Functional morphology of cuttlefish suction cups

A daunting challenge in robotics is gripping complex objects in uncontrolled environments, such as harvesting crops in agriculture. These objects and environments consist of soft, wet tissues and rough surfaces, whose physical properties differ from materials traditionally encountered in engineering. However, through millions of years of evolution and adaptation, biological systems have developed versatile grippers that interface with such complex materials and surfaces. The aim of the proposed study is to uncover the physical working principles underlying biological gripping, with a particular focus on the suction cups of cuttlefish.

Cuttlefish suction cups are unique in their specialised designs, with extreme inter- and intra-species variations in morphology and organisation. Unlike other species bearing suction cups, individual cuttlefish may exhibit up to three different suction surfaces, thought to support various functions like prey capture and manipulation and mating. Moreover, cuttlefish suction cups are supported by a slender stalk, which is hypothesised to enable the passive generation of suction. Specialisation and passive functioning hint at embodied intelligence, an attribute attractive for bioinspired grippers and soft robotics.

This study aims to address the following: what are the roles of muscles in cuttlefish suction cup attachment and detachment and to what extent can reversible grasping be achieved passively without intrinsic muscle activity? A multi-disciplinary approach is proposed, combining in vivo animal observations, with ex vivo morphological and mechanical characterisation, and experiments and simulations using biomimetic models. In vivo observations using high speed videography will identify and quantify stereotypic behaviours involved in gripping and suction cup attachment and detachment. Ex vivo characterisation of the suction cups, including their geometrical and material properties, will reveal their biomechanics. In silico studies with biomimetic models, both physical and digital, will enable exploration of conditions outside those found in nature and offer tractable models to determine the underlying physics.