Bio-inspired autonomy in soft robots

Soft robotic actuation concepts meet and sometimes exceed their natural counterparts. In contrast, artificially recreating natural levels of autonomy is still an unmet challenge. Here, we come to this conclusion after defining a measure of energy- and control-autonomy and classifying a representative selection of soft robots. We argue that, in order to advance the field, we should focus our attention on interactions between soft robots and their environment, because in nature autonomy is also achieved in interdependence. If we better understand how interactions with an environment are leveraged in nature, this will enable us to design bio-inspired soft robots with much greater autonomy in the future.