Hyperspectral Fluid Sensing for Repeatable Compound Concentration
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing near-infrared spectroscopy (NIRS) methods for sensing compound concentrations in fluids suffer from poor repeatability due to environmental noise, affecting chemometric glucose determination models.
Innovation Solution
A non-contact system using hyperspectral imaging and machine learning, specifically a convolutional neural network, to predict compound concentrations by capturing hyperspectral images, preprocessing spectral signals, and applying reference corrections, enabling precise compound concentration monitoring in bioreactor fluids.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If near-infrared spectroscopy (NIRS) is used for sensing compound concentrations, then sensing capability is provided, but measurement precision deteriorates due to environmental noise
Solution Approach 1:
The patent introduces a flow cell as an intermediary component between the NIRS sensor and the fluid sample. This flow cell provides a controlled, isolated environment for the spectral measurement, shielding the measurement process from environmental noise while maintaining the sensing capability. The flow cell acts as a mediator that protects the measurement system from external interference.
Solution Approach 2:
The patent captures multiple spectral copies of the sample by flowing the fluid through the flow cell and acquiring successive spectral measurements. By collecting multiple spectral copies under controlled conditions and processing them through chemometric models, the system improves measurement precision by averaging out random noise while preserving the compound concentration information.
2Reliability
If traditional contact-based sensing methods are used, then compound concentration can be monitored, but reliability deteriorates due to cross-contamination
Solution Approach 1:
The patent replaces traditional contact-based mechanical sensing methods with non-contact near-infrared spectroscopy. Instead of physically inserting probes or sensors into the fluid (which causes cross-contamination), the system uses optical radiation to sense compound concentrations through the flow cell, eliminating the mechanical contact that leads to contamination while maintaining monitoring capability.
3Measurement precision
If single-point spectral information is captured, then sensing operation is simplified, but measurement precision deteriorates
Solution Approach 1:
The patent implements continuous spectral information capture by flowing the fluid continuously through the flow cell and acquiring spectral data over time. This continuous action provides multiple spectral measurements that can be processed by chemometric models to improve precision, while the automated flow system manages the complexity of handling continuous spectral data.
Solution Approach 2:
The patent transitions from single-point spectral information to multi-dimensional spectral data by capturing spectra across different time points as fluid flows through the cell. This adds the time dimension to the spectral measurement, providing richer data for chemometric analysis and improving compound concentration determination precision while systematically managing the increased data complexity.
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
The system provides accurate and repeatable compound concentration predictions, avoiding cross-contamination and improving monitoring capabilities in bioreactor environments.
Implementation Method 1
near-infrared spectroscopy (NIRS) only provides a single point of spectral information of the sample
Data Source
AI summary
A non-contact system for the sensing the concentration of a compound includes a hyperspectral imaging device configured to capture a hyperspectral image of a fluid, a flow cell configured to enable the capturing of a hyperspectral image of a fluid, a process, and a memory. The memory includes instructions stored thereon which, when executed by the processor, cause the system to generate a hyperspectral image of the fluid in the flow cell, generate several spectral signals based on the hyperspectral image, provide the spectral signal as an input to a machine learning network, and predict by the machine learning network the concentration of a compound in a fluid.