BEGIN:VCALENDAR
VERSION:2.0
PRODID:-//Talks.cam//talks.cam.ac.uk//
X-WR-CALNAME:Talks.cam
BEGIN:VEVENT
SUMMARY:Distributed Learning for Scalable Collaboration in Robotic Multi-A
 gent Systems - Guillaume Sartoretti\, National University of Singapore (NU
 S)
DTSTART:20230608T150000Z
DTEND:20230608T160000Z
UID:TALK201775@talks.cam.ac.uk
CONTACT:Amanda Prorok
DESCRIPTION:My research group's work aims to mitigate the curse of dimensi
 onality in high degree-of-freedom (DOF) multi-agent robotic systems: as th
 e number of agents (robots or DOFs) in the system grows\, so does the comb
 inatorial complexity of coordinating them. There are many solutions to man
 aging this complexity growth\, and our work favors decentralized approache
 s\, whether it be for a team of mobile robots or a single articulated robo
 t. Specifically\, we have embraced advances in distributed reinforcement l
 earning (dRL) to let multiple agents learn a common decentralized policy i
 n a time-efficient manner\, with or without explicit communications among 
 agents. This has produced collaborative policies that naturally scale to l
 arge teams of agents while remaining near-optimal. In this talk\, I will p
 resent dRL based approaches to 1) one-shot and lifelong multi-agent path f
 inding (e.g.\, for warehouse automation)\, 2) the multi-agent traveling sa
 lesman problem (mTSP)\, and 3) communication learning for team-level coope
 ration in various tasks. I will present experiments on autonomous ground v
 ehicles as well as in simulated environments that help validate our learne
 d policies\, and finally briefly go over some of my lab's ongoing projects
 .\n\n \n\nSpeaker bio:\nGuillaume Sartoretti joined the Mechanical Enginee
 ring department at the National University of Singapore (NUS) as an Assist
 ant Professor in 2019\, where he founded and is directing the Multi-Agent 
 Robotic Motion (MARMot) lab. Before that\, he was a Postdoctoral Fellow in
  the Robotics Institute at Carnegie Mellon University (USA)\, where he wor
 ked with Prof. Howie Choset. He received his Ph.D. in robotics from EPFL (
 Switzerland) in 2016 for his dissertation on "Control of Agent Swarms in R
 andom Environments\," under the supervision of Prof. Max-Olivier Hongler. 
 He also holds a B.S. and an M.S. degree in Mathematics and Computer Scienc
 e from the University of Geneva (Switzerland). His research focuses on the
  distributed/decentralized coordination of numerous agents\, at the interf
 ace between stochastic modelling\, conventional control\, and artificial i
 ntelligence. Applications range from multi-robot systems\, where independe
 nt robots and systems need to coordinate their actions to achieve a common
  goal\, to high-DoF articulated robots\, where joints need to be carefully
  coupled during locomotion in rough terrain. Guillaume was a Manufacturing
  Futures Initiative (MFI) postdoctoral fellow at CMU in 2018-2019\, and wa
 s awarded an Outstanding Early Career Award from NUS' College of Design an
 d Engineering in 2023.
LOCATION:Department of Computer Science and technology\, FW26
END:VEVENT
END:VCALENDAR