Industrial Fault Detection Employing Meta Ensemble Model Based on Contact Sensor Ultrasonic Signal.

Ultrasonic diagnostics is the earliest way to predict industrial faults. Usually, a contact microphone is employed for detection, but the recording will be contaminated with noise. In this paper, a dataset that contains 10 main faults of pipelines and motors is analyzed from which 30 different features in the time and frequency domains are extracted. Afterward, for dimensionality reduction, principal component analysis (PCA), linear discriminant analysis (LDA), and t-distributed stochastic neighbor embedding (t-SNE) are performed. In the subsequent phase, recursive feature elimination (RFE) is employed as a strategic method to analyze and select the most relevant features for the classifiers. Next, predictive models consisting of k-Nearest Neighbor (KNN), Logistic Regression (LR), Decision Tree (DT), Gaussian Naive Bayes (GNB), and Support Vector Machine (SVM) are employed. Then, in order to solve the classification problem, a stacking classifier based on a meta-classifier which combines multiple classification models is introduced. Furthermore, the k-fold cross-validation technique is employed to assess the effectiveness of the model in handling new data for the evaluation of experimental results in ultrasonic fault detection. With the proposed method, the accuracy is around 5% higher over five cross folds with the least amount of variation. The timing evaluation of the meta model on the 64 MHz Cortex M4 microcontroller unit (MCU) revealed an execution time of 11 ms, indicating it could be a promising solution for real-time monitoring.

Design of an Integrated Micro-Viscometer for Monitoring Engine Oil.

This paper proposes a novel integrated micro-viscometer for engine-oil monitoring. The final solution consists of a capacitive micromachined ultrasonic transducer (CMUT) and an application-specific integrated circuit (ASIC). The CMUT is used to generate and capture acoustic waves while immersed in engine oil. The low power transceiver ASIC is interfaced with the CMUT structure for actuation and reception. An integrated charge pump boosts the supply voltage from 3.3 to 22 V to generate the DC polarization voltage of the CMUT. The receiver has a power consumption of 72 µW with an input-referred noise current of 3.2pAHz and a bandwidth of 7 MHz. The CMUT array occupies an area of 3.5 × 1 mm, whereas the ASIC has a chip area of 1 × 1 mm. The system was tested using engine oils of different types and ages at different temperatures. Measurement results show a significant frequency shift due to the dynamic viscosity change that occurs as oil ages. A shift of -1.9 kHz/cP was measured, which corresponds to a shift of 33 Hz/mile. This work paves the way for high accuracy-integrated solutions for oil condition monitoring and is expected to play a significant role in a more economic and environmentally friendly usage of oil.

Dual-Level Capacitive Micromachined Uncooled Thermal Detector.

This paper presents a novel dual-level capacitive microcantilever-based thermal detector that is implemented in the commercial surface micromachined PolyMUMPs technology. The proposed design is implemented side-by-side with four different single-level designs to enable a design-to-design performance comparison. The dual-level design exhibits a rate of capacitance change per degree Celsius that is over three times higher than that of the single-level designs and has a base capacitance that is more than twice as large. These improvements are achieved because the dual-level architecture allows a 100% electrode-to-detector area, while single-level designs are shown to suffer from an inherent trade-off between sensitivity and base capacitance. In single-level designs, either the number of the bimorph beams or the capacitance electrode can be increased for a given sensor area. The former is needed for a longer effective length of the bimorph for higher thermomechanical sensitivity (i.e., larger tilting angels per degree Celsius), while the latter is desired to relax the read-out integrated-circuits requirements. This thermomechanical response-to-initial capacitance trade-off is mitigated by the dual-level design, which dedicates one structural layer to serve as the upper electrode of the detector, while the other layer contains as many bimorph beams as desired, independently of the former's area.

Hard-Baked Photoresist as a Sacrificial Layer for Sub-180 °C Surface Micromachining Processes.

This letter proposes a method for utilizing a positive photoresist, Shipley 1805, as a sacrificial layer for sub-180 °C fabrication process flows. In the proposed process, the sacrificial layer is etched at the end to release the structures using a relatively fast wet-etching technique employing resist remover and a critical point dryer (CPD). This technique allows high etching selectivity over a large number of materials, including silicon-based structural materials such as silicon-carbide, metals such as titanium and aluminum, and cured polymers. This selectivity, as well as the low processing thermal budget, introduces more flexibility in material selection for monolithic integration above complementary metal oxide semiconductor (CMOS) as well as flexible substrates.