In-line Sensor Research Articles

Mach-Zehnder interferometer fiber optic sensor coated with UiO-66 metal organic framework for creatinine detection

An in-line Mach-Zehnder fiber optic sensor based is proposed and fabricated to measure creatinine (CR) concentration in the physiological range for clinical trials. The shift in valleys wavelength is monitored to assess the sensor response to the variation of CR concentration. Single-mode fiber as the sensing part is coated with UiO-66 metal–organic framework (MOF) adsorbent by dip coating to enhance the performance of the sensor. For further improvement, the length of the SMF and the coating time of the UiO-66 are optimized. The sensitivity and limit of detection (LOD) of the optimized bare sensor are 1530 pm/(mg/dL) and 0.4 mg/dL, respectively. The sensor with optimized UiO-66 coating thickness has a sensitivity of 2400 pm/(mg/dL) and an LOD of 0.2 mg/dL indicating the high performance. The sensors offer fast response, high sensitivity, and proper linear response at low concentrations of CR which reveal their potential in bio and chemical applications.

3D Printed, Single-Use Bioreactor with Integrated Inline Sensors for Microbial and Mammalian Cell Cultivation—A Case Study

The development of upstream bioprocesses necessitates small, instrumented bioreactors for investigating and optimizing production processes in a cost-effective manner. Due to advances in both the equipment and the materials used in additive manufacturing, 3D printing of customized bioreactors is now in the realm of possibilities. In this study, a small-scale 3D printed bioreactor suitable for mammalian and microbial cultivations was developed, featuring a working volume of 90 mL, inline pH and dissolved oxygen probes and a levitating magnetic stirrer. Aeration channels and a sampling port were printed directly into the vessel walls. Additionally, the vessel was equipped with a 3D printed customizable optical biomass-sensor. The bioreactor’s performance was evaluated through technical characterization and proof of concept cultivations, demonstrating that mixing time and oxygen mass transfer were sufficient for cultivating mammalian as well as microbial cells at high cell densities. Specifically, an Escherichia coli fed-batch cultivation achieved a maximum OD600 of 204. Furthermore, a fed-batch cultivation of an IgG antibody-producing Chinese hamster ovary cell line reached a peak viable cell density of 10.2 × 106 cells mL−1 and a maximum product titer of 2.75 g L−1. Using a three-parameter fit, the inline biomass signal could be correlated to the corresponding offline values with satisfactory accuracy, making it possible to monitor cell growth in real-time.

Simultaneous Pipe Leak Detection and Localization Using Attention-Based Deep Learning Autoencoder

Water distribution networks are often susceptible to pipeline leaks caused by mechanical damages, natural hazards, corrosion, and other factors. This paper focuses on the detection of leaks in water distribution networks (WDN) using a data-driven approach based on machine learning. A hybrid autoencoder neural network (AE) is developed, which utilizes unsupervised learning to address the issue of unbalanced data (as anomalies are rare events). The AE consists of a 3DCNN encoder, a ConvLSTM decoder, and a ConvLSTM future predictor, making the anomaly detection robust. Additionally, spatial and temporal attention mechanisms are employed to enhance leak localization. The AE first learns the expected behavior and subsequently detects leaks by identifying deviations from this expected behavior. To evaluate the performance of the proposed method, the Water Network Tool for Resilience (WNTR) simulator is utilized to generate water pressure and flow rate data in a water supply network. Various conditions, such as fluctuating water demands, data noise, and the presence of leaks, are considered using the pressure-driven demand (PDD) method. Datasets with and without pipe leaks are obtained, where the AE is trained using the dataset without leaks and tested using the dataset with simulated pipe leaks. The results, based on a benchmark WDN and a confusion matrix analysis, demonstrate that the proposed method successfully identifies leaks in 96% of cases and a false positive rate of 4% compared to two baselines: a multichannel CNN encoder with LSTM decoder (MC-CNN-LSTM) and a random forest and model based on supervised learning with a false positive rate of 8% and 15%, respectively. Furthermore, a real case study demonstrates the applicability of the developed model for leak detection in the operational conditions of water supply networks using inline sensor data.

Expanding access to CAR T cell therapies through local manufacturing.

Chimeric antigen receptor (CAR) T cells are changing the therapeutic landscape for hematological malignancies. To date, all six CAR T cell products approved by the US Food and Drug Administration (FDA) are autologous and centrally manufactured. As the numbers of approved products and indications continue to grow, new strategies to increase cell-manufacturing capacity are urgently needed to ensure patient access. Distributed manufacturing at the point of care or at other local manufacturing sites would go a long way toward meeting the rising demand. To ensure successful implementation, it is imperative to harness novel technologies to achieve uniform product quality across geographically dispersed facilities. This includes the use of automated cell-production systems, in-line sensors and process simulation for enhanced quality control and efficient supply chain management. A comprehensive effort to understand the critical quality attributes of CAR T cells would enable better definition of widely attainable release criteria. To supplement oversight by national regulatory agencies, we recommend expansion of the role of accreditation bodies. Moreover, regulatory standards may need to be amended to accommodate the unique characteristics of distributed manufacturing models.

In-line open-cavity Mach-Zehnder interferometric refractive-index sensors based on inter-mode and inter-core-mode interferences

We experimentally demonstrated in-line Mach-Zehnder interferometric sensors based on inter-mode interference (IMI) and inter-core-mode interference (ICMI) for refractive index (RI) measurements. The structure of the IMI-based sensor consisted of a C-shaped fiber spliced between two multi-mode fibers (MMF) with the opposite ends spliced to lead-in and lead-out single-mode optical fibers. The ICMI-based sensor has a similar structure as the IMI-based sensor but with an additional in-house fabricated three-core multi-core fiber (MCF) spliced between the MMF and C-shaped fiber. The performance of the two RI sensors was evaluated and compared using water-glycerol solutions with RI values ranging from 1.333 to 1.338. The RI sensitivity, as well as other characteristic properties such as free spectral range and extinction ratio in relation to three lengths of the C-shaped fiber of 2, 2.5 and 3 mm, were investigated in both structures. The open-cavity characteristic of the C-shaped fiber allows the solution to interact with the optical path of the sensors, hence improving their RI measurement sensitivities. Both types of sensors exhibit linear wavelength responses with variation in RI. The RI sensitivity of the IMI-based sensor exhibits minor variations when the length of the C-shaped fiber is altered, with a maximum sensitivity of -7999.76 nm/RIU. In contrast, the RI sensitivity of ICMI-based sensor exhibits variability when changing the C-shaped fiber length, potentially due to variations in the RI distribution among the cores within the MCF, with its maximum sensitivity being -6211.98 nm/RIU. The temperature sensitivity of all the tested sensors was around 0.7 nm/ °C, denoting a low temperature cross-sensitivity of -7.427 × 10−5 RIU/ °C. This work experimentally exhibits a comparison of two sensing effects in the same configuration. The long length (longer than 2 mm) of the sensing element facilitates the fabrication process, and the open-cavity present in the proposed sensors yields a high sensitivity, which is in the same order of magnitude as the IMI-based sensors and is one order of magnitude higher than the ICMI-based sensors, emphasizing their promising candidacy for label-free optical sensing of chemical and biological samples.

All-Fiber In-Line Twist Sensor Based on a Capillary Optical Fiber

Twist sensors have emerged as crucial tools in the field of structural health monitoring, playing a significant role in monitoring and ensuring the integrity of critical infrastructure such as dams, tunnels, bridges, pipelines, and buildings. We proposed and demonstrated an all-fiber in-line twist sensor which was based on a capillary fiber spliced between two single-mode fibers with a transverse offset. Through a series of experiments, the sensor’s performance was evaluated and quantified. The results showcased remarkable twist sensitivities in both clockwise and anticlockwise directions. With a transverse offset of 8.0 µm, the sensor exhibited twist sensitivities of −0.077 dB/° and 0.043 dB/° in the clockwise and anticlockwise directions, respectively, in the measured twist range from 0 to 90°. Furthermore, it was also demonstrated that the sensor was temperature insensitive at the chosen wavelength of 1520 nm, which can assist in increasing measurement accuracy. Our sensor’s low cost, simplicity of manufacture, and improved performance will push forward its adoption in future engineering applications such as structural health monitoring in dams, tunnels, and buildings.

Investigation of a Finite-Difference-Method based real-time viscous heating compensation in a nozzle viscometer for inline viscosity measurement of phenol resins

This work is motivated by the rarely available material data for thermosets and a missing inline viscosity measurement method to monitor melt property deviations. A nozzle viscometer is designed for inline viscosity measurement of phenol-formaldehyde compounds. The nozzle viscometer is mounted at the plasticizing unit of an injection molding machine. A fluid coolant circuit allows a dynamic tempering to pause the transient crosslinking reaction. A 2D FDM model is implemented, to calculate the viscous heating within an injection cycle and to compensate the temperature rising effect in the viscosity measurement. The FDM model processes the inline sensor signals in a closed loop correction in real time to determine a corrected reactive viscosity model. The signal-to-noise-ratio quantifies the reliability of the inline viscosity measurement method.

In-line temperature-compensated vector magnetic field sensor with side-polished fiber.

A novel, to the best of our knowledge, vector magnetic field sensor with temperature compensation is proposed and investigated. The proposed sensor is realized by side polishing a multi-mode optical fiber and adopting the surface plasmon resonance (SPR) effect. The side-polished surface is coated with a magnetic fluid (MF) and polydimethylsiloxane (PDMS) successively along the fiber axis. The as-fabricated sensor can be used not only for magnetic field strength and direction sensing, but also for temperature detection. The achieved magnetic field intensity sensitivities are 1720 pm/mT (90° direction) and -710 pm/mT (0° direction), and the temperature sensitivity is -2070 pm/°C. On top of its temperature compensation ability, the easy fabrication and very high sensitivity of the proposed sensor are attractive features for vector magnetic field sensing applications.

Portable Nanoparticle-Enhanced Sensor for Tigecycline Determination in Biological Fluids and in the Presence of Its Degradation Products

Tigecycline (TGC) is a novel potent antibiotic with recently proven anticancer activity against leukemia, glioma, and lung cancer. In-line TGC potentiometric sensors are fabricated for monitoring TGC in its pure form, pharmaceutical formulation, presence of its degradation products, and spiked human plasma. In-line sensors act as greener, portable, and economical alternatives to the classical off-line separation-based techniques. Classical and advanced liquid-contact (LC) and solid-contact (SC) sensors were fabricated, where the best performance was observed with the modified SC sensor (sensor VI) with potassium tetrakis (4-chlorophenyl) borate as ionic exchanger, β-cyclodextrin ionophore and cobalt oxide nanoparticles, showing a Nernstian response of 30 mV decade−1 in the linear range of 10−2–10−6 M. Statistical comparison was carried out for the results obtained from proposed SC sensors and the official method on TGC pure form. Additionally, method greenness was evaluated using a semi-quantitative analytical eco-scale, scoring approximately 95 points, which was the highest greenness achievement score when compared to the proposed LC sensors or British Pharmacopeial chromatographic method.

Formulation of a Mechanical Stress Dependent Macroscopic Magnetic Model for Incremental Permeability Simulation

Electromagnetic based systems receive ever-growing interest for on-line assessment of steel strip properties for major manufacturers, especially when Advanced High Strength Steel (AHSS) products for automotive purposes are considered. Manufacturing AHSS need optimal control of the thermo-mechanical processing using in-line sensors for real time monitoring. IMPOC, 3MA and PropertyMon are some of the well-known electromagnetic sensors. Identification of a mechanical or metallurgical state from a magnetic signal is of great interest, requiring however sophisticated models to mimic the real behaviour and its sensitivity to many influencing parameters. Our paper presents a proposition to add the dependency of mechanical stress to the Jiles-Atherton hysteresis model. Experimental data consist of B(H) curves of carbon steels strips subjected to different levels of mechanical stress applied along the magnetic measurement direction. The JilesAtherton model has then been used to find out a set of stress-dependent coefficients. The variation of coefficients with stress is in close relationship with the variations of coercive field, remanent induction and permeability with stress. A coupled magneto-mechanical approach allows physics-based functions to be proposed to describe the evolution of JilesAtherton coefficients with stress. In this paper, experimental working conditions will be described, as well as the stressdependent Jiles-Atherton model. The model is finally implemented in FEM software, allowing more complex magneto-mechanical situations to be modelled. After matching with measurement, the new model is applied on 3MA modelling where magnetic transient analysis and magnetic AC steady state analysis are linked through incremental permeability.

How the EU project “Online Microstructure Analytics” advances inline sensing of microstructure during steel manufacturing

Weight savings in mobility and transport are mandatory in order to reduce CO 2 emissions and energy consumption. The steel industry offers weight saving solutions by a growing portfolio of Advanced High Strength Steel (AHSS) products. AHSS owe their strength to their largely refined and complex microstructures, containing multiple metallurgical phases. Optimal control of the thermo-mechanical processing of AHSS requires inline sensors for real-time monitoring of evolution and consistency of microstructure and material properties. To coordinate and accelerate European development activities in this domain, the project ”Online Microstructure Analytics (OMA)” was established in 2019, constituting of a consortium of 14 specialised research organisations. The EU-funded OMA project, with a total budget exceeding 6 MEU, focuses on inline sensing techniques to monitor the consistency and homogeneity of microstructure, texture and mechanical properties for automotive steels and in particular for AHSS.

Maintaining and Assessing Various Tissue and Cell Types of the Eye Using a Novel Pumpless Fluidics System.

Many in vitro models used to investigate tissue function and cell biology require a flow of media to provide adequate oxygenation and optimal cell conditions required for the maintenance of function and viability. Toward this end, we have developed a multi-channel flow culture system to maintain tissue and cells in culture and continuously assess function and viability by either in-line sensors and/or collection of outflow fractions. The system combines 8-channel, continuous optical sensing of oxygen consumption rate with a built-in fraction collector to simultaneously measure production rates of metabolites and hormone secretion. Although it is able to maintain and assess a wide range of tissue and cell models, including islets, muscle, and hypothalamus, here we describe its operating principles and the experimental preparations/protocols that we have used to investigate bioenergetic regulation of isolated mouse retina, mouse retinal pigment epithelium (RPE)-choroid-sclera, and cultured human RPE cells. Innovations in the design of the system, such as pumpless fluid flow, have produced a greatly simplified operation of a multi-channel flow system. Videos and images are shown that illustrate how to assemble, prepare the instrument for an experiment, and load the different tissue/cell models into the perifusion chambers. In addition, guidelines for selecting conditions for protocol- and tissue-specific experiments are delineated and discussed, including setting the correct flow rate to tissue ratio to obtain consistent and stable culture conditions and accurate determinations of consumption and production rates. The combination of optimal tissue maintenance and real-time assessment of multiple parameters yields highly informative data sets that will have great utility for research in the physiology of the eye and drug discovery for the treatment of impaired vision.

Development of a Soft Sensor Using Machine Learning Algorithms for Predicting the Water Quality of an Onsite Wastewater Treatment System.

Developing advanced onsite wastewater treatment systems (OWTS) requires accurate and consistent water quality monitoring to evaluate treatment efficiency and ensure regulatory compliance. However, off-line parameters such as chemical oxygen demand (COD), total suspended solids (TSS), and Escherichia coli (E. coli) require sample collection and time-consuming laboratory analyses that do not provide real-time information of system performance or component failure. While real-time COD analyzers have emerged in recent years, they are not economically viable for onsite systems due to cost and chemical consumables. This study aimed to design and implement a real-time remote monitoring system for OWTS by developing several multi-input and single-output soft sensors. The soft sensor integrates data that can be obtained from well-established in-line sensors to accurately predict key water quality parameters, including COD, TSS, and E. coli concentrations. The temporal and spatial water quality data of an existing field-tested OWTS operated for almost two years (n = 56 data points) were used to evaluate the prediction performance of four machine learning algorithms. These algorithms, namely, partial least square regression (PLS), support vector regression (SVR), cubist regression (CUB), and quantile regression neural network (QRNN), were chosen as candidate algorithms for their prior application and effectiveness in wastewater treatment predictions. Water quality parameters that can be measured in-line, including turbidity, color, pH, NH4+, NO3-, and electrical conductivity, were selected as model inputs for predicting COD, TSS, and E. coli. The results revealed that the trained SVR model provided a statistically significant prediction for COD with a mean absolute percentage error (MAPE) of 14.5% and R2 of 0.96. The CUB model provided the optimal predictive performance for TSS, with a MAPE of 24.8% and R2 of 0.99. None of the models were able to achieve optimal prediction results for E. coli; however, the CUB model performed the best with a MAPE of 71.4% and R2 of 0.22. Given the large fluctuation in the concentrations of COD, TSS, and E. coli within the OWTS wastewater dataset, the proposed soft sensor models adequately predicted COD and TSS, while E. coli prediction was comparatively less accurate and requires further improvement. These results indicate that although water quality datasets for the OWTS are relatively small, machine learning-based soft sensors can provide useful predictive estimates of off-line parameters and provide real-time monitoring capabilities that can be used to make adjustments to OWTS operations.

How to design a low-cost pilot scale magnetic bioseparation process for protein separation from complex mixtures using in-line process analytics

Magnetic separation is a powerful tool for the integrated capture and purification of biomolecules and cells directly from complex feed media without the need for prior clarification and without the risk of equipment blockage. In this study, we equipped a commercially available rotor–stator high-gradient magnetic separator (RS-HGMS) with in-line analytics of hygienic design (absorbance, conductivity, pH, and flow) and used it for a systematic characterization of the equipment. By combining particle agglomeration and separation studies with residence time analysis, magnetic filter capacity, and particle retention experiments for bare iron oxide nanoparticles (BIONs), we developed a generic RS-HGMS operation protocol optimized for minimal particle carry-over and efficient buffer exchange. To demonstrate the versatility of our approach, we customized this operation protocol for the challenging task of lactoferrin separation from crude acid whey, which serves as an illustrative example in this paper. The modifications involved simple adjustments of pump speed, step duration, and cycle repetition without altering the fundamental principles of our generic approach. In this proof-of-concept, a BIONs carry-over of less than 0.037 g L-1 was achieved in the elution fractions, and the in-line sensors enabled continuous monitoring and documentation of the process, providing valuable data for process optimization and validation. This paper demonstrates how a magnetic separation process can be effortlessly tailored for challenging bioseparation tasks, while emphasizing the importance of implementing continuous in-line process monitoring and data logging. We believe this is an essential step toward the wider adoption of this promising technique in the industry.

Real-Time Quantification of Meat Paste Constituents Displaying Nonlinear Blending Behavior Including Salt Using an In-Line NIR MEMS Sensor

Rapid and robust quality monitoring of the composition of meat pastes is of fundamental importance in processing meat and sausage products. Here, an in-line near-infrared spectroscopy/micro-electro-mechanical-system-(MEMS)-based approach, combined with multivariate data analysis, was used for measuring the constituents fat, protein, water, and salt in meat pastes within a typical range of meat paste recipes. The meat pastes were spectroscopically characterized in-line with a novel process analyzer prototype. By integrating salt content in the calibration set, robust predictive PLSR models of high accuracy (R2 > 0.81) were obtained that take interfering matrix effects of the minor and NIR-inactive meat paste recipe component “salt” into account as well. The nonlinear blending behavior of salt concentration on the spectral features of meat pastes is discussed based on a designed mixture experiment with four systematically varied components.

On the Challenges of Applying Machine Learning in Mineral Processing and Extractive Metallurgy

The application of Machine Learning in Mineral Processing and Extractive Metallurgy has important benefits in terms of increasing the predictability and controllability of the processes, optimizing their performance, and improving maintenance. However, this application has significant implementation challenges. This paper analyzes these challenges and proposes ways of addressing them. Among the main identified challenges are data scarcity and the difficulty in characterizing abnormal events/conditions as well as modeling processes, which require the creative use of different learning paradigms as well as incorporating phenomenological models in the data analysis process, which can make the learning process more efficient. Other challenges are related to the need of developing reliable in-line sensors, adopting interoperability data models and tools, and implementing the continuous measurement of critical variables. Finally, the paper stresses the need for training of advanced human capital resources with the required skills to address these challenges.

Towards circular plastics: Density and MFR prediction of PE with IR spectroscopic techniques

The high variety of tailor fitted molecular structures of polyethylene (PE) is very beneficial to fulfill requirements of various applications, however it poses a difficulty in the mechanical recycling of post-consumer PE products. To improve the quality of PE recyclates and increase the amounts of recyclates that can be used in new products, separation of PE waste by density and melt flow rate (MFR) during mechanical sorting is essential. Therefore, 25 virgin PE grades were used to manufacture compression molded plates that were then characterized by means of Attenuated Total Reflection - Fourier transformed IR (ATR-FTIR) and near IR (NIR) spectroscopy, NIR hyperspectral imaging and dual-comb spectroscopy. The results were used to build partial least squares regression (PLS) models to predict MFR and density. ATR-FTIR and laboratory NIR spectroscopy provided sufficient information to predict the density value of PE, whereas the MFR assessments was not possible. The PLS model from the industrial NIR data also only allowed the density-based classification of virgin PE grades. The PLS models built from transmission and reflectance dual comb spectroscopy infrared (DCS-IR) of selected samples clearly showed that density and MFR prediction can be carried out with high accuracy. As DCS-IR could be implemented on plastic sorting systems using a conveyor belt, the addition of this sensor in mechanical sorting line would lead to a significantly higher quality of recycled PE with narrow well-defined density and MFR ranges. Such an improvement would immensely support the targeted recycling rates and amount by the European Union and would make a significant step towards circular plastics.

In-line small high-pressure sensors in anodically bonded microfluidic restrictors

In-line small high-pressure sensors in anodically bonded microfluidic restrictors

Continuous pH monitoring using a sensor for the early detection of anastomotic leaks.

Anastomotic leaks (AL) and staple line leaks are a serious post-operative complication that can develop following bariatric surgery. The delay in the onset of symptoms following a leak usually results in reactive diagnostics and treatment, leading to increased patient morbidity and mortality, and a clinical and economic burden on both the patient and the hospital. Despite support in literature for pH as a biomarker for early detection of AL, the current methods of pH detection require significant clinician involvement and resources. Presented here is a polyaniline (PANI)-based pH sensor that can be connected inline to surgical drains to continuously monitor peritoneal secretion in real time for homeostatic changes in pH. During this study, the baseline peritoneal fluid pH was measured in two pigs using the PANI sensor and verified using a benchtop pH probe. The PANI sensor was then utilized to continuously monitor the changes in the pH of peritoneal effluent, as a gastric leak was simulated. The inline sensors were able to detect the resulting local changes in drainage pH within 10 min of leak induction. The successful implementation of this sensor in clinical practice can both enable high efficiency continuous monitoring of patient status and drastically decrease the time required to detect AL, thus potentially decreasing the clinical and economic burden incurred by gastric leaks.

Ultrasonic Vibration Sensing of Intrinsic Fabry-Perot Interferometer Sensor Array Based on Local Spectral Analysis

This paper presents a low-cost integrated intrinsic Fabry-Perot interferometer (IFPI) sensor array demodulation algorithm based on local spectral analysis, which exploits multiple local spectra instead of the entire spectrum to achieve ultrasonic vibration measurement well beyond the sweeping period of the tunable laser source. By balancing the demodulation range and the demodulation resolution of IFPIs, this paper divides the entire spectrum produced by a distributed feedback (DFB) laser into 152 local spectra by a sliding window for IFPIs demodulation respectively, thereby implementing up to 160 kHz vibration measurement. To reduce the load on the data acquisition system and simplifies the interrogation system, this paper adopts a nonlinear tuning error compensation method based on an in-line IFPI sensor. In addition, chirp Z-transform is used in this paper to replace the traditional zero-padding-based fast Fourier transform to achieve the local fineness of demodulated optical path difference (OPD) spectrum, so as to reduce the waste of computation caused by zero-padding.