The picture shows the hierarchical structure of an artificial system, here in the form of a hand. The schematic structure of the system (S) in the form of a pyramid names the active nanomaterials (NM) at the base, which form the materials (M) at the next level and from there are assembled into components (C), which then form subsystems (SB) from which the system (S) is constructed. Each level is represented by different coloured circles with the respective letter abbreviation and a selection of elements of one level is connected by thin lines to some elements of the next level.

Below this network-like representation is a schematic representation of a Convolutional Neural Network model, which is intended to illustrate the application of machine learning methods in this context. It consists of input parameters represented by arrows that are connected into the next layer inside. A convolutional layer is highlighted. On the output side, connections into the interior and output parameters are shown. Above this depicted mesh is a sphere with finite element mesh shown, showing an active hydrogel in its initial state (swollen, shown in grey) and in its final, unswollen state (shown in blue). This is an active example material for which the hierarchical approach shown in the upper part of the image can be applied.