[Opt-net] PHD Position for research on Cooperative and Robust Scheduling Methods

[Cyril Briand]

PHD POSITION for research on Cooperative Robust Scheduling Methods

The MOGISA research group of the LAAS-CNRS of Toulouse (France) has an 
open 3-year position for a PhD student (full-time), starting from 
September 2009. The PhD student position is proposed in the context of 
the ROBOCOOP project (Robustness and Cooperation in Scheduling) funded 
by the French ANR research agency. Four partners are involved in the 
ROBOCOOP project: the MOGISA group of the LAAS-CNRS of TOULOUSE 
(http://www.laas.fr/MOGISA/MOGISA.html), the LITIS of LE HAVRE 
(http://www.litislab.eu/), the LI of TOURS (http://www.li.univ-tours.fr) 
and collaborators from the ILOG-IBM software company located in Paris.

In most existing approaches, scheduling is often seen as a global 
decision problem where decisions concern the organization of the entire 
resource set, assuming a global knowledge of the system parameters. 
However, in a lot of application fields (supply chain management, 
project management, timetabling, grid computing), resources are often 
distributed among a set of actors which have their own decisional 
autonomy, their own decision objectives and a restricted knowledge of 
their environment. Moreover, this kind of distributed organization is 
precisely the context where robust decision methods can be profitable 
since each decision centre has to face, not only the unexpected events 
coming from its own organization, but also and particularly those 
arising from its environment. Therefore, it is interesting to revisit 
robust scheduling under the assumption of a cooperative approach, where 
scheduling decisions are distributed over a set of actors and have to be 
negotiated intending to converge towards a compromise which satisfies 
local actor performance objectives and global performance objective.

The research will seek to design, to evaluate and to analyse cooperative 
robust scheduling methods used in Shop and Project Scheduling problems 
under uncertainty. Successful applicant will design, develop and 
implement algorithms for solving, in a cooperative way, dynamic 
scheduling problems, implementing original cooperation mechanisms. The 
research may also involve some analysis of data, the design of numerical 
experiments, the numerical evaluation of new algorithms, and the 
analysis of algorithm performances.

Applicants must demonstrate their ability to carry out high-quality 
research in scheduling, a high level of proficiency in computer 
programming, their ability to work as part of a research team, and to 
interact with industry research partners. Good spoken and written 
English communication skills are required. Familiarity with Scheduling, 
Optimisation or Constraint Programming is desirable.

Interested individuals should send an email with their (detailed) CV 
before June 12th 2009 to the following contact:
Cyril Briand
LAAS-CNRS
7, av. du Colonel Roche
31077 Toulouse Cedex 4
FRANCE
Email: briand(at)laas(dot)fr
Tel: (33)  561 337 818
-------------- next part --------------
A non-text attachment was scrubbed...
Name: briand.vcf
Type: text/x-vcard
Size: 254 bytes
Desc: not available
Url : http://listserv.zib.de/pipermail/opt-net/attachments/20090402/cbb057f0/briand-0001.vcf