We are pleased to announce that our coordinator, Dr. Nicolas Mansard from LAAS-CNRS, will present a research seminar titled "Artificial movement: Should we learn or optimize?" on Monday, April 29th, at the Sala d'Actes Facultat de Matemàtiques i Estadística UPC in Barcelona.

The seminar will delve into the challenges of generating precise movements for advanced robots, whether for walking or manipulation tasks. With recent advancements in legged locomotion hinting at breakthroughs, two primary strategies will be explored: predictive control and reinforcement learning. Predictive control involves optimizing predictions of robot movements while in motion, while reinforcement learning focuses on offline optimization of control policies.

Moreover, the discussion will underscore the pivotal importance of progress in robot hardware and underscore the potential for artificial intelligence to enhance the optimization of robot design even further.

Find more information about the seminar here.