Very limited information given.
But, SPC is not used to identify outliers because it use Process Data to derive the control limits.
How about using "Analyze > Screening > Explore Outliers>" tool in JMP to identify the outliers?
For univariate data, you can use Quantile Range or Robust Fit Outliers to detect outliers
But if you want to specify the "sigma = 3", maybe you want to use Robust Fit Outliers using robust method.