures sufficient for accurate gas classification rather than exhaustively testing all possible temperatures. The system identifies and uses only the necessary temperature points required for the specific gas detection task, avoiding the excessive action of processing all potential temperature conditions and thereby reducing power consumption while maintaining adequate accuracy.
Applied Scientific Principles
This section explains which scientific principles are used to turn an abstract innovation direction into a practical engineering solution.
Function Achieved in This Case
This approach enables accurate gas classification and concentration determination with significantly reduced power consumption, potentially less than what is required for RF communication, thereby enhancing the operational efficiency and suitability of MOS-based gas sensors for diverse applications.
Implementation Method 1
Metal oxide semiconductor (MOS) sensors can be operated as chemiresistors... a change in resistance of the MOS sensing element is measured, and this change in resistance is proportional to the gas concentrations in a fluid sample
Implementation Method 2
a measurement circuit operatively coupled to the gas sensing element and configured to provide dielectric excitation to, and to measure dielectric excitation responses of, the gas sensing element
Data Source
AI summary
Gas sensors are disclosed having an on-board, low-power data processor that uses multivariable gas classification and/or gas quantitation models to perform on-board data processing to resolve two or more gases in a fluid sample. To reduce computational complexity, the gas sensor utilizes low-power-consumption multivariable data analysis algorithms, inputs from available on-board sensors of ambient conditions, inputs representing contextual data, and/or excitation responses of a gas sensing material to select suitable gas classification and/or gas quantitation models. The data processor can then utilize these gas classification and quantitation models, in combination with measured dielectric responses of a gas sensing material of the gas sensor, to determine classifications and/or concentrations of two or more gases in a fluid sample, while consuming substantially less power than would be consumed if a global comprehensive model were used instead. Thus, the data processor is utilized for linear, nonlinear, and non-monotonic multivariate regressions.