I have used the daewr and rsm packages in the past to create very simple designs. However they seem to have a lot of limitations (ex. maximum number of 7 factors for Box Behnken designs in rsm, number of factors must be between 4 and 7 for Definitive Screening Designs in daewr) that I don't experience in JMP.
Are there other packages I should consider? Is this an area where commercial software is more advanced than R?