I need a quick tip regarding best practice. I have too many records for a product that need to be tracked. So which one is more suitable for best practice?
1- Keep all record in the same table with lots of column. (One table to maintain with huge data)
2- Keeping them in a couple of tables with less amount of columns. (More than one table to maintain with lesser amount of data)
Hey @OzanB, this is a fantastic question. The answer to this question is going to depend heavily on the format you want the data to take at the end, and any analytics or processing you need to do after it’s collected.
It’s important to note that analytics in Tulip can’t have multiple tables as sources. Often, this will lead folks to format their data in a single table with many rows.
@OzanB, I want to flag that this is also something that my team (The Education Team) is currently working on, and that we’ll have content to share around more complex data problems in Tulip in the near future.
Rather than think of a table as just data organization we prefer to define a table by what it represents and how it should be used. Tulip table data is most effective when it represent a real world physical or operational artifacts.
Physical artifacts are things that you can feel or touch in the industrial operation. Eg. order, equipment, material, etc.
Operational artifacts are tangible things that are central to industrial operations. Eg. order, job, action, etc.
Each row in the a table is therefore the digital twin representation of something in your productions operations. For example in a Table of materials each unique material is a row and the id is the actual id you will find the material labeled with. If you take this perspective the answer to your question is much clear. Also the data in Tulip becomes much more transparent and interpretable, which means can be widely used. Collaboration on apps is more seamless and people building apps can easily extend tables following some simple rules.
I just need to create a product report that will consist of different steps of process parameters. So, In this case the it would be create a table for each different process step right? So, My each table will indicate one op. artifact and will not mess up later on?