Abstract

In recent years robotics has been increasingly focusing on the transition of robots from industrial cages to unstructured environments. This shift aims to achieve stable and safe execution of complex tasks in dynamic and unpredictable settings. Energy-based methods, such as passive controllers, have emerged as promising solutions, as the stability of closed-loop systems is, in principle, independent of external environmental interactions. At the same time, recent advancements in Artificial Intelligence (AI) have gained significant interest, positioning AI techniques as promising tools for designing controllers capable of handling complex tasks in unknown environments. However, these approaches struggle to guarantee stability and safety, critical in human-robot interaction, and they require large amounts of data, which is usually prohibitive for robotic applications. Energy is a fundamental concept across various engineering and scientific disciplines, and it has the potential to facilitate communication between AI experts and control theorists. It simplifies the analysis of complex systems by breaking them into subsystems, where their combined energies determine overall behavior. These principles not only enhance the understanding of control actions but also support stability analysis. This workshop seeks to explore the role that energy can play in the ongoing transformation of robotics. Moreover, it aims to investigate how energy principles can be harnessed to bridge the gap between AI and traditional robot modeling and control to overcome their respective limitations and unlock new, promising capabilities. 

Speakers

Session  1: Control

What insights can energy considerations provide for controlling robots in complex tasks?

Session 2: Learning

How can energy concepts be used to bring learning-based methods more effectively into robotics?

Organizers