A postdoctoral research grant is available at the Institute for
Systems and Robotics (ISR) of the Instituto Superior Técnico, Lisbon.
The topic of the grant is

Planning under Uncertainty for Multiagent/Multi-robot Systems.

Due to restrictions in communication, centralized planning solutions
are not always viable.  Decentralized partially observable Markov
decision processes (Dec-POMDPs) form a formal approach for multiagent
planning.  They handle uncertainty in acting and sensing in a
principled way, and recent years have seen great progress in planning
algorithms for Dec-POMDPs.  In this grant, we intend to advance the
state of the art in planning algorithms and their scalability.
Furthermore, we will consider a case study concerning the
decentralized control of irrigation channels, in which we will apply
planning under uncertainty techniques to design an autonomous
controller for each gate in a channel.  More details can be found at
http://decpucs.isr.ist.utl.pt/ .

ISR-Lisbon is a research institute of the Instituto Superior Técnico,
the oldest and largest engineering school in Portugal.  ISR-Lisbon has
a long standing tradition of research and development and offers a
modern and enthusiastic research environment with strong
interdisciplinary and international links.

The grant has duration of (a maximum of) two years, and candidates
should have obtained a relevant PhD degree (e.g., Computer Science,
Artificial Intelligence, Electrical Engineering).  The ideal candidate
has expertise in Markov decision processes and decentralized control.
Applications should include CV, a statement of research interests and
two recommendation letters, and be sent to Dr. Matthijs Spaan at
[EMAIL PROTECTED] or fax: +351-21-8418291, who can also be
contacted for more information.  Application deadline is December 23,
2008.

--
[ Matthijs Spaan ]               [ http://www.isr.ist.utl.pt/~mtjspaan ]
_______________________________________________
uai mailing list
uai@ENGR.ORST.EDU
https://secure.engr.oregonstate.edu/mailman/listinfo/uai

Reply via email to