Distinguished Lecture: Dr. Anton Shiriaev – NTNU Norway
Date and time: 11 April 2024, 11:00-12:00 CEST
Speaker: Dr. Anton Shiriaev, NTNU, Trondheim, Norway
Title: On developing new abilities for robotic systems
Where: Digital Futures hub, Osquars Backe 5, floor 2 at KTH main campus OR Zoom
Directions: https://www.digitalfutures.kth.se/contact/how-to-get-here/
OR
Zoom: https://kth-se.zoom.us/j/69560887455
Meeting ID: 695 6088 7455
Moderator: Xiaoming Hu, hu@math.kth.se
Administrator: Emil Björnson, emilbjo@kth.se
Abstract: One of the great advantages of model-based approaches in robotics is an opportunity to separate the task of motion/trajectory planning for a robotic system from the task of a synthesis of feedback controller for stabilizing its pre-planned behaviour. This is quite different from the way humans learn motions. Indeed, human trials in learning new behaviours are embodied and, therefore, should be accompanied by feedback actions to make them insensitive to various perturbations and uncertainty. This clearly limits a variety of human behaviours and requires time and effort to perform most of the common motions for each of us.
Meanwhile, already-learned behaviours become useful for searching and training new abilities. What can be a representation for organizing such a learning process if the new ability is demanding and should comprehend the physics of the environment and a human as a mechanical system put to operate in it? What do humans train in performing trials for such tasks? How can we incorporate such insights into searching for new robotic systems abilities and consider it a math problem?
The talk will discuss assignments on developing/learning agile movements of human-like robotic systems for anthropomorphic machines or for developing/learning robot-like behaviours that can be difficult or impossible for humans to perform. The illustrative examples can be useful for vision in creating novel robotic systems with advanced capabilities.
Biography: Anton S. Shiriaev, Professor at the Norwegian University of Science and Technology (NTNU), Trondheim, Norway, is an internationally recognized researcher in applied math, control systems, mechanical engineering and robotics subjects. He received MSc and PhD degrees in Applied Math from St. Petersburg State University, Russia, in 1993 and 1997.
Throughout his academic career, Dr Shiriaev held (permanent or long-time visiting) positions in various universities and academic institutions worldwide, including universities at Lund, Umea, Aalborg and Odense in Scandinavia, at the CNRS and universities at Montpellier and Nantes in France, at the Tokyo Institute of Technology, at the Russian Academy of Sciences, at the University of Texas at Dallas and others. He has made several scientific contributions and tangible observations in stability analysis of non-linear dynamical systems, in modelling and calibrating robotic systems, in developing algorithms for planning, controlling and analyzing agile behaviours of mechanical systems, in analyzing and reconstructing human motions, in planning and performing dynamic manipulations by robotic systems.
Dr. Shiriaev has actively tutored and supervised students and young researchers, bringing them to cutting-edge scientific and technological problems through lectures and original PhD courses.
Anton S. Shiriaev is a Digital Futures Scholar in Residence from February to April 2024.