Process Downtime Research Articles

Machine Learning Based Fault Classification in Pilot Plant Batch Reactor: Using Support Vector Machine.

Identifying and diagnosing faults is a critical task in process industries to maintain effective monitoring of process and plant safety. Minimizing process downtime is critical for enhancing the quality of the product and minimizing production costs. Real-time categorization of issues across several levels is essential for the monitoring of processes. However, there are still notable obstacles, that must be addressed, such as the existence of robust correlations, the complexity of the data, and the lack of linearity. This study introduces a novel fault identification technique in batch reactor experimental trials that employs multikernel support vector machines (SVMs) to categorize internal and external issues, specifically reactor temperature, coolant temperature, and jacket temperature. The data set was obtained from empirical research. The classification has been conducted using a multikernel SVM. This article identified that the nonlinear classifier using the radial bias function results in an accuracy that is at least 22.08% superior to other methods.

Reliability Improvement by Fault-Tolerant Operation of NPC Inverter for Motor Driving

The high-reliability operation of three-level inverters is crucial to prevent equipment damage, process downtime, and economic losses. This paper investigates a three-level neutral-point-clamped inverter under all possible combinations of open-circuit and short-circuit faults and proposes a new postfault operation method without adding extra hardware. This method provides comprehensive solution for operations after single and multiple-device failures, increasing the inverter reliability by 24%. This paper classified post-fault modulations and uncovered previously unknown fault scenarios that can be addressed using the proposed control method. A new postfault modulation based on space vector modulation with virtual vectors is proposed. The feasibility of the proposed control method is verified by simulation and an experiment for one fault scenario in a three-level neutral-point-clamped inverter with a 3.73 kW motor load. This paper contributes to improving the reliability of three-level inverters.

Curvature measurement in profile bending - proposal and analysis of two application oriented measurement concepts

The curvature of bent profiles is the key parameter for the adaptation of process parameters in kinematic bending processes such as three-roll bending. Till today industrial standards include manual and time-consuming measurement procedures to determine the curvature, which can result in an increased process downtime. This is due to the fact that almost no efficient and also accurate curvature measuring systems exist for three-roll bending processes. This paper introduces two new application-oriented measurement systems. One is a mobile measuring device using a low cost inertial sensor, which appeals due to its easy applicability. The other is a measurement system mounted on a three-roll bending machine using laser scanners, with the purpose to provide the basis for a closed loop control in three-roll bending. The accuracy of both systems is investigated under industry-oriented conditions and error influences are analyzed. Concluding perspectives for improvements are discussed and the application potentials are evaluated.

Determination of manufacturing process failures priority under type 2 fuzzy environment: Application of genetic algorithm and Variable neighborhood search

The aim of this research is to propose the two-stage model to select a set of failures that need to be eliminated or reduced which leads to improved reliability and effectiveness of the manufacturing process in the automotive industry. In the first stage, uncertainties under the relative importance of risk factors and costs of manufacturing process downtime due to failure are modeled by type 2 fuzzy sets. The weights vector of risk factors is obtained by analytical hierarchical Process which is extended with type 2 fuzzy sets. Evaluation of risk factors at the level of each identified failure is based on failure mode and effect analysis which is widely used in practice. Determining the set of failures to be eliminated is set as a knapsack problem. The linear fitness function is defined as the ratio of the overall risk priority index and total costs. Maintenance costs incurred due to the realization of failure are limited by the available budget and in this knapsack problem are presented by a linear inequality. The solution to this problem is found by using the Genetic Algorithm and Variable Neighborhood Search. The model is verified with real-life data originating from automotive companies that exist in Serbia. Authors have managed to obtain suitable results on different knapsack problem instances. It is shown that the enhancement of the manufacturing process can be based on the proposed model.

Maintenance management in an industrial process: Case study in a company of the Industrial Pole of Manaus

In an increasingly competitive market, and given the current economic problems, it is necessary that companies invest in improving their production processes. Based on the above, the objective of this work was to develop research aiming to apply quality management tools that are widely applied in maintenance management in a production process of a company from the Manaus Industrial Pole to eliminate the process downtime due to problems with sticks with thick layer of paint and varnish. During the process, important steps were fulfilled within the research and starting from the initial diagnosis developed as well as the application of the mentioned tools, results were obtained that allow us to affirm the importance of the application of the maintenance management tools in this area, by obtaining satisfactory results both in quality gains and work time, translating into efficiency.

Enhance the Injection Molding Quality Prediction with Artificial Intelligence to Reach Zero-Defect Manufacturing

With the spread of the Industry 4.0 concept, implementing Artificial Intelligence approaches on the shop floor that allow companies to increase their competitiveness in the market is starting to be prioritized. Due to the complexity of the processes used in the industry, the inclusion of a real-time Quality Prediction methodology avoids a considerable number of costs to companies. This paper exposes the whole process of introducing Artificial Intelligence in plastic injection molding processes in a company in Portugal. All the implementations and methodologies used are presented, from data collection to real-time classification, such as Data Augmentation and Human-in-the-Loop labeling, among others. This approach also allows predicting and alerting with regard to process quality loss. This leads to a reduction in the production of non-compliant parts, which increases productivity and reduces costs and environmental footprint. In order to understand the applicability of this system, it was tested in different injection molding processes (traditional and stretch and blow) and with different materials and products. The results of this document show that, with the approach developed and presented, it was possible to achieve an increase in Overall Equipment Effectiveness (OEE) of up to 12%, a reduction in the process downtime of up to 9% and a significant reduction in the number of non-conforming parts produced. This improvement in key performance indicators proves the potential of this solution.

STRIDE‐based threat modeling and DREAD evaluation for the distributed control system in the oil refinery

Industrial control systems (ICSs) used to be operated in closed networks, that is, separated physically from the Internet and corporate networks, and independent protocols were used for each manufacturer. Thus, their operation was relatively safe from cyberattacks. However, with advances in recent technologies, such as big data and internet of things, companies have been trying to use data generated from the ICS environment to improve production yield and minimize process downtime. Thus, ICSs are being connected to the internet or corporate networks. These changes have increased the frequency of attacks on ICSs. Despite this increased cybersecurity risk, research on ICS security remains insufficient. In this paper, we analyze threats in detail using STRIDE threat analysis modeling and DREAD evaluation for distributed control systems, a type of ICSs, based on our work experience as cybersecurity specialists at a refinery. Furthermore, we verify the validity of threats identified using STRIDE through case studies of major ICS cybersecurity incidents: Stuxnet, BlackEnergy 3, and Triton. Finally, we present countermeasures and strategies to improve risk assessment of identified threats.

Model-Based Dynamic Categorization of Alarm Trip Points for Manufacturing Process Disruption Minimization

ABSTRACT Faults in manufacturing processes can cause disruptions and downtime, which adversely affects process efficiency and increases production costs. Process downtime can be reduced by monitoring manufacturing processes for faults, categorizing those faults, and prioritizing action. A new approach is proposed for minimizing function downtime, which involves monitoring system faults and generating a system response depending upon the status of the individual workstations in the manufacturing process line. The proposed model dynamically categorizes the faults considering the fault repair time, the weightage assigned to each workstation, and the process status, i.e. idle or running. The proposed model generates the threshold values for setting up trip points and categorizes critical faults into low, medium, and high values. The proposed model is tested using software simulations and an experimental test rig, and the results demonstrate that it can significantly reduce downtime in manufacturing processes. The model has been successfully implemented on a test rig and Asphalt manufacturing plant. In the test rig, 12% improvement is seen while on the Asphalt plant, 60% reduction is noticed in terms of start/ stops incidents, 60 minutes in downtime, and 800 kg of a valuable processed batch is saved in one week.

An Autonomous Electrical Signature Analysis-Based Method for Faults Monitoring in Industrial Motors

Rotating machines are widely used in industries, manufacturing, and oil and gas plants as a critical component for process availability. The inadvertent failure of these rotating machines causes significant process downtime and incurs higher repair costs and loss of revenue. These failures can belong to electrical, thermal, or mechanical fault categories. Early detection of all these faults is very much critical for avoiding complete failure of these machines and requires continuous 24/7 online monitoring using sensors or intelligent electronic devices (IEDs). However, an affordable low-cost and efficient online monitoring is desired to practice specifically for medium-voltage (MV) machines which have a larger install base. Electrical signature analysis (ESA) technology offers such flexibility and requires measuring just current and/or voltage at a motor control panel for machine health diagnosis, unlike vibration analysis (VA) requiring installation of sensors and its wiring on the machine. Furthermore, an intelligent, self-reliable, and autonomous ESA procedure is required for monitoring the machines due to the lack of ESA standards in practice unlike for VA. This article proposes a new autonomous electrical signature analysis (AESA)-based measurement technique implemented in IED, that is, Protection Relay, offering 24/7 online monitoring. The proposed technique implemented in Relay not only avoids dependence on standalone ESA hardware but also provides earlier detection of failures using peak and energy magnitudes computation approach at fault frequencies. To validate the proposed method, various tests were performed on the actual 1000- and 300-HP motors with/without mechanical faults in a machine repair shop and the results are discussed. The performance of the proposed method is also compared with the commercially available third-party ESA device results proving the efficacy.

Multi-stage and multi-objective decision-support tool for biopharmaceutical drug product manufacturing: Equipment technology evaluation

The emergence of single use technology (SUT) for equipment promises to be a revolutionary change in biopharmaceutical production. SUT equipment reduces non-value adding working time and provides operational flexibility. However, there are increased process risks such as possible contamination with leachable compounds or supply dependencies, as well as required redesign efforts. This work provides a multi-stage approach for decision-support for optimal equipment technology applied in a case study for a sterile filling line. Three design stages are identified. The first two stages explore the design landscape using the economic performance to show the general trends of favorable regions for equipment technology and the sensitivity to process parameters according to their relative importance. The last stage is more detailed but incorporates additional assessment indicators such as environmental impact and risks of process downtime. From an economic perspective SUT options are more favorable for short term and small-scale production, such as the case for preparing drugs for clinical trials. Traditional stainless-steel options had a significant economic advantage at large scale and long term production. The ranking of favorable options varied as other assessment factors, namely life-cycle CO2 emissions as well as risks in contamination and supply delay, were included in the analysis. The proposed methodology offers an agile decision-making tool, accounts for uncertainties across various design phases, and offers a systematic exploration of the available design landscape.

Математичні моделі верифікації ергатичних систем засобів залізничної автоматики

Issues related to the impact of operational, technical and inspection personnel on the safety of the use of ergatic railway automation systems have been addressed. In the conditions of interactive interaction of all types of personnel and equipment mathematical models for calculation of parameters of safety of use of such systems which consider their functioning in a regular and pre-auxiliary mode in various contexts of relations of human and technical factors are developed. It is established and proved that different modes of interactive interaction form a differentiated structure of calculation and logic schemes for assessing the safety of the use of ergatic control systems in railway transport. The functioning of ergatic control systems in railway transport has significant features associated with the inadmissibility of technological process downtime at infrastructure facilities. In this regard, in such systems there is an auxiliary control mode, in the framework of which blocking dependencies are completely or partially removed, which ensures the technological safety of the use of such systems. In recent years, a piecewise-defined character of mathematical models has been established on the basis of which a numerical safety assessment of the use of such control systems can be performed, which is due to the predominance of technical and human factors in different modes of functioning of such systems. At the same time, the development of forms of interactive interaction between different types of personnel in the active modes of operation of control systems requires a fundamentally new approach to safety assessment – taking into account the influence of differentiated human and integrated technical factors. In this regard, in this study, mathematical models are developed that allow you to take into account the interactive interaction of man and technology in the process of assessing the safety of the use of automated ergatic systems of railway automation. The graphical 3D dependencies built on their basis make it possible to carry out a predictive safety assessment of different systems in the process of their functioning during one working shift of operational, technical and audit personnel. In particular, the formation of mathematical models for calculating safety at different time intervals, which determine the auxiliary and standard operating modes of control systems, provides a differentiation of accounting for various factors when they dominate. An approach based on taking into account the influence of all types of personnel during their interaction with technology is reduced to considering different types of element connections in the calculation and logic schemes of the functional safety of systems, which reduces to different types of formulas in the final piecewise-defined expression for assessing the safety of using such systems.

Human immune system inspired framework for disruption handling in manufacturing Process

ABSTRACTDisruptions have a direct impact on the process downtime and efficiency pertaining to process industry. Anomalies, in any automated process, have an impeccable impact on consumer-centric values, high rejection of raw materials and cost, thus demanding special attention and techniques to be efficiently dealt with. Human immune system (HIS) presents an astounding example of such system, wherein the disruptions caused by viruses and bacteria are addressed by deploying B and T cells, which either destroys the pathogen (virus or bacteria) by killing it (Phagocytosis) or renders it harmless. Inspired from that, this research proposes a model analogous to HIS for industrial applications, to deal with disruptions. Addressing that, Immune based ontologies are developed for artificial immune system (AIS). This model works on the disruption detection and isolation with automatic response generation. Furthermore, the proposed model is capable of handling the disruptions in a dynamic way via weight assignment. The final reaction is assessed based on the assigned weights. Ontology was developed using Protégé (software). Experimentation was carried out in a controlled laboratory environment on a test bed by analysing 15 input/outputs (IOs) influencing the process downtime, system efficiency and consumer centric value.

Bayesian Network-Based Process Downtime Cost Determination of an Industrial Plant

Industrial plants utilize sensitive equipment to produce their products and meet their financial targets. Equipment downtime caused by power quality issues such as voltage sag affects production and entails cost hence poses a threat to their ability to deliver their financial objectives. This research aims to determine the response of industrial equipment to sag events and quantify the downtime cost caused by interruption in the production process. The study used the voltage tolerance curve to determine the individual equipment response to sag events and the Bayesian Network to establish the network structure of the production process. The probability of process interruption and the associated downtime losses was computed using a mathematical software. The research shows a strong relationship between the equipment’s response to voltage sag events and the production downtime cost and highlights the importance of the immunity of equipment to voltage sags.

Monitoring of a machining process using kernel principal component analysis and kernel density estimation

Tool wear is one of the consequences of a machining process. Excessive tool wear can lead to poor surface finish, and result in a defective product. It can also lead to premature tool failure, and may result in process downtime and damaged components. With this in mind, it has long been desired to monitor tool wear/tool condition. Kernel principal component analysis (KPCA) is proposed as an effective and efficient method for monitoring the tool condition in a machining process. The KPCA-based method may be used to identify faults (abnormalities) in a process through the fusion of multi-sensor signals. The method employs a control chart monitoring approach that uses Hotelling’s T2-statistic and Q-statistic to identify the faults in conjunction with control limits, which are computed by kernel density estimation (KDE). KDE is a non-parametric technique to approximate a probability density function. Four performance metrics, abnormality detection rate, false detection rate, detection delay, and prediction accuracy, are employed to test the reliability of the monitoring system and are used to compare the KPCA-based method with PCA-based method. Application of the proposed monitoring system to experimental data shows that the KPCA based method can effectively monitor the tool wear.

Probabilistic forecasting of sudden power supply interruptions of electromechanical and heat-power equipment of regionally isolated electrotechnical complex

There is considered probabilistic forecasting of the most important characteristics of sudden power supply interruptions — damage and downtime of electromechanical and heat-and-power equipment — for process plants typical for installations of a regionally isolated electrotechnical complex. These specifications are considered as random functions of the power supply interruptions duration. A feature of the specifications considered is the presence of finite jumps when the power supply exceeds the critical values. Generalized theoretical equations are obtained for continuous functions of mathematical expectation and variance of process downtime and damage when there is any number of finite jumps of certain implementations and there is consideration of the discontinuity points as random variables with different distribution laws. The obtained specifications are used to optimize the cost of creating power supply systems considering their reliability.

Evaluation and validation of an inline Coriolis flowmeter to measure dynamic viscosity during laboratory and pilot-scale food processing

Abstract Inline process analytical technology tools are increasingly employed in industry as they facilitate real time process parameter monitoring and optimization of product quality. Use of an inline viscometer in milk powder manufacture allows greater process control by monitoring the dynamic viscosity of the process concentrate pumped to the spray drier. In this study, an inline Promass I300 was used to measure the dynamic viscosity of both a Newtonian fluid (10–60% w/w sucrose solutions) and a non-Newtonian fluid (10–40% w/w skim milk concentrate) at laboratory scale at 25 °C. Validation of the instrument for measurement of a non-Newtonian fluid was completed at pilot-scale under similar operating conditions. Coefficient of determination (R2 = 0.99) was obtained between inline and offline viscosity measurements for both Newtonian and non-Newtonian fluids investigated. This study demonstrated the potential of an inline Coriolis flowmeter for rapid and accurate measurement of dynamic viscosity during processing of dairy streams. Industrial relevance Control of milk concentrate viscosity during the manufacture of milk powder is critical to reduce energy consumption, reduce fouling and meet powder functional properties e.g. dispersability, flowability, particle size and powder hydration. The use of inline instruments for continuous monitoring of viscosity offers economic benefits to the producer as it can contribute to reduction in commonly encountered process issues such as fouling, blocking of nozzles and evaporator, which can result in process downtime and in extreme cases, product rework. Inline instruments can thus be employed to improve process control and reduce production waste during the manufacture of dairy powders.

Data-Driven Optimization of Processes with Degrading Equipment

In chemical and manufacturing processes, unit failures due to equipment degradation can lead to process downtime and significant costs. In this context, finding an optimal maintenance strategy to ensure good unit health while avoiding excessive expensive maintenance activities is highly relevant. We propose a practical approach for the integrated optimization of production and maintenance capable of incorporating uncertain sensor data regarding equipment degradation. To this end, we integrate data-driven stochastic degradation models from condition-based maintenance into a process level mixed-integer optimization problem using robust optimization. We reduce computational expense by utilizing both analytical and data-based approximations and optimize the robust optimization parameters using Bayesian optimization. We apply our framework to five instances of the state-task-network and demonstrate that it can efficiently compromise between equipment availability and cost of maintenance.

Inactivation of Foodborne Bacteria Biofilms by Aqueous and Gaseous Ozone.

In this study, the efficacy of treatments with ozone in water and gaseous ozone against attached cells and microbial biofilms of three foodborne species, Pseudomonas fluorescens, Staphylococcus aureus, and Listeria monocytogenes, was investigated. Biofilms formed on AISI 304 stainless steel coupons from a mixture of three strains (one reference and two wild strains) of each microbial species were subjected to three types of treatment for increasing times: (i) ozonized water (0.5 ppm) by immersion in static condition, (ii) ozonized water under flow conditions, and (iii) gaseous ozone at different concentrations (0.1–20 ppm). The Excel add-in GinaFit tool allowed to estimate the survival curves of attached cells and microbial biofilms, highlighting that, regardless of the treatment, the antimicrobial effect occurred in the first minutes of treatment, while by increasing contact times probably the residual biofilm population acquired greater resistance to ozonation. Treatment with aqueous ozone under static conditions resulted in an estimated viability reduction of 1.61–2.14 Log CFU/cm2 after 20 min, while reduction values were higher (3.26–5.23 Log CFU/cm2) for biofilms treated in dynamic conditions. S. aureus was the most sensitive species to aqueous ozone under dynamic conditions. With regard to the use of gaseous ozone, at low concentrations (up to 0.2 ppm), estimated inactivations of 2.01–2.46 Log CFU/cm2 were obtained after 60 min, while at the highest concentrations a complete inactivation (<10 CFU/cm2) of the biofilms of L. monocytogenes and the reduction of 5.51 and 4.72 Log CFU/cm2 of P. fluorescens and S. aureus respectively after 60 and 20 min were achieved. Considering the results, ozone in water form might be used in daily sanitation protocols at the end of the day or during process downtime, while gaseous ozone might be used for the treatment of confined spaces for longer times (e.g., overnight) and in the absence of personnel, to allow an eco-friendly control of microbial biofilms and consequently reduce the risk of cross-contamination in the food industry.

Biofiltration of feedwater to control organic fouling of low pressure membranes

ABSTRACTThe application of low pressure membrane (LPM) technology, i.e., ultrafiltration and microfiltration, for drinking water treatment and secondary effluent reclamation has increased rapidly over the past two decades. A major drawback in the use and operation of LPM processes is fouling due to the deposition of organic matter present in the feedwater on the membranes which leads to reduced water productivity necessitating membrane cleaning, process downtime and eventual membrane attrition. Pretreatment of feedwater using biological processes (particularly slow sand filtration and biological activated carbon) has been investigated as a simple and cost-effective means to control organic fouling of the membrane. In this paper, membrane fouling by the organic matter in drinking water and secondary effluent, the mechanisms of biological treatment systems, and the effectiveness of biological processes for fouling reduction are reviewed. It has been demonstrated that biological processes have great potential for controlling the organic fouling of membranes.

Sensor deployment for variation diagnosis considering heterogeneity in single-station multi-step assembly processes

Sensor deployment to robustly monitor operation parameters is the cornerstone for timely diagnosing variations that lead to product quality defects. A sensor system with optimal placement can help manufacturers improve product quality and reduce process downtime. However, current literature lacks investigation in methodologies that consider the mechanism of variation propagation, heterogeneity among sensor and variation properties, and multi-objective optimization involved in the sensor layout for variation diagnosis in an assembly process. In our approach, sensing information from fixture error and part machining error is integrated into a step-indexed state space model based on an understanding of variation propagation mechanism for the stream of assembly set variation described in a single-station multi-step assembly process. The effectiveness of such sensor system is quantified by some indexes, which are used to quantitatively characterize the diagnosability. Based on digraph theory, a quantitative fuzzy binary graph-based approach is developed to model the cause–effect relationship between part variations and sensor measurements considering the heterogeneous properties of sensor and variation, thus an improved shuffled frog-leaping algorithm is proposed to determine the sensor types, numbers, and locations to achieve full diagnosability. An example for the differential assembly illustrates the methodology.