Search

Industrial revolutions have gone through four phases: steam, electricity, electronics, and Industry 4.0. Through all these four industrial revolutions, efficiency, productivity, quality, and automation have been greatly improved. However, the manufacturing processes created by humans have had disastrous consequences on the environment leading to a gigantic “climate change” problem. To mitigate climate change, engineers, and manufacturers all over the world have stepped up the research into cradle-to-cradle designs and sustainable manufacturing practices inspired by the designs and value cycles in nature. Bio-inspired designs have been gaining momentum to create products and manufacturing methods that are eco-friendly. All manufacturing (of a fruit, an organism such as a human baby) in nature happens in microfactories such as a womb, a leaf, a flower, or a chicken oviduct whose products are eggs. The product (egg) and the manufacturing process (chicken oviduct) are both green (eco-effective), lean (built with minimal resources), and smart (sensors and Internet of Things). Using a chicken as a model, this book chapter presents a set of metrics for green, lean, and smart attributes, which engineers can use to design products and microfactories.

Manufacturing has adopted technologies such as automation, robotics, industrial Internet of Things (IoT), and big data analytics to improve productivity, efficiency, and capabilities in the production environment. Modern manufacturing workers not only need to be adept at the traditional manufacturing technologies but also ought to be trained in the advanced data-rich computer-automated technologies. This study analyzes the data science and analytics (DSA) skills gap in today's manufacturing workforce to identify the critical technical skills and domain knowledge required for data science and intelligent manufacturing-related jobs that are highly in-demand in today's manufacturing industry. The gap analysis conducted in this paper on Emsi job posting and profile data provides insights into the trends in manufacturing jobs that leverage data science, automation, cyber, and sensor technologies. These insights will be helpful for educators and industry to train the next generation manufacturing workforce. The main contribution of this paper includes (1) presenting the overall trend in manufacturing job postings in the U.S., (2) summarizing the critical skills and domain knowledge in demand in the manufacturing sector, (3) summarizing skills and domain knowledge reported by manufacturing job seekers, (4) identifying the gaps between demand and supply of skills and domain knowledge, and (5) recognize opportunities for training and upskilling workforce to address the widening skills and knowledge gap.

The computational aero-acoustic study of an isolated passenger car tire is carried out to understand the effect of dimensions of longitudinal tire grooves and operational parameters (velocity and temperature) on tire noise. The computational fluid dynamics and acoustic models are used to obtain aero-acoustic tire noise at near-field and far-field receivers around the tire and artificial neural networks-based regression are used to study the highly non-linear and interactive causal relationships in the system. Unsteady Reynolds-Averaged Navier-Stokes based realizable k-epsilon model is used to solve the flow field in the computational domain. The Ffowcs Williams and Hawkings model is used to obtain aero-acoustic tire noise at far-field positions. Spectral analysis is used to convert the output time domain to frequency domain and to obtain A-weighted sound pressure level. Artificial neural network–based response surface regression is conducted to understand casual relationships between A-weighted sound pressure level and control variables (Groove depth, Groove width, Temperature and velocity). Maximum A-weighted sound pressure level is observed in the wake region of the tire model. The interaction study indicates that ∼10% reduction in the aero-acoustic emissions is possible by selecting appropriate combinations of groove width and groove depth. The interaction of velocity with width is found to be most significant with respect to A-weighted sound pressure level at all receivers surrounding the tire. The interaction of operational parameters, that is, velocity and temperature are found to be significant with respect to A-weighted sound pressure level at wake and front receivers. Therefore, the regional speed limits and seasonal temperatures need to be considered while designing the tire to achieve minimum aero-acoustic emissions.

Perturbations in networked dynamical complex systems could cause cascading failures, consequently instability, and eventually collapse of the systems. In this work, we derive a set of formulations that help us understand the mechanism of high-dimensional interactions among components and uncover the principles that control the dynamics of interacting components. Our formulation reduces the system’s high dimensional dynamics to a parsimonious resilience function of a parameter that solely depends on the topology but effectively captures the complex interaction structure. The experimental results demonstrate that the formulation can accurately predict the resilience loss brought about by perturbations and identify the tipping point where the systemic collapse occurs. These predictive results can be used for enhancing a complex system’s ability to withstand perturbations and avert catastrophic collapses. The study highlights the topological properties that can be used as principles for improving the resilience of ecosystems, biological systems, economic systems and technological infrastructures.

In the current work, the effects of design (groove depth and groove width) and operational (temperature and velocity) parameters on aerodynamic performance parameters (coefficient of drag and coefficient of lift) of an isolated passenger car tire have been investigated. The study is conducted by using neural network-based Monte-Carlo analysis on computational fluid dynamics (CFD). The computer experiments are designed to obtain the causal relationship between tire design, operational, and aerodynamic performance parameters. The Reynolds-averaged Navier–Stokes equations-based RealizableK-εmodel has been employed to analyze the variations in flow patterns around an isolated tire. The design parameters are varied over wide range and full factorial design, while considering temperature and velocity is completely explored to draw conclusive results. The multi-layer perceptron type neural network with the back-propagation algorithm is trained to map any non-linearity in causal relationships. The sensitivity analysis is performed to find the relationship between control variables and performance indicators. The importance of control variable is determined by both sensitivity and significance analyses and the paired interaction analysis is performed between selected control variables to find the interactive behavior of corresponding variables. The design parameter of groove width with 6.8% and 41% reduction in drag and lift coefficient, respectively, and conventionally overlooked operational parameter of velocity with 4% and 35% impact on drag and lift coefficient, respectively, are found to be the most significant variables. The air trapped between the longitudinal grooves and the road is found to follow the beam theory. The interaction of the groove depth and width is found to be significant with respect to coefficient of lift based on the air beam concept. The interaction of groove width and velocity is found to be significant with respect to both coefficients of lifts and drag.

With the potential of Industry 4.0 technologies to enable sustainable manufacturing, asset life cycle management (ALCM) has been gaining increasing attention in recent years. This study explores the evolution of Industry 4.0 technology applications to sustainable ALCM from 2002 to 2021. This study is based on keywords collected from 3896 ALCM-related scientific articles published in the Web of Science, IEEE Xplore and Engineering Village between 2002 and 2021. We conducted a review analysis of these keywords using a network science-based methodology, which unlike the tedious traditional literature review methods, gives the capability to analyze a huge number of scientific articles efficiently. We built keyword co-occurrence networks (KCNs) from the keywords and explored the network characteristics to uncover meaningful knowledge patterns, knowledge components, knowledge structure, and research trends in the body of literature at the intersection of ALCM and Industry 4.0. The network modeling and data analysis results identify the emerging Industry 4.0-related keywords in ALCM literature and indicate the recent explosion of connectivity among keywords. We found IoT, predictive maintenance and big data to be the top three most popular Industry 4.0-related keywords in ALCM literature. Furthermore, this study maps relevant ALCM keywords in contemporary literature to the nine pillars of Industry 4.0 to help the responsible manufacturing community identify research trends and emerging technologies for sustainability.

In current clinical settings, typically pain is measured by a patient’s self-reported information. This subjective pain assessment results in suboptimal treatment plans, over-prescription of opioids, and drug-seeking behavior among patients. In the present study, we explored automatic objective pain intensity estimation machine learning models using inputs from physiological sensors. This study uses BioVid Heat Pain Dataset. We extracted features from Electrodermal Activity (EDA), Electrocardiogram (ECG), Electromyogram (EMG) signals collected from study participants subjected to heat pain. We built different machine learning models, including Linear Regression, Support Vector Regression (SVR), Neural Networks and Extreme Gradient Boosting for continuous value pain intensity estimation. Then we identified the physiological sensor, feature set and machine learning model that give the best predictive performance. We found that EDA is the most information-rich sensor for continuous pain intensity prediction. A set of only 3 features from EDA signals using SVR model gave an average performance of 0.93 mean absolute error (MAE) and 1.16 root means square error (RMSE) for the subject-independent model and of 0.92 MAE and 1.13 RMSE for subject-dependent. The MAE achieved with signal-feature-model combination is less than 1 unit on 0 to 4 continues pain scale, which is smaller than the MAE achieved by the methods reported in the literature. These results demonstrate that it is possible to estimate pain intensity of a patient using a computationally inexpensive machine learning model with 3 statistical features from EDA signal which can be collected from a wrist biosensor. This method paves a way to developing a wearable pain measurement device.

Craig Versek,1 S Mohammad Ali Banijamali,2 Peter Bex,3 Kameran Lashkari,4 Sagar Kamarthi,2 Srinivas Sridhar1,5,6 1NeuroFieldz Inc, Newton, MA, USA; 2Department of Mechanical and Industrial Engineering, Northeastern University, Boston, MA, USA; 3Department of Psychology, Northeastern University, Boston, MA, USA; 4Department of Bioengineering, University of Massachusetts Dartmouth, Dartmouth, MA, USA; 5Departments of Physics, Bioengineering, and Chemical Engineering, Northeastern University, Boston, MA, USA; 6Division of Radiation Oncology, Harvard Medical School, Boston, MA, USACorrespondence: Srinivas SridharDepartment of Physics, Northeastern University, 110 Forsyth St., 111 Dana Research Center, Boston, MA, 02115, USAEmail s.sridhar@northeastern.eduBackground: Delayed Dark-Adapted vision Recovery (DAR) is a biomarker for Age-related Macular Degeneration (AMD), however its measurement is burdensome for patients and examiners.Methods: In this study, we developed a portable, wireless and user-friendly system that employs a headset with a smartphone to deliver controlled photo-bleach and monocular pattern reversal stimuli, while using custom electroencephalography (EEG) electrodes and electronics in order to measure Dark-Adapted Visual Evoked Potentials (DAVEP) objectively and separately at the peripheral and central visual field. This is achieved in one comfortable 20-minute session, without requiring subject reporting. DAVEP responses post photo-bleach for up to 15 minutes were measured concurrently from both eyes in 12 AMD-patients, 1 degenerative myopia patient, and 8 controls who had no diagnosed macular vision loss.Results: Robust positive polarity DAVEP responses were observed at 200– 500 ms from stimulus onset to scotopic stimuli that have been seldom reported and analyzed previously. The amplitude recovery of the DAVEP response was significantly delayed in AMD patients as compared to controls. We developed DAVEP1 scores, a simple metric for DAR, which classified 90% of subject eyes correctly, indicating the presence of AMD in at least one eye of all pre-confirmed subjects with this diagnosis.Conclusion: We developed a user-friendly, portable VEP system and DAVEP1 metric, which show a high potential to identify DAR-deficits in AMD-patients. This novel technology could aid in early diagnosis of AMD.Keywords: age-related macular degeneration, EEG, visual evoked potential, dark adaptation recovery, scotopic vision, portable diagnostics

In this work, we explored permutation entropy and complexity representation for classification of machine component fault signals. The proposed method is computationally inexpensive and has the ability to handle nonlinear and non-stationary signals. We used vibrations signals collected from normal and faulty ball bearings to demonstrate the effectiveness of the permutation entropy and complexity representation. We performed a sensitivity analysis to study the effect of representation parameters and signal length on the signal classification. We observed that embedding dimension \(D = 3, 4, 5, 6\) and embedding delay \(\tau = 1, 5\) are appropriate for good class separability in this application and the separability of classes improves with increasing signal length N. The overall accuracy of signal classification varies from 90 to 100% depending on signal length.

Background It is well established that lack of physical activity is detrimental to the overall health of an individual. Modern-day activity trackers enable individuals to monitor their daily activities to meet and maintain targets. This is expected to promote activity encouraging behavior, but the benefits of activity trackers attenuate over time due to waning adherence. One of the key approaches to improving adherence to goals is to motivate individuals to improve on their historic performance metrics. Objective The aim of this work was to build a machine learning model to predict an achievable weekly activity target by considering (1) patterns in the user’s activity tracker data in the previous week and (2) behavior and environment characteristics. By setting realistic goals, ones that are neither too easy nor too difficult to achieve, activity tracker users can be encouraged to continue to meet these goals, and at the same time, to find utility in their activity tracker. Methods We built a neural network model that prescribes a weekly activity target for an individual that can be realistically achieved. The inputs to the model were user-specific personal, social, and environmental factors, daily step count from the previous 7 days, and an entropy measure that characterized the pattern of daily step count. Data for training and evaluating the machine learning model were collected over a duration of 9 weeks. Results Of 30 individuals who were enrolled, data from 20 participants were used. The model predicted target daily count with a mean absolute error of 1545 (95% CI 1383-1706) steps for an 8-week period. Conclusions Artificial intelligence applied to physical activity data combined with behavioral data can be used to set personalized goals in accordance with the individual’s level of activity and thereby improve adherence to a fitness tracker; this could be used to increase engagement with activity trackers. A follow-up prospective study is ongoing to determine the performance of the engagement algorithm.

The market for solar renewable energy has expanded extensively due to decreased levelized cost of energy and the appeal of zero carbon emissions during generation. Solar PV systems provide low-cost, flexible opportunities for utilizing renewable energy in manufacturing sectors, which globally consume 33% of all primary energy and produce 37% of carbon dioxide emissions. This study explores three U.S. states (California, New Jersey, and Texas) for economic and environmental consequences of mounting onsite solar PV systems at manufacturing facilities. The objective of the study is to measure the effect of various financial incentives and locational differences on the economic viability and environmental performance of solar PV systems.

BACKGROUND Total joint replacements are high-volume and high-cost procedures that should be monitored for cost and quality control. Models that can identify patients at high risk of readmission might help reduce costs by suggesting who should be enrolled in preventive care programs. Previous models for risk prediction have relied on structured data of patients rather than clinical notes in electronic health records (EHRs). The former approach requires manual feature extraction by domain experts, which may limit the applicability of these models. OBJECTIVE This study aims to develop and evaluate a machine learning model for predicting the risk of 30-day readmission following knee and hip arthroplasty procedures. The input data for these models come from raw EHRs. We empirically demonstrate that unstructured free-text notes contain a reasonably predictive signal for this task. METHODS We performed a retrospective analysis of data from 7174 patients at Partners Healthcare collected between 2006 and 2016. These data were split into train, validation, and test sets. These data sets were used to build, validate, and test models to predict unplanned readmission within 30 days of hospital discharge. The proposed models made predictions on the basis of clinical notes, obviating the need for performing manual feature extraction by domain and machine learning experts. The notes that served as model inputs were written by physicians, nurses, pathologists, and others who diagnose and treat patients and may have their own predictions, even if these are not recorded. RESULTS The proposed models output readmission risk scores (propensities) for each patient. The best models (as selected on a development set) yielded an area under the receiver operating characteristic curve of 0.846 (95% CI 82.75-87.11) for hip and 0.822 (95% CI 80.94-86.22) for knee surgery, indicating reasonable discriminative ability. CONCLUSIONS Machine learning models can predict which patients are at a high risk of readmission within 30 days following hip and knee arthroplasty procedures on the basis of notes in EHRs with reasonable discriminative power. Following further validation and empirical demonstration that the models realize predictive performance above that which clinical judgment may provide, such models may be used to build an automated decision support tool to help caretakers identify at-risk patients.

BACKGROUND It is well established that lack of physical activity is detrimental to the overall health of an individual. Modern-day activity trackers enable individuals to monitor their daily activities to meet and maintain targets. This is expected to promote activity encouraging behavior, but the benefits of activity trackers attenuate over time due to waning adherence. One of the key approaches to improving adherence to goals is to motivate individuals to improve on their historic performance metrics. OBJECTIVE The aim of this work was to build a machine learning model to predict an achievable weekly activity target by considering (1) patterns in the user’s activity tracker data in the previous week and (2) behavior and environment characteristics. By setting realistic goals, ones that are neither too easy nor too difficult to achieve, activity tracker users can be encouraged to continue to meet these goals, and at the same time, to find utility in their activity tracker. METHODS We built a neural network model that prescribes a weekly activity target for an individual that can be realistically achieved. The inputs to the model were user-specific personal, social, and environmental factors, daily step count from the previous 7 days, and an entropy measure that characterized the pattern of daily step count. Data for training and evaluating the machine learning model were collected over a duration of 9 weeks. RESULTS Of 30 individuals who were enrolled, data from 20 participants were used. The model predicted target daily count with a mean absolute error of 1545 (95% CI 1383-1706) steps for an 8-week period. CONCLUSIONS Artificial intelligence applied to physical activity data combined with behavioral data can be used to set personalized goals in accordance with the individual’s level of activity and thereby improve adherence to a fitness tracker; this could be used to increase engagement with activity trackers. A follow-up prospective study is ongoing to determine the performance of the engagement algorithm.

Unexpected component failures in a mechanical system always cause loss of performance and functionality of the entire system. Condition-based maintenance decisions for a multi-component mechanical system are challenging because the interdependence of individual components’ degradation is not fully understood and lack of physical models. Most existing literature commonly assumes that degradation and failure of individual components within a mechanical system are independent, which could lead to inaccurate diagnostic and prognostic results. In this research, state-rate dependence denoting interaction between component health condition (degradation state) and failure rate is proposed for degradation and failure analysis for a two-component repairable system. A state discretization technique is proposed to model how health state of one component affects the hazard rate of another. An extended proportional hazard model (PHM) is used to characterize the failure dependence and estimate the influence of degradation state of one component on the hazard rate of another. An optimization model is developed to determine the optimal hazard-based threshold for a two-component repairable system. A case study on a generic industrial gearbox has been conducted to show the effectiveness of the proposed model.

In this study, we evaluated the diagnostic accuracy of shear wave elastography (SWE) for differentiating high-risk non-alcoholic steatohepatitis (hrNASH) from non-alcoholic fatty liver and low-risk non-alcoholic steatohepatitis (NASH). Patients with non-alcoholic fatty liver disease scheduled for liver biopsy underwent pre-biopsy SWE. Ten SWE measurements were obtained. Biopsy samples were reviewed using the NASH Clinical Research Network Scoring System and patients with hrNASH were identified. Receiver operating characteristic curves for SWE-based hrNASH diagnosis were charted. One hundred sixteen adult patients underwent liver biopsy at our institution for the evaluation of non-alcoholic fatty liver disease. The area under the receiver operating characteristic curve of SWE for hrNASH diagnosis was 0.73 (95% confidence interval: 0.61-0.84, p < 0.001). The Youden index-based optimal stiffness cutoff value for hrNASH diagnosis was calculated as 8.4 kPa (1.67 m/s), with a sensitivity of 77% and specificity of 66%. SWE may be useful for the detection of NASH patients at risk of long-term liver-specific morbidity and mortality.

Delayed Dark Adapted vision Recovery (DAR) is a biomarker for Age-related Macular Degeneration (AMD), however its measurement is burdensome for patients and examiners. We developed a portable, wireless, quick-setup system that employs a headset with a smartphone to deliver controlled photobleach and monocular pattern reversal stimuli, while using custom electroencephalography (EEG) electrodes and electronics in order to measure objective Dark Adapted Visual Evoked Potentials (DAVEP) at multiple locations of the visual field - all in one comfortable 20-minute session, without requiring subject reporting. DAVEP responses post photobleach (up to 15 minutes), were measured concurrently in both eyes of 13 patients with AMD and 8 others not diagnosed with AMD. New, unexpected features at high latencies were observed in the DAVEP responses to scotopic stimulus intensities. The amplitude recovery of the DAVEP response was significantly delayed in AMD patients compared with controls. We developed DAVEP1 scores, a simple metric for DAR, using it to successfully identify all 100% of AMD subjects and classify 90% of subject eyes correctly. Deficits in DAR in patients with AMD can be identified using this objective VEP based system and the DAVEP1 metric, a promising new objective biomarker for this disease that can be easily tested in a clinic.

It is well established that lack of physical activity is detrimental to overall health of an individual. Modern day activity trackers enable individuals to monitor their daily activity to meet and maintain targets and to promote activity encouraging behavior. However, the benefits of activity trackers are attenuated over time due to waning adherence. One of the key methods to improve adherence to goals is to motivate individuals to improve on their historic performance metrics. In this work we developed a machine learning model to dynamically adjust the activity target for the forthcoming week that can be realistically achieved by the activity-tracker users. This model prescribes activity target for the forthcoming week. We considered individual user-specific personal, social, and environmental factors, daily step count through the current week (7 days). In addition, we computed an entropy measure that characterizes the pattern of daily step count for the current week. Data for training the machine learning model was collected from 30 participants over a duration of 9 weeks. The model predicted target daily count with mean absolute error of 1545 steps. The proposed work can be used to set personalized goals in accordance with the individual’s level of activity and thereby improving adherence to fitness tracker.

Current Supplier-Manufacturer (SM) networks are highly complex and susceptible to local and global disruptions, due to connectivity and interdependency among suppliers and manufacturers. Resiliency of supply chains is critical for organizations to remain operational in the face of disruptive events. Existing quantitative analyses oversimplify the mutualistic nature of SM networks, in which failure of individual entities affects not only the directly connected entities but also those connected indirectly. In this work we investigate resiliency of SM networks using the quantitative methods employed to study mutualistic ecological systems. Much like in ecological systems, catastrophic failures of SM networks are difficult to predict due to high dimensionality of their interactive space. To address this, first we create a bipartite representation and generate a multidimensional nonlinear model that captures the dynamics of a SM network. We transform the multidimensional model into a two-dimensional model without sacrificing the model’s ability to predict the point of collapse. We extensively validate the model using real-world global automotive SM networks. We observe that the resiliency of a SM network depends on both the network structure and parameters. The current work offers a means for designing resilient supply chains that can remain robust to local and global perturbations.

BackgroundType 1 diabetes mellitus (T1DM) is characterized by chronic insulin deficiency and consequent hyperglycemia. Patients with T1DM require long-term exogenous insulin therapy to regulate blood glucose levels and prevent the long-term complications of the disease. Currently, there are no effective algorithms that consider the unique characteristics of T1DM patients to automatically recommend personalized insulin dosage levels.ObjectiveThe objective of this study was to develop and validate a general reinforcement learning (RL) framework for the personalized treatment of T1DM using clinical data.MethodsThis research presents a model-free data-driven RL algorithm, namely Q-learning, that recommends insulin doses to regulate the blood glucose level of a T1DM patient, considering his or her state defined by glycated hemoglobin (HbA1c) levels, body mass index, engagement in physical activity, and alcohol usage. In this approach, the RL agent identifies the different states of the patient by exploring the patient’s responses when he or she is subjected to varying insulin doses. On the basis of the result of a treatment action at time step t, the RL agent receives a numeric reward, positive or negative. The reward is calculated as a function of the difference between the actual blood glucose level achieved in response to the insulin dose and the targeted HbA1c level. The RL agent was trained on 10 years of clinical data of patients treated at the Mass General Hospital.ResultsA total of 87 patients were included in the training set. The mean age of these patients was 53 years, 59% (51/87) were male, 86% (75/87) were white, and 47% (41/87) were married. The performance of the RL agent was evaluated on 60 test cases. RL agent–recommended insulin dosage interval includes the actual dose prescribed by the physician in 53 out of 60 cases (53/60, 88%).ConclusionsThis exploratory study demonstrates that an RL algorithm can be used to recommend personalized insulin doses to achieve adequate glycemic control in patients with T1DM. However, further investigation in a larger sample of patients is needed to confirm these findings.

The data in this article provide details about MRI lesion segmentation using K-means and Gaussian Mixture Model-Expectation Maximization (GMM-EM) algorithms. Both K-means and GMM-EM algorithms can segment lesion area from the rest of brain MRI automatically. The performance metrics (accuracy, sensitivity, specificity, false positive rate, misclassification rate) were estimated for the algorithms and there was no significant difference between K-means and GMM-EM. In addition, lesion size does not affect the accuracy and sensitivity for either method. Keywords: Ischemic stroke, Lesion, Magnetic resonance image (MRI), Segmentation

Magnesium oxide (MgO) is a good candidate for an interface layer in multifunctional metal-oxide nanoscale thin-film heterostructures due to its high breakdown field and compatibility with complex oxides through O bonding. In this research, molecular beam epitaxy (MBE) is used to deposit 10 nm to 15 nm MgO single-crystal films on silicon carbide with hexagonal polytype 6H (6H-SiC) to serve as an interface layer for effective integration of functional oxides. In this work, the effect of MBE process control variables on the growth rate of the MgO film measured in nanometers per minute is investigated. Experiments are conducted at various process conditions and the resulting MgO film growth rate at each combination of process conditions is measured. The process control variables studied are the substrate temperature (100 °C – 300 °C), magnesium source temperature (328 °C – 350 °C), plasma intensity (0 mV – 550 mV), and percentage oxygen on the starting surface of 6H-SiC substrate (9 % – 13 %) after the substrate is prepared by high-temperature hydrogen etching. The film thickness is computed using the effective attenuation length (EAL) of silicon photoelectron peak intensity as measured by x-ray photoelectron spectroscopy (XPS). The film thickness is converted to growth rate by dividing it with the duration of film growth. Using the experimental data, a neural network model is developed to estimate growth rate for any given process variable combination. From this neural network model, multiple replications of data were generated to conduct a 3-level full factorial design of experiments and response surface-based analysis. The study reveals that the plasma intensity has the most significant influence on growth rate. The results indicate that growth rate is relatively low on high-quality substrates with √3 × √3 R30° reconstructed 6H-SiC (0001) surface with optimum oxygen content (approximately 10 %); in contrast, the growth rate is relatively high on substrates with high surface roughness and excessive oxygen on the starting substrate surface.

Surface roughness is one of the important factors in all areas of tribology and in evaluation of the quality of machining operations such as turning, milling, or grinding. This paper describes a method for in-process estimation of workpiece surface roughness in a turning process from acoustic emission signals generated by the sliding friction between a graphite probe and the workpiece. Acoustic emission signals are first transformed into recurrence plots and from these plots computed are a set of recurrence statistics using the recurrence quantification analysis. The surface roughness parameters are estimated using an artificial neural network, taking the recurrence statistics of the acoustic emission signals as inputs. This method is verified by conducting an extensive set of experiments on AISI 1054 steel workpiece and K420 grade uncoated carbide inserts. Three surface roughness parameters, namely arithmetic mean (Ra), maximum peak-to-valley roughness (Rmax), and mean roughness depth (Rz) are estimated. The estimation accuracy of the proposed method is in the range of 90.13 to 91.26 %. The merit of the proposed method lies in its ability to circumvent the limitations of the existing optical and induction-based techniques which are hampered by the rotation of the workpiece, presence of coolants, and interference of chips. Furthermore, these accurate results indicate that the proposed surface roughness estimation method is very effective and amenable for practical implementation.

Despite the many touted benefits of nanomaterials, concerns remain about their possible environmental, health, and safety (EHS) risks in terms of their toxicity, long-term accumulation effects, or dose-response relationships. The published studies on EHS risks of nanomaterials have increased significantly over the past decade and half, with most focused on nanotoxicology. Researchers are still learning about health consequences of nanomaterials and how to make environmentally responsible decisions regarding their production. This article characterizes the scientific literature on nano-EHS risk analysis to map the state-of-the-art developments in this field and chart guidance for the future directions. First, an analysis of keyword co-occurrence networks is investigated for nano-EHS literature published in the past decade to identify the intellectual turning points and research trends in nanorisk analysis studies. The exposure groups targeted in emerging nano-EHS studies are also assessed. System engineering methods for risk, safety, uncertainty, and system reliability analysis are reviewed, followed by detailed descriptions where applications of these methods are utilized to analyze nanomaterial EHS risks. Finally, the trends, methods, future directions, and opportunities of system engineering methods in nano-EHS research are discussed. The analysis of nano-EHS literature presented in this article provides important insights on risk assessment and risk management tools associated with nanotechnology, nanomanufacturing, and nano-enabled products.

The cross correlation coefficient has been widely applied in financial time series analysis, in specific, for understanding chaotic behaviour in terms of stock price and index movements during crisis periods. To better understand time series correlation dynamics, the cross correlation matrices are represented as networks, in which a node stands for an individual time series and a link indicates cross correlation between a pair of nodes. These networks are converted into simpler trees using different schemes. In this context, Minimum Spanning Trees (MST) are the most favoured tree structures because of their ability to preserve all the nodes and thereby retain essential information imbued in the network. Although cross correlations underlying MSTs capture essential information, they do not faithfully capture dynamic behaviour embedded in the time series data of financial systems because cross correlation is a reliable measure only if the relationship between the time series is linear. To address the issue, this work investigates a new measure called phase synchronization (PS) for establishing correlations among different time series which relate to one another, linearly or nonlinearly. In this approach the strength of a link between a pair of time series (nodes) is determined by the level of phase synchronization between them. We compare the performance of phase synchronization based MST with cross correlation based MST along selected network measures across temporal frame that includes economically good and crisis periods. We observe agreement in the directionality of the results across these two methods. They show similar trends, upward or downward, when comparing selected network measures. Though both the methods give similar trends, the phase synchronization based MST is a more reliable representation of the dynamic behaviour of financial systems than the cross correlation based MST because of the former’s ability to quantify nonlinear relationships among time series or relations among phase shifted time series.

Background Total joint replacements are high-volume and high-cost procedures that should be monitored for cost and quality control. Models that can identify patients at high risk of readmission might help reduce costs by suggesting who should be enrolled in preventive care programs. Previous models for risk prediction have relied on structured data of patients rather than clinical notes in electronic health records (EHRs). The former approach requires manual feature extraction by domain experts, which may limit the applicability of these models. Objective This study aims to develop and evaluate a machine learning model for predicting the risk of 30-day readmission following knee and hip arthroplasty procedures. The input data for these models come from raw EHRs. We empirically demonstrate that unstructured free-text notes contain a reasonably predictive signal for this task. Methods We performed a retrospective analysis of data from 7174 patients at Partners Healthcare collected between 2006 and 2016. These data were split into train, validation, and test sets. These data sets were used to build, validate, and test models to predict unplanned readmission within 30 days of hospital discharge. The proposed models made predictions on the basis of clinical notes, obviating the need for performing manual feature extraction by domain and machine learning experts. The notes that served as model inputs were written by physicians, nurses, pathologists, and others who diagnose and treat patients and may have their own predictions, even if these are not recorded. Results The proposed models output readmission risk scores (propensities) for each patient. The best models (as selected on a development set) yielded an area under the receiver operating characteristic curve of 0.846 (95% CI 82.75-87.11) for hip and 0.822 (95% CI 80.94-86.22) for knee surgery, indicating reasonable discriminative ability. Conclusions Machine learning models can predict which patients are at a high risk of readmission within 30 days following hip and knee arthroplasty procedures on the basis of notes in EHRs with reasonable discriminative power. Following further validation and empirical demonstration that the models realize predictive performance above that which clinical judgment may provide, such models may be used to build an automated decision support tool to help caretakers identify at-risk patients.

In recent years, significant advancements in Fused Filament Fabrication (FFF) have enabled this technology to become one of the most leading techniques of Additive Manufacturing (AM) for the production of functional products. The poor mechanical properties of manufactured parts have traditionally imposed considerable limitations on use of FFF processes. These shortcomings have been overcome using new advanced filaments with nanoparticle reinforced components, short-length and continuous fibres, and other composite material processing technologies. Polymers reinforced with graphene nanoplatelets (GNP) have been an effective solution for improving electrical, thermal, and mechanical properties. However, the geometric properties of functional products manufactured with GNP reinforced polymers have not been analysed in spite of being crucial for the manufacture, assembly, and service life of functional products. The aim of this study was to compare an improved PLA polymer (PLA-3D) with a GNP reinforced PLA composite (PLA-Graphene) by analysing the geometric properties of dimensional accuracy, flatness error, surface texture, and surface roughness. The effect of the 3D printing parameters − build orientation (Bo), layer thickness (Lt), and feed rate (Fr) − on the geometric properties of two PLA-based filaments were evaluated. The results showed dimensional accuracy was mainly affected by the build orientation, where an increase in the layer area on the X–Y plane showing the highest dimensional deviation owing to the longer displacements of the extruder accumulating positioning errors. The dimensional accuracy along the Z-axis was not affected by any of the printing parameters nor the accumulation of layers, with results close to nominal ones. The flatness error and surface roughness were strongly conditioned by building orientation, with the best results obtained in the flat orientation. Neither of the compared materials showed significant variations between them in geometric properties, with similar results in the tested printing conditions.

Coronavirus disease (COVID-19) has created substantial detrimental impacts on the global manufacturing industry The COVID-19 outbreak across the globe has caused a short supply of raw materials and components required to keep production running in the automobile, chemical, electronics, and aircraft sectors In this technical note, the authors present their views on how different manufacturing stages are affected by various COVID-related factors The authors discuss their ideas on employing emerging industry 4 0 technologies as building blocks of resilience themes necessary to circumvent COVID-related disruptions and make future manufacturing resilient, efficient, and cost-effective © 2020 by ASTM International

BACKGROUND Diabetes is a serious chronic disease marked by high levels of blood glucose. It results from issues related to how insulin is produced and/or how insulin functions in the body. In the long run, uncontrolled blood sugar can damage the vessels that supply blood to important organs such as heart, kidneys, eyes, and nerves. Currently there are no effective algorithms to automatically recommend insulin dosage level considering the characteristics of a diabetic patient. OBJECTIVE The objective of this work is to develop and validate a general reinforcement learning framework and a related learning model for personalized treatment and management of Type 1 diabetes and its complications. METHODS This research presents a model-free reinforcement learning (RL) algorithm to recommend insulin level to regulate the blood glucose level of a diabetic patient considering his/her state defined by A1C level, alcohol usage, activity level, and BMI value. In this approach, an RL agent learns from its exploration and response of diabetic patients when they are subject to different actions in terms of insulin dosage level. As a result of a treatment action at time step t, the RL agent receives a numeric reward depending on the response of the patient’s blood glucose level. At each stage the reward for the learning agent is calculated as a function of the difference between the glucose level in the patient body and its target level. The RL algorithm is trained on ten years of the clinical data of 87 patients obtained from the Mass General Hospital. Demographically, 59% of patients are male and 41% of patients are female; the median of age is 54 years and mean is 52.92 years; 86% of patients are white and 47% of 87 patients are married. RESULTS The performance of the algorithm is evaluated on 60 test cases. Further the performance of Support Vector Machine (SVM) has been applied for Lantus class prediction and results has been compared with Q-learning algorithm recommendation. The results show that the RL recommendations of insulin levels for test patients match with the actual prescriptions of the test patients. The RL gave prediction with an accuracy of 88% and SVM shows 80% accuracy. CONCLUSIONS Since the RL algorithm can select actions that improve patient condition by taking into account delayed effects, it has a good potential to control blood glucose level in diabetic patients.

Objectives To investigate the impact of insurance coverage on the adoption of customized individually made (CIM) knee implants and to compare patient outcomes and cost effectiveness of off-the-shelf and CIM implants. Methods A system dynamics simulation model was developed to study adoption dynamics of CIM and meet the research objectives. The model reproduced the historical data on primary and revision knee replacement implants obtained from the literature and the Nationwide Inpatient Sample. Then the dynamics of adoption of CIM implants were simulated from 2018 to 2026. The rate of 90-day readmission, 3-year revision surgery, recovery period, time savings in operating rooms, and the associated cost within 3 years of primary knee replacement implants were used as performance metrics. Results The simulation results indicate that by 2026, an adoption rate of 90% for CIM implants can reduce the number of readmissions and revision surgeries by 62% and 39%, respectively, and can save hospitals and surgeons 6% on procedure time and cut down cumulative healthcare costs by approximately $38 billion. Conclusions CIM implants have the potential to deliver high-quality care while decreasing overall healthcare costs, but their adoption requires the expansion of current insurance coverage. This work presents the first systematic study to understand the dynamics of adoption of CIM knee implants and instrumentation. More broadly, the current modeling approach and systems thinking perspective could be used to consider the adoption of any emerging customized therapies for personalized medicine.