BEGIN:VCALENDAR
VERSION:2.0
PRODID:-//Talks.cam//talks.cam.ac.uk//
X-WR-CALNAME:Talks.cam
BEGIN:VEVENT
SUMMARY:Real time control with lots of humans in the loop - University of 
 Southampton
DTSTART:20131216T100000Z
DTEND:20131216T110000Z
UID:TALK48696@talks.cam.ac.uk
CONTACT:Microsoft Research Cambridge Talks Admins
DESCRIPTION:Did you know that humans can control traffic light junctions b
 etter than the leading automatic adaptive control systems? The optimal swi
 tching policy for traffic lights on a network of junctions is computationa
 lly intractable (EXP-time complete) and current systems use approximate op
 timisations to determine the switching policy. Embodied simulation experim
 ents and also computer based simulation experiments\, which employ a "traf
 fic control computer game" have shown that human controllers can significa
 ntly outperform these automated controllers on many metrics. Furthermore b
 y applying supervised learning to the traffic control computer game it is 
 possible to capture player's strategies and thus new trained automated con
 trollers with high performance have been developed.\n\nA key difficulty in
  traffic control is that any signal control system is not the only control
  system 'in the loop'. Each vehicle is controlled by an independent human 
 who is not aware of (or even motivated towards) the actions they could tak
 e for optimal system performance. Furthermore\, assuming that a benevolent
  controller had this information\, their available control actions are onl
 y weakly acting (changing the traffic lights at junctions). In fact there 
 are several other control problems that share these properties: Crowd cont
 rol\, queueing problems\, building evacuation\, battlefield operations. In
  these "lots-of-humans-in-the-loop" systems real-time control actions are 
 limited or absent and optimal control strategies are computationally intra
 ctable.\n\nIn this talk I will discuss the potential for building on the t
 raffic control computer game to develop large multi-player games for simul
 ating lots-of-humans-in-the-loop systems and 'gaming' scenarios where real
  time control is applied. As in the traffic control computer game the deve
 loped control strategies may be in part learned from human players thus ex
 ploiting the powerful intuition and empathy of humans in solving the appar
 ently intractable problem of how to control them.\n\n
LOCATION:Auditorium\, Microsoft Research Ltd\, 21 Station Road\, Cambridge
 \, CB1 2FB
END:VEVENT
END:VCALENDAR