A split plot experiment is usually a (fractional) factorial, often replicated, with a restriction on randomization. So, you would create your design in the appropriate DOE platform, and then in a data table, you could re-arrange the rows to restrict the randomization.
Using Custom Design, you can specify factors as "Easy", "Hard", or "Very Hard" to change. When a factor is specified as Hard, that tells JMP that a split plot design is being made. When the table is made, it creates a column for Whole Plot and restricts the randomization.
"Very Hard" is used with split-split-plot design. This is all detailed in the DOE Guide. Look up split plot design in the index.