BEGIN:VCALENDAR
VERSION:2.0
PRODID:-//Memento EPFL//
BEGIN:VEVENT
SUMMARY:Learning and control for complex and multi-agent systems
DTSTART:20210217T090000
DTEND:20210217T100000
DTSTAMP:20260415T235848Z
UID:94d7cc77c84e2b7f8dc2bdc25de371d40714c02ed51d9d9b93f41efe
CATEGORIES:Conferences - Seminars
DESCRIPTION:Dr. Yvonne Stürz\, Model Predictive Control Laboratory\, UC B
 erkeley\nAbstract: The increased availability of inexpensive hardware\, se
 nsors and computational power enables a widespread use of systems in compl
 etely novel application areas and with ever-growing complexity. While auto
 mation and multi-agent systems have a significant potential to increase th
 e efficiency\, sustainability and safety in many industries (such as energ
 y\, traffic\, robotics\, construction)\, there are still major open challe
 nges in the control of complex and multi-agent systems. In particular\, th
 eir large scale and spatial distribution give rise to the need of efficien
 t and scalable local control algorithms.\n\nMoreover\, safety and performa
 nce have to be guaranteed despite uncertainties and unknown environments. 
 The talk will illustrate this need for novel advanced control methods with
  two examples:\n\nThe first part of the talk introduces a novel control ap
 plication in the area of digital fabrication. Unlike many other branches o
 f industry\, the construction sector has not followed the trend of automat
 ion\, mainly because it has to deal with unstructured environments and non
 -standardized complex tasks. However\, as one of the largest consumers of 
 resources worldwide\, the construction industry has a high potential for i
 mproved sustainability. A novel method to precisely control the form of a 
 tensioned cable-net based formwork on site is an example of how this poten
 tial can be achieved. The presented approach enables the construction of l
 ightweight building elements. The developed optimization-based control alg
 orithm tackles the complexity of the flexible cable-net system by exploiti
 ng the mathematical structure and convexity of the problem. Experimental v
 alidation on a cable net for a doubly-curved roof shell will be presented.
 \n\nThe second part of the talk focuses on data-driven control of autonomo
 us multi-agent systems. While the operation of single robots in structured
  environments\, such as autonomous driving on the highway\, is well studie
 d\, the coordination of complex tasks of fleets of multiple agents still p
 oses major challenges. In particular\, designing distributed controllers w
 hich guarantee safety and a high performance for multi-agent systems is di
 fficult. This is due to the inherent requirements for scalability and the 
 thereby introduced limitations in local sensing\, computational power and 
 communication. These informational constraints can lead to a decreased con
 trol performance. Therefore\, a novel distributed learning model predictiv
 e control scheme is presented\, which uses local data in order to learn lo
 cal control policies. This approach is data-efficient and guarantees safet
 y as well as\, under mild assumptions\, global optimality\, of the central
 ized closed-loop system.\n\nBio: Yvonne Stürz is a postdoctoral researche
 r at the Model Predictive Control Laboratory at UC Berkeley\, USA. Through
  the Marie-Curie fellowship that she was awarded\, she is also affiliated 
 with the Automatic Control Department at KTH\, Sweden. In 2019\, she obtai
 ned her PhD from the Automatic Control Laboratory at ETH Zurich\, Switzerl
 and. She holds a Bachelor's and Master's degree in mechanical engineering 
 from TU Munich\, and a Diploma degree in general engineering from Centrale
 Supelec Paris\, France. She was awarded scholarships from the Hans Werthen
  Foundation\, the German National Academic Foundation and the Ulderup Foun
 dation. She gained industrial experience through internships with Bosch\, 
 Porsche and Siemens in Germany and with Siemens SFAE in China. Her researc
 h interests include distributed optimal control and data-driven methods\, 
 with applications to large-scale heterogeneous interconnected systems\, mu
 lti-robot systems\, autonomous driving and digital fabrication.
LOCATION:https://epfl.zoom.us/j/81443631619?pwd=Ym5SWFZDa0M0N3hWcU1FRnNhUn
 BNUT09
STATUS:CONFIRMED
END:VEVENT
END:VCALENDAR