Core Concept behind recording process and meta data for process tags
Litmus Production Record Database allows to record, beside Production Records for production events, also data for process tags similar to a Historian but intended for a small selection of tags used for production event monitoring or production record event root cause investigation instead of long term data storage.
It has to be understood, that Litmus Production Record Database is not a Historian and should not be seen or used as a replacement to a true historian.
Similar to the recording of Production Record event data, process as well as meta data for process tags are also using the EAV database model to achieve this high level of flexibility.
In the case of process data, the entity is a unique combination of a tag and its data source also called device. This concept is borrowed directly from Litmus Edge, as Litmus Edge will be the most commonly but not only data source from where these data are send from.
Users are then able to define the attributes they want to record for each tag.
But unlike Litmus Edge, which has a normalized data object, Litmus Production Record Database does not enforce this but it does encourage to use a similar approach.
As an example, let us imagine a user has 4 tag entities, they are able to define for each tag the attribute "Value" for example, which is the same as having a normalized data object.
But users do have the freedom and maybe the need to call this attribute for each tag differently like "Value_Tag1", "Value_Tag2" and so on.
Using the EAV model, users are also not restricted to "just" the value and timestamp to be recorded. They are able to define additional attributes like "5 Minute Average", "User Comment" or "PLC Timestamp" to be recorded for the tag they define.
Additionally, can each attribute be classified as either as process data or as meta data.
The difference is, that process data record data based on a regular frequency for example every 5 seconds. These data will also get a retention period like mechanism applied to, allowing to control how much data is online available.
Values for attributes set as meta data will and are expected to be recorded very infrequently. An example for a meta data attribute could be "Lower_Limit", "High Limit", "Engineering Unit" or "Process Area". The biggest difference to the process data is, that meta data will not get the retention period applied like process data and are therefore not purged automatically by Litmus Production Record Database.
The picture below shows, how much flexibility Litmus Production Record Database can provide, when defining tags. The green boxes represent data recorded as process data, while the purple boxes represent meta data. And while all tags may share some common items, it is totally within the capabilities of the system to add tag specific items as well.
When defining tags inside Litmus Production Record Database, they should always have a direct relation to data which can be collected from a process or have a relation to signals used by a process. As otherwise, no value can be recorded.
The picture below shows, how taking for example the signals provided by a process as template and then translate them into tags recorded by Litmus Production Record Database.