There are already approaches in this direction e.g. such as providing a standardized experimental hardware-software interface with Antha [1]. The complexity of the problems in questions (biological domain, biophysical, biochemical) is daunting - we do not understand many things, "there is plenty of room at the bottom".