Cosimo Della Santina is Assistant Professor at TU Delft, and Research Scientist at the German Aerospace Institute (DLR). He received his PhD in robotics (cum laude, 2019) from University of Pisa.
He was then a visiting PhD student and a postdoc (2017 to 2019) at the Computer Science and Artificial Intelligence Laboratory (CSAIL), Massachusetts Institute of Technology (MIT). He was also a post doc (2020) at the department of Mathematics and Informatics, Technical University of Munich (TUM). He is now a guest lecturer at the same university. Cosimo has been awarded with euRobotics Georges Giralt Ph.D. Award (2020), and the “Fabrizio Flacco” Young Author Award of the RAS Italian chapter (2019). He also has been a finalist of the European Embedded Control Institute PhD award (2020). His main research interests include (i) Modelling for Control and Model Based Control of Soft Robots, (ii) Combining Machine Learning and Model Based Strategies with application to soft robotics, (iii) Soft Robotic Hands/prostheses.
Model Based Learning and Control for Continuum Soft Robots
Taking inspiration from the animal body, researchers have started introducing compliant elements into robot mechanical structure, leading to the so-called Soft Robotic field. Leveraging the capabilities provided by their continuously deformable bodies, Continuum Soft Robots promise to have a disruptive impact on several fields of science, industry, and society. However, soft robotic systems need to robustly manage the intelligence embedded in their complex structure in order to generate reliable and repeatable behaviors. Substantial progress has been made in the development of soft robotic bodies, but developing control strategies suited for soft body control has remained challenging. In this talk, I will discuss the model based view to this grand challenge within the soft robotic field. I will show how simplified models can be combined with standard techniques in nonlinear control theory, leading to the execution of tasks that are precise and dynamic. I will also propose ways of integrating such views with data driven techniques, towards higher levels of control performance.
Talk: Model Based Learning and Control for Continuum Soft Robots