• Bio-inspired modeling
• Continuum curve
• curvature
• Elephant trunk
• Kinematic analysis
• torsion

The introduction of robotic manipulators in the industry has changed the horizon of robotics. Traditional, or hard, robotic manipulators are characterized by a rigid structure consisting of links and a finite number of joints. They are widely used in industry and although they have been proven extremely effective for many tasks, they are not without limitations, mainly due to their lack of manoeuvrability and number of degrees of freedom.
To try to overcome these limitations, robot designers have started to create structures with more and more degrees of freedom, moving from traditional robotics, to hyper-redundant robots, to soft robots or continuous robots. This type of robot is characterized by a high, ideally infinite, number of degrees of freedom which gives it high manoeuvrability, dexterity and movement capabilities. These characteristics make soft robots particularly suitable for use in unstructured environments and in contact with humans. For these and other reasons, research in soft robotics is growing rapidly.
One of the main trends assisting the development of soft robotics is bioinspiration. The term bioinspiration refers to the process of abstracting principles from nature that can be exploited in other areas. This mechanism also caught on in robotics through the study of the movement capabilities of certain animals, such as earthworms, serpents, octopuses, and elephants. The movement of these animals could be exploited for both innovative methods of locomotion and gripping.
Within this context, the PROBOSCIS project is developed with the objective of studying the dexterity and agility of the elephant trunk for the construction of a robotic manipulator that emulates it. The topics covered in this thesis represent a portion of this study. The primary goal of this thesis is to create a kinematic model that is capable of reproducing the trunk's reaching strategies. To do this, first a detailed analysis of kinematic data provided by the University of Geneva was performed, focusing on the movement of the proboscis during the execution of different tasks. With these analyses, we were able to reconstruct and characterize the curve representing the proboscis during its motion. Once the curve had been characterized, we moved on to the kinematic analysis of the movement of the proboscis’ tip. These two steps served for the extraction of the fundamental parameters for the construction and validation of the model. This model was obtained from the study of those present in the literature, which were compared and expanded to meet the needs of this project.