BEGIN:VCALENDAR
VERSION:2.0
PRODID:-//Talks.cam//talks.cam.ac.uk//
X-WR-CALNAME:Talks.cam
BEGIN:VEVENT
SUMMARY:Distributed Learning and Control for Networked Autonomy - Konstant
 inos Gatsis\, University of Oxford
DTSTART:20230223T140000Z
DTEND:20230223T150000Z
UID:TALK197671@talks.cam.ac.uk
CONTACT:Fulvio Forni
DESCRIPTION:Systems that operate autonomously\, with limited human involve
 ment\, are becoming ubiquitous and they facilitate applications such as Sm
 art Cities\, Connected Mobility\, Resilient Energy Systems\, and the Indus
 trial Internet-of-Things. Modern autonomous systems exploit the accelerati
 ng convergence of the physical world with the digital world due to a) sens
 ing available at scale\, b) novel networking and computing capabilities\, 
 c) powerful algorithms for processing sensing data\, and d) the ability to
  actuate on the physical world\, e.g.\, by technologies such as autonomous
  robots or autonomous driving vehicles. In the main part of this seminar\,
  I will study how these advances enable autonomous systems to learn and ad
 apt to data collected in a distributed fashion. I will discuss approaches 
 to address the fundamental challenge of communication efficiency in this d
 istributed learning setup\, combining the problem with tools from networke
 d control systems. If time permits\, I will also discuss recent work on ho
 w control systems can incorporate learning architectures\, such as Neural 
 Networks.\n\nThe seminar will be held in the JDB Seminar Room\, Department
  of Engineering\, and online (zoom): https://us06web.zoom.us/j/87986687566
 ?pwd=MGJScmMwd2lwT0tVMHNmWmxSa05XZz09\n\n
LOCATION:Department of Engineering / Online (Zoom)
END:VEVENT
END:VCALENDAR