About the application Pythia

Pythia is an application that is independent of any other brand of software or sensoring hardware from any company.

 

Pythia can be situated on top of process monitoring systems like ERP, MES or SCADA or similar applications, but is completely independent of them. It merely uses the results to predict events, based on historical data that were collected earlier.

 

Pythia is created for identifying, designing, and/or developing a computing system environment and protocols to facilitate interactions between analysis tools and common delivery mechanisms in use or under development. These computational techniques promise a significant cost savings and increased productivity that will maximize knowledge gained from more costly laboratory experimentation, and will provide analytical input to future automated risk assessment and reduction applications for service providers and end users.

 

The basis of Pythia are “Events”, “Processes”, "Installations" and "Objects".

Within Pythia you create a model of your core production or service process which is under investigation.

Processes and installations are the items of interest, and they consist of "Objects". Installations and/or services are the asset of your business. They can be latterly anything as far as Pythia is concerned.

(In this document an “Installation” can also be read as a “Process”, if not explicitly stated otherwise.)

 

Pythia; Process Behavior:

Some definitions to start with:

  • Installation or Service; highest level of the item of interest.

  • Object; Installations or services are build up from the constituent objects.

  • Parameter or Key Production/Service Indicators; any measurement in an object that can be recorded by means of sensoring technique.

  • High and low tolerance; the values that define the highest and lowest measurement between the parameter measurement that is allowed and valid.

  • Deviation factor; a percentage which allows the measurement to be offside of the “Event” value.

  • Behavior; the total of measurements of all parameters in all objects that are in an installation or service.

  • Pattern; A limited dataset of behavior measurements preceding the actual moment that an event happened.

  • Event (or Failure); Every failure is an event; but not all events are failures. Events happen on a certain moment in time and the dataset is stored in the EventPattern database. (see Pattern)

  • Internal Object parameter(s); these are the parameters that can be influenced by the object itself like its operating pressure or temperature.

  • External Object parameter(s); these are the “environmental” circumstances that an object has to run in like climate, geotechnical or other influences.

     

    Pythia defines “Process Behavior” as the “Presentation of the process data during a certain timeslot”.

    Pythia uses all kinds of data that is produced in a process or service during a certain timeslot. It takes the measurement and timestamp of a process parameter (like delivery time, cooling temperature or pressure) and stores it in its database. Pythia combines and determines the correlation of the internal and external parameters to make it possible to take all possible causes in effect. All this information together forms the “Behavior” of an installation, a service or an object.

     

    Pythia; Failures And Events:

    In general, a failure is considered to be a "negative" event; it causes a "loss". All events are caused or triggered by one or more specific combinations of situations or states that is a system in. On the other hand; a positive event is a "gain". The purpose of the installation or process. Pythia does not distinguish between “Loss” or “Gain”; it just uses “Events”.

    Any type of production facility is thought of as being an "Installation" or a “Service”. So, by generalizing this as a concept, it is possible to create a model of virtually every kind of productive organization, regardless the real (or virtual) product/service they deliver.

     

    During production cycles, every installation facility is in any sort of "state" and the most important ones are closely monitored, because they are the main source of information that tells you whether the production cycle is "in control" and the produced goods are within acceptable quality standards. In general, most of this information data is gathered by sensors, but data may also come from (historical) data(bases).

     

    This brings us into the world of "Tolerances".

     

    Tolerances are everywhere and always present. They are the mean measurement of "Quality" because they define whether a situation, product, good or services is acceptable to be used by or sold to always more demanding customers.

    Tolerances have to ensure that all the produced goods are continuous, consistent and reliable without variances outside these tolerances.

    But processes have tolerances to! Within Pythia, every process parameter has its own tolerances, and these can individually be set and maintained. This enables Pythia to monitor whether a process is “in control” or not. When subsequent readings are reaching the tolerance borders Pythia will detect this and will issue warnings to be send out before the state of event (failure) is reached.

     

    The chain of steps.

     

    Customers and users regard a product as "Acceptable" when they receive their purchase within certain limits of acceptance. What these "Limits of Acceptance" are is greatly depended of the nature of the client, and the designated use of the services or goods they expect, buy or use. But the client is always the last part of the chain. This means that all the steps before that are due to influence the satisfaction of the client. Thus all these steps can be the cause of quality degradation in the eye of the client.

    And this is the point that Pythia kicks in!

    Every step in any production process can be identified as an (more or less) individual set of activities in the production/service system. And all these individual parts have their own "Key Production Indicators". These indicators are the items that define the "state" that this step is in. By guarding these indicators and by controlling their values (inside or outside tolerances), it is possible to ensure that every production/service step is performed well. But what happens when the quality of an outcome of a step is good, but the end control of the product says "NO!" Obvious there is a problem when individual steps deliver results that appear to be good, but in the end turn out to result in poor end products or services. This needs a Root Cause Analysis.

    Root Cause Analysis is the art of recognizing what is the cause and effect of any combination of "State" and "Event" (regardless good or bad) in any type of installation or service.

     

    Because this application records all of the pre-defined "Key Production Parameters" (Internal and External) and also records all the failures or any deviation of produced products, it creates the possibility of analyzing the steps from start to end, i.e. the behavior of the process.

     

    Pythia; Pattern Recognition, catch the event:

    So what you do is select an "Event of Interest" and determine the period of measurements that precede this moment. In this way Pattern Recognition makes it possible to detect certain issues, due to the fact that the combination of parameters lead to this failure. So, when you have selected an "Event of Interest", Pythia has to create the pattern of measurement points that go in advance of the moment of event. Now Pythia sets up the pattern. This needs only to be done once for any "Event of Interest" and can be done as a function within the application by selecting the event, determine the pattern points and the treshold factor (between 0 and 1) and then generate the pattern. Now, when running the analysis again, Pythia will re-examine the whole timeslot and sets a marker on every moment the pattern is found. This enables the process analyst to detect weather the event has been occurred earlier and is thereby able to reconstruct the circumstances under which an event will appear.

     

    Pythia; Predictitive Analytics, real-time and on-line:

    Predictive analytics is the science about being able to foresee the (near) future, based on historical and new incoming data in a process and its behaviour.

    Pythia performs predictive analytics by “learning” from the past by comparing each new incoming dataset by the patterns in the database. Pythia has to connect to the installation’s sensors directly thru a so-called API or a connection to a SCADA system. By evaluating each new incoming dataset from the sensors against the patterns in the event database, Pythia is able to detect events arising as they come in, therefore Pythia is able to “see” and recognize the probability of an known event to happen. We are developing the function in which Pythia is also able to determine when there is a deviation from the “normal” process behaviour, even when no pattern is present.

     

    Pythia; Root Cause analysis, learn from the past:

    Root cause analysis (RCA) is a method of problem solving that tries to identify the root causes of faults or problems. A root cause is a cause that once removed from the problem fault sequence, prevents the final undesirable event from recurring. A causal factor is a factor that affects an event's outcome, but is not a root cause. Though removing a causal factor can benefit an outcome, it does not prevent its recurrence for certain.

    Because Pythia has all the relevant data in its database, it can present the analysts with the chain of events that preceded the “Root Cause”. The next manager in line can then take appropriate actions to solve the problem(s).

     

    Pythia; Predictitive Maintenance:

    Predictitive Maintenance techniques are designed to help determine the condition of in-service equipment in order to predict when maintenance should be performed. This approach promises cost savings over routine or time-based preventive maintenance, because tasks are performed only when warranted.

    Pythia performs an analysis in (semi)real time and online by reading all the relevant (sensor)data from sensors and present detailed information about the status that a process is in and weather any event might occur.

     

    Pythia; Reliability Centered Maintenance:

    Reliability Centered Maintenance (RCM) is a process to ensure that assets continue to do what their users require in their present operating context.

    It is generally used to achieve improvements in fields such as the establishment of safe minimum levels of maintenance, changes to operating procedures and strategies and the establishment of capital maintenance regimes and plans.

    With Pythia you can implement RCM to minimize maintenance costs. Pythia informs the maintenance officer that certain processes are at risk because the behavior of the installation, or even an individual object, is triggering a warning signal.

     

    Pythia; Process Stability and Process Capability:

  • Process Stability: The process produces its product in acceptable quality, but not necessary in acceptable quantity.

  • Process Capability: The process is able to produce the product in sufficient quantity, but not always in the right quality.

     

    Process Stability and Process Capability are both extremely important aspects of any manufacturing process. Often the concepts behind process stability and process capability and the relationship between them are misunderstood. Pythia attempts to clarify the relationship between them. Defining Process Stability and Process Capability, Process Stability refers to the consistency of the process with respect to important process characteristics such as the average value of a key dimension or the variation in that key dimension. If the process behaves consistently over time, then we say that the process is stable or in control. Within Pythia this is done by applying Pareto analysis and stability determination.

    Pythia offers all the necessary information to the management and process analysts on all objects that is needed to determine whether the process as a whole is stable and/or capable to perform its tasks.

     

    Pythia; Overall Equipment Effectiveness (O.E.E.):

    OEE (Overall Equipment Effectiveness) is the gold standard for measuring manufacturing productivity. Simply put – it identifies the percentage of manufacturing time that is truly productive. An OEE score of 100% means you are manufacturing only Good Parts, as fast as possible, with no Stop Time. In the language of OEE that means 100% Quality (only Good Parts), 100% Performance (as fast as possible), and 100% Availability (no Stop Time).

    Pythia offers all the necessary information to the management and process analysts to determine the effectiveness of the underlying process in order to enhance productiveness and gain better process quality.

     

    Pythia; An enhancement on ERP, MES and SCADA systems:

    Pythia is the link between the more administrative oriented ERP, MES and SCADA systems and the actual organization’s process.

    ERP systems can deliver a lot of data that an analyst can use with Pythia to find correlations in any process investigation. Pythia can READ data from ERP systems and can use these as extra means for process analysis.

    MES and SCADA systems are able to produce loads of (production)data that reflect the state of an installation or a process. Pythia can dive into this (big)data and turns it into BIG INFORMATION. Pythia can be used as an add-on on any MES or SCADA system thru direct connection to their databases or indirect thru data sharing by means of file transfer. In this way Pythia is fully independent of any legacy system already installed in your organization.

     

    Recapitulation:

     

    Pythia’s Unique Selling Points:

    Pythia offers a wide range of possible use by the way it operates on processes because it performs an in-depth data analysis on (big)data. It presents the results in the appropriate level of detail in easy-to-read graphics to the user with a specific interest and responsibility like 

    Open process modeling feature for the:

  • Data analyst

  • Process analyst

  • Process developer

  • Risk manager

  • Quality manager

  • Maintenance officer

     

    It offers functionality in terms of:

  • Predictive Analytics

  • Root Cause Analysis

  • O.E.E. analysis

  • Six Sigma analysis

  • Oscillation analysis

  • Early Warning System

  • Process control and monitoring

  • Quality control

  • Risk and damage control

  • Asset management

  • Energy consumption management

  • Predictitive Maintenance

  • Risk Based Maintenance

  • Reliability Centered Maintenance

  • Pattern Recognition

  • Bump test

  • Parameter/Event correlation

  • What-If analysis

  • Testing processes

  • Before-After analysis

     

    Applicable in (amongst others):

  • Service organisations

  • Distinct production facilities

  • Continueous production facilities

  • Testing facilities

  • General building management

  • Domotica

  • Wastewater management

  • Water and soil management

  • Cooling or heating facilities

  • Agricultural environments

  • Demographic environment

  • Research laboratories

     

    Conclusion:

     

    Pythia in your organisation:

    Pythia is an independent either stand-alone or web based application. It takes the data by reading a text file (*.txt or *.csv) for either the operational data or the event occurrences and process these directly into the analysis per installation or object. These files can be output by any legacy system, as long as the record structure is maintained.

    We are currently working on direct connected sensors that feed Pythia real-time and on-line thru an API to get to any problem as quick as possible and depending on the market needs.

    We are also investigating the possibility and necessity of the development of Pythia to be a dedicated “Embedded Application” that can be built into an installation all together.

    Warning messages are generated based on all kind of triggers and can be sent directly to any receiver’s mail address or any other means of communication.

    Pythia can be used for your own learning aims, your production process monitoring or as a service to your clients. Especially when your organization delivers maintenance services to other parties, Pythia can assist you by its early warning function even before your client knows that something bad is bound to happen!

     

    Benefits:

    In short:

  • Cost reduction

  • Enhancement of any existing ERP, MES or SCADA application

  • Early warning system

  • Asset management

  • Risk management

  • Maintenance management

  • Organisation enhancement

  • Process enhancement

  • Quality enhancement

  • Analysis tools

  • Ease of use