By reducing high-dimensional biological movements into simplified linear relationships, we provide insights into natural muscle coordination
Using motion-capture technologies developed for the movie industry, we have previously demonstrated that the complex behaviours of the elephant trunk emerge from the combination of a finite set of basic movements such as the propagation of an inward curvature and the formation of pseudo-joints. In addition, we have shown that the elephant trunk velocity obeys a mathematical law observed in human hand drawing movements. These results were published in the journal Current Biology.
We are happy to announce our contribution to a new study in iScience —led by Camilla Agabiti and Egidio Falotico (Sant’Anna School of Advanced Studies, Pontedera, Italy)— where the dexterity of the elephant trunk is modelled for Soft Robotics applications.
Modelling continuous, skeleton-free biological structures is notoriously difficult. We address this by developing a 3D rod-based dynamic model to simulate planar reaching tasks. We introduce linear "stereotypical" laws that map desired trunk shapes —based on length and curvature—directly to the internal, muscle-analogue forces required to produce them. Using the data we collected on real elephants, we show that this model predicts biological trajectories with tip-position errors under 8% while accurately preserving the trunk's constant volume during motion.
By reducing high-dimensional biological movements into simplified linear relationships, this research provides deep insights into natural muscle coordination. Furthermore, it establishes a foundational framework for developing new, model-based control systems for advanced soft robotics.
Much additional information is available in the original article:
Stereotypical force patterns of the elephant trunk in planar reaching movements
Agabiti, Donato, Setti, Dagenais, Milinkovitch, Laschi, Sabatini, Mazzolai & Falotico
iScience 29, 115108 (2026)
https://doi.org/10.1016/j.isci.2026.115108