Sometimes you don’t know what kind of hidden information is within your process measures, and feeding your data to a control chart chosen from a library can hide as much as it reveals. In contrast, Control Chart Builder in JMP 10 lets you dive deeper into your process control data using the drag-and-drop flexibility of Graph Builder and the functionality of the Control Chart platform.
“Control Charts are the fourth most-used platform in JMP,” says Director of Statistical Research and Development Chris Gotwalt. “We wanted to do something new and exciting with the JMP 10 release that will revolutionize the way you interact with your process data.”
What Chris and the team of JMP developers came up with is next-generation for control charting tools. “The Control Chart Builder lets you see your data in the light of the questions or ideas you have rather than forcing you to follow a prescribed workflow,” Gotwalt explains. “And most importantly, Control Chart Builder allows you to be surprised and open up new avenues of discovery with a visual exploration paradigm that JMP is known for -- it’s really quite exciting!”
This sequence of images illustrates the use of JMP 10 and the Control Chart Builder to explore textile manufacturing defect data.
Data for textile quality from multiple suppliers, dates, lots and roll samples can be explored with new Control Chart Builder in JMP 10.
Automatically creating a Control Chart is easy: Just drag the “Points Per 100 Yards” (our defect measurement) and the chart is automatically made. It’s clear that there are certain samples that are out of control. Is it all suppliers or maybe just a certain supplier?
Dragging and dropping the Supplier column to the phase of the Control Chart allows us to visually flow to the conclusion that Supplier ‘A’ has rolls that are out of control, whereas supplier ‘B’ and ‘C’ seem to be OK.