Rodrigo Aldana López, professeur émérite à l'université de Saragosse dans le cadre du programme de bourses Beatriz Galindo
  • Search

Seminar by Rodrigo Aldana López

The Department of Automation is pleased to invite you to a seminar by Rodrigo Aldana López, currently professor emeritus at the University of Zaragoza, as part of the Beatriz Galindo Fellowship Program.

  • Le 18/06/2026

  • 15:00 - 16:00
  • Seminar
  • Mont Houy Campus
    Claudin Lejeune 2 building
    amphi E2

Abstract

This paper addresses distributed differentiation, in which agents receiving time-varying local signals must cooperatively estimate the network-wide mean and the derivative of that mean relying solely on communication between neighbors.

The main contribution lies in a Lyapunov framework that extracts the essential structural characteristics common to the classical super-torsion differentiator and integrates them into an abstract super-torsion model that accounts for the error dynamics within the network.

Using tools from convex analysis and homogeneity, we construct a Lyapunov function based on a strictly convex homogeneous potential and its convex conjugate, which leads to explicit gain conditions and global convergence in finite time to consensus starting from arbitrary initial conditions.

We then show that the distributed differentiator fits into this abstract model by selecting a potential constructed from the edge disagreements induced by the graph structure.

Extensions of this framework include “leader-follower” distributed differentiation and affine formation tracking based on the HOSM model.

Short Biography

Rodrigo Aldana López was born in Guadalajara, Mexico, and earned his Ph.D. from the University of Zaragoza in Spain. He received the EECI Award for the best Ph.D. thesis in systems and control in Europe for the year 2024.

He currently holds the position of professor emeritus at the University of Zaragoza as part of the Beatriz Galindo Fellowship Program.

His areas of interest include control theory, distributed optimization, robotics, machine learning, networked systems, and autonomous decision-making.

He has collaborated with Intel Labs in the fields of artificial intelligence, robotics, and large-scale computing systems, resulting in more than 20 issued patents.

He is associate editor of the Journal of the Franklin Institute and is the author of numerous publications in leading journals and conference proceedings in the fields of systems and control, robotics, and machine learning.

Title

Super-torsion on Networks: A Lyapunov Approach to Distributed Differentiation.

Contact

Michael Defoort