Process Monitoring Techniques Research Articles

Evaluating early predictive performance of machine learning approaches for engineering change schedule – A case study using predictive process monitoring techniques

By applying machine learning algorithms, predictive business process monitoring (PBPM) techniques provide an opportunity to counteract undesired outcomes of processes. An especially complex variation of business processes is the engineering change (EC) process. Here, failing to adhere to planned implementation dates can have severe impacts on assembly lines, and it is paramount that potential negative cases are identified as early as possible. Current PBPM research, however, has seldomly investigated the predictive performance of machine learning approaches and their applicability at early process steps, let alone for the EC process. In our research, we show that given adequate feature encoding, shallow learners can accurately predict schedule adherence after process initialisation. Based on EC data from an automotive manufacturer, we provide a case sensitive performance overview on algorithm-encoding combinations. For that, three algorithms (XGBoost, Random Forest, LSTM) were combined with four encoding techniques. The encoding techniques used were the two common aggregation-based and index-based last state encoding, and two new combinations of these, which we term advanced aggregation-based and complex aggregation-based encoding. The study indicates that XGBoost-index-encoded approaches outclass regarding average predictive performance, whereas Random-Forest-aggregation-encoded approaches perform better regarding temporal stability due to reduced influence by dynamic features. Our research provides a case-based reasoning approach for deciding on which algorithm-encoding combination and evaluation metrics to apply. In doing so, we provide a blueprint for an early warning and monitoring method within the EC process and other similarly complex processes.

LS-ICE: A Load State Intercase Encoding framework for improved predictive monitoring of business processes

Research on developing techniques for predictive process monitoring has generally relied on feature encoding schemes that extract intra-case features from events to make predictions. In doing so, the processing of cases is assumed to be solely influenced by the attributes of the cases themselves. However, cases are not processed in isolation and can be influenced by the processing of other cases or, more generally, the state of the process under investigation. In this work, we propose the LS-ICE (load state intercase encoding) framework for encoding intercase features that enriches events with a depiction of the state of relevant load points in a business process. To assess the benefits of the intercase features generated using the LS-ICE framework, we compare the performance of predictive process monitoring models constructed using the encoded features against baseline models without these features. The models are evaluated for remaining trace and runtime prediction using five real-life event logs. Across the board, a consistent improvement in performance is noted for models that integrate intercase features encoded through the proposed framework, as opposed to baseline models that lack these encoded features.

Demonstration of temperature monitoring of the curing process in resin transfer moulding using integrated phase-shifted Bragg grating sensors

Composite materials have increased in use across multiple industries due to their low weight, thermal stability and design flexibility. As applications grow, the importance of accurate process-monitoring techniques grows, with various research teams investigating a variety of in situ cure-monitoring sensors. Photonic integrated circuits (PICs) can be a feasible solution in tracking curing cycles due to their resistance to harsh conditions and ease of integration into the production tool. This paper demonstrates the integration and performance evaluation of an embedded phase-shifted Bragg grating solution, instead of commonly used fiber sensors, based on 220 nm height silicon-on-insulator platform for in-situ monitoring of the RTM-6 resin curing process in a composite tool. The embedded optical sensor presents high sensitivity (75 pm/°C) and linearity (R2=0.995) up to 120 °C, enabling precise temperature measurements throughout the curing cycle. The experimental results show consistency with previously reported simulations, throughout the calibration, dry run characterization and the curing process, up to 120 °C, allowing for accurate monitoring of the curing process.

Statistical process monitoring creates a hemodynamic trajectory map after pediatric cardiac surgery: A case study of the arterial switch operation.

Postoperative critical care management of congenital heart disease patients requires prompt intervention when the patient deviates significantly from clinician-determined vital sign and hemodynamic goals. Current monitoring systems only allow for static thresholds to be set on individual variables, despite the expectations that these signals change as the patient recovers and that variables interact. To address this incongruency, we have employed statistical process monitoring (SPM) techniques originally developed to monitor batch industrial processes to monitor high-frequency vital sign and hemodynamic data to establish multivariate trajectory maps for patients with d-transposition of the great arteries following the arterial switch operation. In addition to providing multivariate trajectory maps, the multivariate control charts produced by the SPM framework allow for assessment of adherence to the desired trajectory at each time point as the data is collected. Control charts based on slow feature analysis were compared with those based on principal component analysis. Alarms generated by the multivariate control charts are discussed in the context of the available clinical documentation.

Control charting methods for monitoring high dimensional data streams: A conceptual classification scheme

There are always challenges in various industrial or non-industrial processes in which the product quality/service is described by a large number of quality characteristics. Thus, statistical process monitoring (SPM) techniques for capturing the quality of high-dimensional processes are becoming increasingly important, and various control charts have been developed to monitor different types of quality characteristics in this area. In general, the construction of conventional control charts to monitor multivariate processes in a high-dimensional setting has some statistical limitations and leads to misleading interactions. As a result, novel control charting techniques have recently been suggested to ameliorate the efficiency of monitoring schemes under high-dimensional data streams. This paper undertakes a conceptual classification structure using content analysis to classify the existing literature in this context for the period 2004–2024. Furthermore, on the basis of 72 selected papers, the research gaps are identified and some directions to stimulate potential for future research are suggested.

About the Data Analysis of Optical Emission Spectra of Reactive Ion Etching Processes—The Method of Spectral Redundancy Reduction

In this study, we present the Method of Spectral Redundancy Reduction (MSRR) for analyzing OES (optical emission spectroscopy) data of dry etching processes based on the principles of spectral clustering. To achieve this, the OES data are transformed into abstract graph matrices whose associated eigenvectors directly indicate anomalies in the data set. We developed an approach that allows for the reduction in temporally resolved optical emission spectra from plasma structuring processes in such a way that individual emission lines can be algorithmically detected, which exhibit a temporal behavior different from the collective behavior of the temporally resolved overall spectrum. The proportion of emission lines that behave consistently throughout the entire process duration is not considered. Our work may find applications in which OES is used as a process-monitoring technique, especially for low-pressure plasma processing. The major benefit of the developed method is that the scale of the original data is kept, making physical interpretations possible despite data reductions.

Monitoring multistage healthcare processes using state space models and a machine learning based framework

Monitoring healthcare processes, such as surgical outcomes, with a keen focus on detecting changes and unnatural conditions at an early stage is crucial for healthcare professionals and administrators. In line with this goal, control charts, which are the most popular tool in the field of Statistical Process Monitoring, are widely employed to monitor therapeutic processes. Healthcare processes are often characterized by a multistage structure in which several components, states or stages form the final products or outcomes. In such complex scenarios, Multistage Process Monitoring (MPM) techniques become invaluable for monitoring distinct states of the process over time. However, the healthcare sector has seen limited studies employing MPM. This study aims to fill this gap by developing an MPM control chart tailored for healthcare data to promote early detection, confirmation, and patient safety. As it is important to detect unnatural conditions in healthcare processes at an early stage, the statistical control charts are combined with machine learning techniques (i.e., we deal with Intelligent Control Charting, ICC) to enhance detection ability. Through Monte Carlo simulations, our method demonstrates better performance compared to its statistical counterparts. To underline the practical application of the proposed ICC framework, real data from a two-stage thyroid cancer surgery is utilized. This real-world case serves as a compelling illustration of the effectiveness of the developed MPM control chart in a healthcare setting.

Image-Based Process Monitoring using Projective Nonnegative-Tensor Factorization

In this study, a novel process monitoring method using image data is presented. The new method is referred to as projective nonnegative-tensor factorization (PNTF). Video monitoring systems are increasingly being deployed in complex industrial processes such as manu-facturing processes, food industries, and pharmaceutical process, for extracting information from images to achieve the purpose of process monitoring. Traditional process-monitoring techniques fail to take full advantage of the rich information from color images due to the complex data characteristics including the high-dimensionality and correlation structures. The proposed PNTF is capable of modeling the high-dimensionality structure of red-green-blue data of color images using a tensor analysis technique. Moreover, in contrast to the traditional data-processing algorithms such as principal component analysis, linear discriminant analysis and so on, the PNTF is more applicable to analyzing image data for its nonnegative constraints on matrix factors. In addition, the PNTF allows us to extract features from both grayscale and color images, making process monitoring using multivariate control charts achiev- able. At last, based on the PNTF, methods for detecting and isolating process faults are developed and evaluated through a case study on a production process of electrically melting magnesia.

A multi-analytical approach to investigate DOM dynamics and alum dose control in enhanced coagulation process using wide-ranging surface waters

In this study, enhanced coagulation (EnC) − as a recommended water treatment process for disinfection by-product control − was investigated using jar testing, multi-analytical techniques, and a wide range of surface water sources. In addition to commonly used dissolved organic matter (DOM) concentration indices, DOM characteristics using UV–Vis spectroscopy, fluorescence EEM (fEEM) data, high-performance size exclusion chromatography (HPSEC), as well as turbidity, zeta potential and chlorophyll-a were studied. The results showed similar removal trends for the DOC, A254 and fluorescent DOM (fDOMs) signals with increasing alum dosing. Absorbance at ∼230 nm and fEEM data near Peak C in raw waters showed the highest correlations with enhanced dose. A decrease in weight-average apparent molecular weight following EnC, and a higher coagulability of DOM fractions with higher molecular weight, was noticed. The optimum zeta potential window was found to be −13 to +4 mV for EnC. Charge neutralization is suggested as the dominant mechanism for the EnC process. In addition to providing comprehensive information on the DOM dynamics during the EnC process, this study helps assess the potential of different water quality analysis techniques for EnC process monitoring and alum dose control, particularly using a field-deployable fDOM probe, offering potential for online, real-time applications.

A change-point–based control chart for detecting sparse mean changes in high-dimensional heteroscedastic data

Because of the “curse of dimensionality,” high-dimensional processes present challenges to traditional multivariate statistical process monitoring (SPM) techniques. In addition, the unknown underlying distribution of and complicated dependency among variables such as heteroscedasticity increase the uncertainty of estimated parameters and decrease the effectiveness of control charts. In addition, the requirement of sufficient reference samples limits the application of traditional charts in high-dimension, low-sample-size scenarios (small n, large p). More difficulties appear when detecting and diagnosing abnormal behaviors caused by a small set of variables (i.e., sparse changes). In this article, we propose two change-point–based control charts to detect sparse shifts in the mean vector of high-dimensional heteroscedastic processes. Our proposed methods can start monitoring when the number of observations is a lot smaller than the dimensionality. The simulation results show that the proposed methods are robust to nonnormality and heteroscedasticity. Two real data examples are used to illustrate the effectiveness of the proposed control charts in high-dimensional applications. The R codes are provided online.

Control charts for monitoring relative risk rate in the presence of Weibull competing risks with censored and masked data

ABSTRACT Competing risks data frequently appear in real-world operations like quality inspections, survival analysis, reliability tests, and clinical trials. From the quality point of view, relative risk rate can be considered an interesting quality indicator in analyzing the competing risks data for statistical process monitoring purposes. The relative risk rate measures the proportion of failures caused by the primary risk among a set of competing risks. This paper introduces two Shewhart-type control charts for monitoring the relative risk rate when the lifetimes of competing risks are independent Weibull random variables. The former chart is constructed based on the maximum likelihood estimation method, while the latter is developed based on the Bayesian approach. The proposed control charts can be applied in Phase II. The calculation of the Bayesian control charts and the evaluation of both process monitoring techniques have been done based on Monte Carlo simulations. The performance of the proposed control charts has been examined based on the average run length metric. The illustrative example is also discussed in detail to demonstrate the applicability of the proposed methods.

Digitalization of hydraulic rotary drilling process for continuously mechanical profiling of siliciclastic sedimentary rocks

Hydraulic rotary drilling can offer the essential information and core samplesa for the researches on solid earth. Recording the factual field drilling data and analyzing the hydraulic rotary coring process are challenging yet promising to utilize the massive drilling information in geophysics and geology. This paper adopts the drilling process monitoring (DPM) technique and records the four parameters of displacement, thrust pressure, upward pressure, and rotation speed in real-time series for profiling the siliciclastic sedimentary rocks along 108 m deep drillhole. The digitalization results with 107 linear zones represent the spatial distribution of drilled geomaterials including superficial deposits (fill, loess, gravelly soil), mudstone, silty mudstone, gritstone, and fine sandstone. The constant drilling speeds varying from 0.018 to 1.905 m/min present the in-situ coring resistance of drilled geomaterials. Furthermore, the constant drilling speeds can identify the strength quality of soils to hard rocks. The thickness distributions of the six basic strength quality grades are presented for all the sedimentary rocks and each individual type of the seven soil and rocks. The in-situ strength profile determined in this paper can be used to assess and evaluate the in-situ mechanical behavior of geomaterial along the drillhole and can provide a new mechanical-based assessment for determining the spatial distribution of geological strata and structures in subsurface. They are important since the same stratum at different depths can have different mechanical behavior. The results provide a novel quantitative measurement for continuously in-situ mechanical profiling by digital drilling data. The findings of the paper can offer a new and effective method for refinement and upgrading of in-situ ground investigation, and can provide researchers and engineers with a novel tool and valuable reference to digitize and utilize factual data of current drilling projects.

A Review on Statistical Process Control in Healthcare: Data-Driven Monitoring Schemes

In recent years, the adoption of statistical process monitoring (SPM) techniques in healthcare has been successful. For instance, biosurveillance and biosignal monitoring have demonstrated direct benefits. As the latest reviews of the literature show, parametric SPM techniques have been implemented to evaluate the quality-of-service hospitals provide, track medical equipment, monitor safety markers, or assess the improvements made by quality projects. However, as shown in this research, world-trending topics in data science that include data-driven approaches integrated with SPM have not been reviewed. To bridge this gap and shed light on new research, a systematic review of scientific databases and a taxonomic literature analysis were performed. For the scientometric analysis, a set of bibliometric indicators were obtained to portray the performance of each subtopic, such as examining growth kinetics, identifying top authors, journals, countries and affiliations, as well as creating network maps of co-authorship and keyword co-occurrence. Additionally, the taxonomic analysis involved grouping proposals by methodological approach. Each approach was explained and discussed to identify the advantages, limitations, and challenges that researchers and practitioners may encounter. SPM researchers and practitioners require more flexibility in data-driven approaches to account for frequency unbalance, complexity, dimensionality problems, and speed. Those working in data-driven and computer-oriented areas can expand their toolbox by incorporating sequential approaches to enhance the power of their classifiers, assess risk, reduce misspecification, and adopt model-oriented mindsets.

Addressing Random Variations in MWD Penetration Rate with the DPM Algorithm

Measurement While Drilling (MWD) is an in situ technique for the assessment of ground conditions. It records the drilling parameters of a drilling machine when its drill bit is penetrating into new geomaterials below the bottom of a drill hole. Its penetration rate along drill hole depth, however, has various random variations. Such random variations are inconsistent with the fact that the geomaterials in the ground have piece-wise homogeneity. Many methods have been adopted to filter and normalize the random variations in penetration rate for ground characterization. However, they have not completely resolved the issue and obtained the piece-wise homogeneous distribution of geomaterials in the ground. This paper uses the time-series algorithm of the Drilling Process Monitoring (DPM) technique to remove the random variations in MWD penetration rate. It further obtains the piece-wise constant variations in the newly addressed DPM penetration rate along the drill hole depth. With this algorithm, the DPM penetration rate is consistent with the fact that ground geomaterials have the property of piece-wise homogeneous distribution. Consequently, the piece-wise constant variations in DPM penetration rate can be used to characterize the strength variations in geomaterials along the drill hole depth to substantially upgrade the MWD technique for ground investigation.

Real-time analysis of magnetic nanoparticle clustering effects by inline-magnetic particle spectroscopy

The requirements for magnetic nanomaterials in life science strongly differ depending on the envisaged specific biomedical application. Especially, the magnetic properties of magnetic nanoparticles might - often unintended - change depending on their microscopic physiological environment with the consequence that the performance is reduced for the application. This frequently occurs if magnetic nanoparticles get internalized within distinct physiological entities in an organism due to immobilization or aggregation. As a consequence, magnetic nanoparticles obtain further surface modifications after synthesis to maintain the magnetic properties. Magnetic particle spectroscopy (MPS) is a fast, sensitive measurement technique that probes the dynamic magnetic response of MNP to a sinusoidal magnetic excitation and is very sensitive to detect changes of the local environment of MNP. In this study, we present the capability of inline-MPS to sensitively monitor the level of colloidal stability of electrostatically stabilized nanomaterials against saline-driven aggregation and to validate the successful aggregation prevention of particles by additional coating. Inline-MPS is a valuable technique for process monitoring as well as for investigating, supporting, and predicting the MNP behavior in biological environments.

Application potential of vibration sensors for online process monitoring of hydrocyclones

ABSTRACT An attempt has been made to develop a non-intrusive and robust technique for online process monitoring of hydrocyclones using commercially available vibration sensors. Vibration signals were captured first at the spigots of 50.8 mm and 76.2 mm diameter hydrocyclones while using water only at different operating conditions. For denoising the captured raw vibration signals, a wavelet-based method coupled with median filtering-based technique was used. The Root Mean Square (RMS) values of vibration accelerations induced in the hydrocyclone were found to be proportional to the total volumetric feed rate. Based on this observation, it has been demonstrated how suitable empirical correlations can be developed which can ultimately be used as an online process monitoring tool for hydrocyclones.

In Situ Thermoset Cure Sensing: A Review of Correlation Methods.

Thermoset polymer composites have increased in use across multiple industries, with recent applications consisting of high-complexity and large-scale parts. As applications expand, the emphasis on accurate process-monitoring techniques has increased, with a variety of in situ cure-monitoring sensors being investigated by various research teams. To date, a wide range of data analysis techniques have been used to correlate data collected from thermocouple, dielectric, ultrasonic, and fibre-optic sensors to information on the material cure state. The methods used in existing publications have not been explicitly differentiated between, nor have they been directly compared. This paper provides a critical review of the different data collection and cure state correlation methods for these sensor types. The review includes details of the relevant sensor configurations and governing equations, material combinations, data verification techniques, identified potential research gaps, and areas of improvement. A wide range of both qualitative and quantitative analysis methods are discussed for each sensing technology. Critical analysis is provided on the capability and limitations of these methods to directly identify cure state information for the materials under investigation. This paper aims to provide the reader with sufficient background on available analysis techniques to assist in selecting the most appropriate method for the application.

Direct Image Feature Extraction and Multivariate Analysis for Crystallization Process Characterization

Small-scale crystallization experiments (1–8 mL) are widely used during early-stage crystallization process development to obtain initial information on solubility, metastable zone width, as well as attainable nucleation and/or growth kinetics in a material-efficient manner. Digital imaging is used to monitor these experiments either providing qualitative information or for object detection coupled with size and shape characterization. In this study, a novel approach for the routine characterization of image data from such crystallization experiments is presented employing methodologies for direct image feature extraction. A total of 80 image features were extracted based on simple image statistics, histogram parametrization, and a series of targeted image transformations to assess local grayscale characteristics. These features were utilized for applications of clear/cloud point detection and crystal suspension density prediction. Compared to commonly used transmission-based methods (mean absolute error 8.99 mg/mL), the image-based detection method is significantly more accurate for clear and cloud point detection with a mean absolute error of 0.42 mg/mL against a manually assessed ground truth. Extracted image features were further used as part of a partial least-squares regression (PLSR) model to successfully predict crystal suspension densities up to 40 mg/mL (R2 > 0.81, Q2 > 0.83). These quantitative measurements reliably provide crucial information on composition and kinetics for early parameter estimation and process modeling. The image analysis methodologies have a great potential to be translated to other imaging techniques for process monitoring of key physical parameters to accelerate the development and control of particle/crystallization processes.

An assessment for the conditional performance of an support vector data description (SVDD)‐based chart

AbstractThe usual practice in Statistical Process Monitoring (SPM) techniques assumes that the data distribution is known and the related parameters are accurately estimated. In practice, the underlying distribution and its parameters are rarely known, and control charts need to be constructed with parameters being estimated. Such issues have recently received an increasing attention in evaluating the properties of both parametric and nonparametric charts. However, the same study is seldom conducted for the control charts based on the data‐driven tools. In this paper, we investigated the in‐control performance of a nonparametric control chart based on the Support Vector Data Description (SVDD) theory. More specifically, we discuss the conditional effect of the training Phase‐I samples on the Phase‐II efficiency when different distributions are considered. Simulation results show that the conditional performance of the SVDD‐based chart can be strongly affected by the Phase‐I samples. It this situation, adjusted control limits with a specific number of available training sample is suggested.

A Multiblock Kernel Dynamic Latent Variable Model for Large-Scale Industrial Process Monitoring

For industrial processes, the multivariate statistical process monitoring (MSPM) technique is an effective way to perform data-driven process modeling and fault detection. Compared to static multivariate statistical models (SMSMs), dynamic multivariate statistical models (DMSMs) can effectively model the autocorrelation and intercorrelation of data and improve the performance of fault detection. However, the performance of using DMSMs degrades when one uses data collected from large-scale processes, as it cannot effectively model the local behavior of large-scale processes. In contrast, multi-block methods can reflect these local behaviors well. Meanwhile, considering that kernel mapping is a proven technique for revealing nonlinear relationships in large-scale data. In this paper, a multivariate statistical model based on the multi-block kernel dynamic latent variable (MBKDLV) is proposed to monitor large-scale industrial processes. It divides high-dimensional process variables into blocks and monitors the local behavior by analyzing these blocks’ dynamic and static relationships. MBKDLV-based process data modeling is composed of in-block and inter-block algorithms, which also consider in-block and inter-block nonlinearity and correlations. Three statistics are proposed for monitoring the local behavior of large-scale processes. The effectiveness of the proposed approach is illustrated by a case study of a hot strip mill process (HSMP).