From Elephant Biology to Soft Robotics: A New Bio-Inspired Robotic Arm

A new study published in Advanced Intelligent Systems presents ELEANOR, a large-scale soft robotic arm inspired by the extraordinary versatility of the African elephant trunk. The work illustrates how knowledge gained from biological systems can help engineers develop more adaptable and efficient robots.

Unlike conventional robotic arms built from rigid segments and articulated joints, ELEANOR features a continuous and highly compliant structure. The 85-cm-long arm was produced using large-scale 3D printing and actuated by a tendon system inspired by the organization of muscles in the elephant trunk.

The project builds on previous research investigating how elephants achieve remarkable dexterity despite the absence of bones and joints in their trunks. These studies showed that trunk movements arise from combinations of a limited number of biomechanical strategies, including bending, elongation, and the temporary formation of pseudo-joints.

By incorporating these principles into its design, ELEANOR can perform large, smooth deformations, wrap around objects, and adapt its shape to interact with items of different sizes and mechanical properties. The results support a central idea in bio-inspired robotics: complex behaviour can emerge from the physical structure of a system itself, reducing the need for sophisticated control algorithms.

The research was led by Lucia Beccai's Soft BioRobotics Perception Laboratory at the Istituto Italiano di Tecnologia (IIT) and involved an international team of collaborators from Korea Aerospace University, Photocentric, Scuola Superiore Sant'Anna, and Professor Michel Milinkovitch of the Department of Genetics and Evolution, University of Geneva, within the framework of the EU Horizon 2020 PROBOSCIS project.

Publication

Continuous and Large-Scale: ELEANOR, the Soft Architected Arm Inspired by the Elephant Trunk

Naselli, Nardin, Joe, Drinkwater, Donato, Bianchi, Falotico, Milinkovitch & Beccai
Advanced Intelligent Systems (2026), e70532
 DOI: 10.1002/aisy.70532