Process Capability (CPK)
Use Tulip to create control charts and measure CPK
Many operations measure their components against a specification. This is commonly found in machining centers and inspection areas. Ideally, you'd like your process to be in control - with random variation around a mean. This "normal" bell curve is representative of a stable process - however, we'd also want the control limits (typically 3 standard deviations from average) to be within the specification limits. Charting these measurements continuously is what is referred to as control charting. This application provides a framework to generate these charts from a Tulip terminal as well as calculating the Cp (process potential) and Cpk (process capability). It also showcases techniques to set alerts for measurements falling on one side of nominal or trending towards a limit.
General Process Overview
This application is a framework and can be used however you'd like. As such, it doesn't have any other connections except the charting and alerting functions. Enter a part number on the landing page and select whether you'd like to view the charting techniques that use Completions or Tables. The benefit of using tables vs. completions is that you can also pull in the data from the Table API and generate alerts.
Once in the charting step, simply enter in your spec limits, control limits, and your measurement. If you are using the tables technique then you will be be able to add an "alert threshold" which will look at the last [x] records to determine your alert trends. For example, if you enter "3" it will alert you if 3 consecutive entries are trending towards a limit. Press "ADD" to begin charting.
Getting Started - Application Overview
Below is a video that walks through the application in more detail:
This application makes use of Tulip Table APIs, which will need to be set up before the app will function. Once those are set up you will be able to begin using the app and generating charts (necessary for the table-based charts only).
You will also need to create your table analytics, since those do not transfer over when importing apps.
Setting up the Tulip Table API
First, create a bot using the Tulip Table API by clicking on settings in the top right
Then Create a Bot
Grant the Bot all Scopes
This will generate credentials for you to set up your API connector. It is a good idea to save these somewhere. Leave this window open while you open up a new window.
In your other window, navigate to your connectors and select the new HTTP connector called "Table API" and edit the connection settings
Choose "Cloud Connector Host" and enter in your tulip instance as the host. Then click edit headers.
Choose Basic Auth and enter in your API Key (from the previous window) as the Username and your Secret as your Password.
Your table API is now set up!
Setting up the Analytics
Click on the "select from existing" button on the empty analytic within the "Process Capability (tables)" step. When you add your analysis, you'll need to select that it will be created from tables.
Choose the "*SPC" table and click "create new analysis". The table analytic is fairly simple to create, below is an image of how it is configured:
How the alerts work
In order to create the alerts, we are actually pulling from the Tulip table API and sending that information into the Connectors demo to let SQL evaluate the data and return the alerts. This part isn't very "no code" but it is a useful technique that we wanted to share.
You won't need to make any changes to this but you are free to make adjustments if you'd like.