Use your historical experiment data as a starting point for producing new useful configurations and recipes. Usually smart_DoE starts from scratch and tries to produce useful results in as few steps as possible. But priming new experiments with existing data is also possible.
smart_DoE A.I. learning algorithms always try to improve on what is currently known about the problem.
For priming and data synthesis to work, your data has to be transformed into smart_DoE readable format according to specified experiment settings.