Transcription
Plantweb Optics AnalyticsProduct Data SheetOctober 2020Plantweb Optics Analytics Modeling StudioTransform data into knowledge and profits.In the manufacturing and energy industries, new challengesare pushing managers to optimize and improve operationalefficiency. From the obvious challenge of managing a finiteresource to the ever changing environmental regulationsand the competitive pressures, it is easy to understandwhy industry leaders are continually striving to find ways tomaximize productivity. In order to sustain the performanceand profitability of their plant operations, they have to makeinformed decisions which are dependent on the ever increasingmassive amount of data scattered in many sources. The effortsof collecting, cleaning, aggregating and analyzing such datais laborious without the guarantee of any results.To address this challenge, Emerson is pleased to offer PlantwebOptics Analytics Modeling Studio which takes away the burdenof processing millions and trillions of data measurementssuch as sensor readings, lab data, simple and complex events,and automatically extract knowledge and predictive models,identify optimization opportunities and behavioral patterns.
Plantweb Optics Analytics Modeling StudioPlantweb Optics Analytics Modeling Studio is an out of the-boxprocess data analytics software that allows endusers withoutextensive domain expertise to: Inspect, clean and transform raw data into informationand knowledge. Create predictive models that can be deployed in real time. Eliminate erroneous data and pinpoint problem areas. Perform proactive decision making to prevent abnormalconditions and identify opportunities. Recognize behavioral patterns and correlate data with events.BenefitsOctober 2020MethodologyAutomated Knowledge CapturingThe procedure adopted by Plantweb Optics Analytics ModelingStudio consists in acquiring raw data from the plant, cleaningthe acquired dataset, building models and interpreting them.This process can lead to the extraction of highly valuableknowledge in form of expert rules and hybrid models thatdetect special operating modes and states, abnormal plantbehavior, and best operating conditions. Plantweb OpticsAnalytics Modeling Studio automates knowledge capturingwithout the need to interview the experts and document theknowledge manually.Beyond Analytics Maximize efficiency and profitability to gaincompetitive advantage. Discover new opportunities for time and money savings. Gain more flexibility in anticipating changes.Not only can Plantweb Optics Analytics Modeling Studio helpyou uncover hidden process knowledge, build predictivemodels, correlate data but it also allows you to determinethe relationship between the dynamic behavior of the plantand complex events. Make efficient decisions with confidence.Online DeploymentUncovered knowledge and models built in Plantweb OpticsAnalytics Modeling Studio can be deployed online to deliverreal-time visibility and insight into operations for optimizedperformance management, maximized throughput, efficientmanagement of abnormal conditions, and tight control ofoperating costs.www.emerson.com2
Plantweb Optics Analytics Modeling StudioOctober 2020Realtime deployment of AnalyticsPPReal-time DataDecision TreeArtificial NeuralNetworkPRoot CauseAnalysisRulesComplex EventsDynamic TargetsData ValidationProject StudioRegressionDLLClusteringPWorkflowsModeling StudioHistorical DataPExport DLLKnowledge to set upexpert rulesTransformationData Driven AnalysisFeaturesMachine LearningMachine learning generates models that can capture and mimicthe behavior of a plant. These models can be used to predictfailures and recognize operating modes. Plantweb OpticsAnalytics Modeling Studio offers several algorithms for machinelearning such as Neural Networks, Decision Trees, LinearClassifier, and others.www.emerson.comOptimizationOptimization consist of finding the best available values of anobjective function given a set of constraints. With this feature,Plantweb Optics Analytics Modeling Studio helps operationsmaximize productivity and optimize costs.3
Plantweb Optics Analytics Modeling StudioOctober 2020WorkflowWorkflow in Plantweb Optics Analytics Modeling Studiois a powerful feature that allows you to build and automatesequential process made of data analysis, learning, clustering,and modeling. Once the workflows are built, they areautomatically memorized and can be later deployedgenerically and applied on multiple data sets withoutthe need of human intervention.Pattern RecognitionThis feature allows the recognition of patterns and regularitiesin data. It consists, for example, in identifying operatingmodes by detecting clusters with related data in scatterplots or in discovering recurrent sequences in alarms andevents databases.Graphical EnvironmentUnlike traditional data analytics technology, Plantweb OpticsAnalytics Modeling Studio offers a highly intuitive anduser-friendly environment with a Microsoft look and feelfeaturing drop down menu, drag and drop graphical iconswithout the need of scripting or programming.ExplorationThe exploration step is important to clean the input data setbefore performing advanced analysis. Plantweb Optics AnalyticsModeling Studio provides different tools to explore data such ascleaning, sampling, smoothing, correlation, and others.ModelingPlantweb Optics Analytics Modeling Studio offers toolsto perform descriptive and predictive models based on actualdata. Descriptive models are used to better understandthe process behavior by identifying key variables and theirrelationships, while predictive models are used to predictthe system behavior using hybrid models and combiningmultiple techniques.www.emerson.com4
Plantweb Optics Analytics Modeling StudioTime SeriesTime Series features in Plantweb Optics Analytics ModelingStudio allow transforming time series and analyzing theirbehavior for trends, seasons, and correlations with past values.Time series models include AR, MA, ARMA and GARCH modelsas well as time series Neural Networks.October 2020ApplicationsAutomatic Identification of Plant Operating Modesand StatesPlantweb Optics Analytics Modeling Studio combines clusteringmethods with classification and modeling to build expert rules.These expert rules can be deployed online to automaticallyidentify plant operating modes or to detect and predictabnormal situations, for example drop in RON(Research Octane Number) and high energy consumption.KPI Dynamic TargetingPlantweb Optics Analytics Modeling Studio offers methodsto set dynamic targets that consider process behavior in termsof operating modes, equipment availability, valves line-ups,special process scenarios, and others.Sensors ValidationVisualizationPlantweb Optics Analytics Modeling Studio embeds differenttools offering efficient data visualization such as scatter plot,line chart, pie chart, area chart, box plots, radar chart,and polar chart.Plantweb Optics Analytics Modeling Studio includes a sensorvalidation module that identifies the faulty sensors andvalidates the values. In addition, Plantweb Optics AnalyticsModeling Studio enables determining soft sensors for complexmeasurements such as analyzers or labs.OptimizationBased on the system configuration, Plantweb Optics AnalyticsModeling Studio provides users with appropriate configurationfor an optimal system performance and reduced downtime.Asset Performance ModelingPlantweb Optics Analytics Modeling Studio allows modelingthe performance of the asset as a function of key process inputs.This model might be used to identify the reference performancecurve to be used to identify performance gaps.www.emerson.com5
Plantweb Optics Analytics Modeling StudioOctober 2020Alarm AnalyticsBatch AnalyticsThe alarm analytics module combines several data analyticsand advanced techniques allowing end users to extractknowledge from alarm data and go beyond classical alarmmanagement. This module allows the user to:Plantweb Optics Analytics Modeling Studio MultivariateAnalysis feature monitors the production process and deliverskey insights into process variability issues. Plantweb OpticsAnalytics Modeling Studio batch specific algorithms help with: Audit the performance of alarm systems and pinpointnuisance alarms such as frequent, chattering, stale andstanding alarms. Identify alarm correlations and clustering to removeredundant alarms and reduce alarm floods. Use pattern recognition tools such as sequence mining andassociation rules to predict the future behavior of alarms forincreased safety and reliability. Correlate alarm data and process data for a bettertroubleshooting and root cause analysis.www.emerson.com Maintaining product quality by preventing waste on highquality batches. Identifying the variations in batches by monitoring key batchvariable contributors in real-time. Predicting process faults early on and determining the rootcause behind any process deviations.6
Plantweb Optics Analytics Modeling StudioOctober 2020Minimum Hardware/Software RequirementsOperating SystemWindows 7, Windows 8,Windows 8.1, Windows 10Windows Server 2008, Windows Server 2008 R2,Windows Server 2012, Windows Server 2012 R2,Windows Server 2016Processor Speed3 GHz (higher recommended)3 GHz (higher recommended)Memory8 GB16 GBDisk Space10 GB10 GB 2020, Emerson. All rights reserved.The Emerson logo is a trademark and service mark of Emerson Electric Co. All other marks are theproperty of their respective owners.Contact Uswww.emerson.com/contactusThe contents of this publication are presented for informational purposes only, and whilediligent efforts were made to ensure their accuracy, they are not to be construed as warrantiesor guarantees, express or implied, regarding the products or services described herein or theiruse or applicability. All sales are governed by our terms and conditions, which are available onrequest. We reserve the right to modify or improve the designs or specifications of our productsat any time without notice.
Plantweb Optics Analytics Modeling Studio ctoer 2020 Plantweb Optics Analytics Modeling Studio is an out of the-box process data analytics software that allows endusers without extensive domain expertise to: Inspect, clean and transform raw data into information and knowledge. Create predictive models that can be deployed in real time.
