System Behavior¶
In this section, the system behavior around various topics, for example, data ingest, is described a little more in detail. This should help you to understand the application a little bit better.
System Limits¶
In this section, the current system limits of Insights Hub OEE are described.
| Type | Limit | Description |
|---|---|---|
| Time Range | 90 days | Max. time range for KPI calculation is 90 days |
| Calendars | 100 | Create up to 100 calendars per tenant |
| Calendar Events | 5000 | Create up to 5000 calendar events per tenant |
| Measure Collection | 100 | Create up to 100 measure collections per tenant |
| Measures | 1000 | Create up to 1000 measures per measure collection |
| Product collections | 100 | Create up to 100 product collections per tenant |
| Products | 10000 | Create up to 10000 products per product collection |
| Quality Codes | 1000 | Create up to 1000 quality code definitions per tenant |
| Reason Trees | 100 | Create up to 100 reason trees per tenant |
| Reasons | 2500 | Create up to 2500 reasons per reason tree |
| Status Mappings | 200 | Create up to 200 status mappings per tenant |
| Time Categories | 300 | Create up to 300 time categories per tenant |
Visualization¶
Overview Page empty while other pages show data¶
Sometimes, it might appear that the OEE application isn't working properly when the Overview page is empty while other pages, e.g. quality or availability, show data. This behavior is related to the following principle:
The overview page only shows data when all KPIs related to the OEE calculation, meaning Availability, Performance and Quality, can be calculated. If one of the KPIs cannot be calculated, then the overview page will be shown as empty, whereas the sub-pages for the other KPIs would still be available. The following example will illustrate this:
You are receiving status information and counters from the machine, but unfortunately, something with the product code ingestion is not working as expected. The system will now behave as follows: The availability page, the downtime analysis page, and the downtime management page will be shown as the status can be evaluated. The quality page can also be shown, since the quality can be calculated based on the counters. The performance page cannot be displayed because the theoretical output and therefore performance cannot be calculated. Following that, the overview page and the loss overview page can also not be displayed since OEE and TEEP can't be calculated.
So, whenever you see that some pages are being shown but not all of them, you should investigate in the data related to the main KPI sub-pages, availability, performance and quality.
Machine status not automatically loaded for longer time ranges¶
Whenever you calculate OEE and the other KPIs for longer time ranges, it means that a lot of data needs to be processed and also visualized. In many times, when there are too many state changes in the time range, the chart is not able to display this properly anymore. To improve the application performance, Insights Hub OEE now decides based on the number of status changes in the selected time range, if the machine status will be displayed on the overview or availability page. If the application detects more than 1000 status changes in the period, the machine status charts will remain empty and display a message. This ultimately reduces the application loading times since less data needs to transferred in the beginning. In any case, the application will show a status distribution on the availability page and the machine status can still be fetched through a button click if really needed.
Data Ingest¶
Data Ingest and post-processed data¶
Insights Hub OEE fetches the data from the time series storage and stores it in its own database to be able to contextualize the data properly and to perform fast OEE calculations. To fetch the data, Insights Hub OEE uses the subscription services of Insights Hub, which notifies the application when new data was ingested in the asset and then the application starts to read only the new values of the asset and stores it in the OEE database. This is only performed one time.
So, whenever you overwrite data in the past, which has already been read by the application, it will not be considered by OEE anymore, as for this timestamp the application already fetched data. If you are post-processing data and want to use it in OEE, it is recommended to set up the post-processed variable in a separate aspect, since it is then separately fetched from the other variables that can be stored in a single aspect. In general, the idea of an aspect is, that the data from the different variables arrives at the same time and can therefore also be read in one go. If the data is coming from different sources that are not in sync, it is recommended to have separate aspects for the different variables.
Variable Switch in the machine or line configuration¶
When you set up an asset in Insights Hub OEE, you need to map the operands in the application, e.g., status, counters, and alike to asset variables from the assets you configured in Asset Manager. Once you succeeded this and everything is set up properly, Insights Hub OEE will start to pull the data from the time series database to its own database. This is done for performance reasons when you want to calculate OEE and the other relevant KPIs.
At some point in time, you might consider switching a mapped variable. Before you do this, you should know what the consequences are and what a suitable strategy might be. If you start to touch the configuration and change a variable, this can lead to issues moving forward. Once you performed the switch of the variable, Insights Hub OEE will start to pull the time series values from the new variable and store it in the database. This can lead to inconsistencies when you want to calculate and view KPIs that includes data from both worlds – from the old variable and the new one. For instance, you pulled data from one status variable with a specific mapping and then choose another status variable to pull the data from. Both variables have different status mappings. Since only one status mapping can be active at a time, the old status codes might not be mapped in the new status mapping and therefore queries might return errors when you start to request data from an overlapping period.
To avoid these kinds of inconsistencies, it is recommended to remove the asset from the Insights Hub OEE configuration entirely and set it up again with the new variable. When an asset is set up in Insights Hub OEE, the application will pull data from the last two weeks if there is data available. With this, you will avoid that OEE will try to calculate KPIs based on a data mix of the old and new variable.
Configuration¶
Mode changes on different variables¶
The manual input related capabilities add another level of complexity to the application. The configuration now supports multiple combinations of manual and automatic value capture. This also comes with a couple of restrictions:
- Reject Reasons: Once one manual entry was created for this asset the reject reason collection can't be changed anymore. If there hasn't been a collection assigned yet, it is still possible to add one when manual entries were already created.
- Product Collection: Product collections can be changed when needed
- Machine/Line Status: Once one manual input was received for this asset, the status mode can't be changed anymore - neither from manual to connected, nor from connected to manual.
In general, changes can be made from connected mode to manual mode but not the other way around, since data can't be fetched from the past.