Two different Spec limits for Process Capability based on the value in one column.

[#1]

I use one measurement process to assess the capability of two different products on our production line. For various reasons, the spec limits for the two products are different but the measurement data is combined with one field being used to indicate which product is being processed.

I am trying to generate a simple, tabular process capability report that splits the data into two separate reports based on the product type in such a way that I can apply different specs to the two groups. I have called them Product A and Product B in this example. My thought was to change the script to look like the example below but the resulting Capability report uses the second set of Spec Limits for both product groups. The data come from the same source and I need to track them together so I don't really want to double-up the process with filters on the data source. Any Ideas please?

Process Capability(
	SendToByGroup( Bygroup Default ),
	Process Variables( :D1953_NpMin, :D1953_NpMax, :D1953_NpParallelism, :D1953_Act1LhsActualX, :D1953_Act1LhsActualY, :D1953_Act1Par ),
	SendToByGroup(
		{:prd_ERPItemNo == "PRODUCT A"},
		Spec Limits(
			:D1953_NpMin( LSL( 19.875 ), Target( 20 ), USL( 20.125 ) ), :D1953_NpMax( LSL( 19.875 ), Target( 20 ), USL( 20.125 ) ),
			:D1953_NpParallelism( LSL( 0 ), Target( 0.09 ), USL( 0.16 ) ), :D1953_Act1LhsActualX( LSL( 29.743 ), Target( 29.763 ), USL( 29.783 ) ),
			:D1953_Act1LhsActualY( LSL( -13.53 ), Target( -13.49 ), USL( -13.45 ) ), :D1953_Act1Par( LSL( -0.016 ), Target( -.006 ), USL( 0.004 ) )
		),
		{:prd_ERPItemNo == "PRODUCT B"},
		Spec Limits(
			:D1953_NpMin( LSL( 19.875 ), Target( 20 ), USL( 20.125 ) ), :D1953_NpMax( LSL( 19.875 ), Target( 20 ), USL( 20.125 ) ),
			:D1953_NpParallelism( LSL( 0 ), Target( 0.09 ), USL( 0.16 ) ), :D1953_Act1LhsActualX( LSL( 29.743 ), Target( 29.763 ), USL( 29.783 ) ),
			:D1953_Act1LhsActualY( LSL( -13.53 ), Target( -13.49 ), USL( -13.45 ) ), :D1953_Act1Par( LSL( -0.008 ), Target( .002 ), USL( 0.012 ) )
		)
	),
	Moving Range Method( Average of Moving Ranges ),
	Capability Box Plots( 0 ),
	Overall Sigma Summary Report( 1 ),
	Goal Plot( 0 ),
	Capability Index Plot( 0 ),
	Process Performance Plot( 0 ),
	By( :prd_ERPItemNo )
);

[#2]

Hi @RichardSlack ,

Can you provide the data table to work with? Or find a similar example in the sample index?

Thanks,
Ben

[#3]

JMP Data Table attached

[#4]

Spec Limits are a Column Property, so if the data for the two products are in the same Column, they will inherit the same Spec Limits. Maybe there is a way around this using some clever scripting...but I would recommend simply splitting the column:

• Tables -> Split.
• Split By: your Product Type column (Product A or Product B).
• Split Column: your process data columns.
• "Keep All" other columns as needed

Would this work for you? 

Just as a side note: If you are also using Control Charts, you could use a Short Run Control Chart to visualize processes that are running on the same line but with different Control Limits. (Keeping in mind Control Limits are very different from Spec Limits) 

[#5]

I am able to do it with Distributions by performing a Process Capability with different spec limits for each grouped parameter and then saving the script to repeat the process later. Unfortunately this doesn't give me such a nice output to then transfer to an Excel report, where I need the numbers in tabular format. The Process Capability tool gives me the tabular output I need but seems to not allow separate spec limits between the two groups of data.

[#6 · Accepted solution]

Hi @RichardSlack , 

There doesn't look to be an easy way to script/command to bygroup - but if you use a separate spec limit table it will work, see here.

Thanks,

Ben

[#7]

It will be a bit long-winded to complete the spec table but it definitely seems to be the solution I was looking for. Many thanks!

[#8]

Hi @RichardSlack 

I wanted to let you know that if you are using JMP 19, the Process Capability platform now has the ability to do grouped process capability. 

For example, the data below contains some cheese manufacturing data.  It has 3 process variables: pH, Salt Concentration, and Moisture Content.  It is organized by the grouping column, Cheese Type.  There are 3 cheese types:  Brie, Cheddar, and Swiss.

The specification limits for the cheese manufacturing data are in the Cheese Manufacturing Limits data table below.  Notice that there are 9 sets of spec limits.  One set for each of the 3 process variables for each of the 3 cheese types.

To analyze the capability of this grouped cheese manufacturing data, you can launch the Process Capability platform by adding the 3 process variables columns to the Y,Process role and Cheese Type column to the Grouping role.  See the image of the launch dialog below.

Once you click OK, you will get a Spec Limits dialog to allow you to load your grouped spec limits data table.

Once you select the spec limits table and click OK, you will get a grouped Process Capability analysis that includes an Overall Sigma Capability Summary Report like the following:

I hope this helps.

Laura

P.S.  Also... In JMP 19 the Manage Limits utility also works with grouped data that is organized with a grouping column.