Postdoc position LAAS-CNRS / CNES

We are seeking applicants for a one year postdoc position at LAAS in Toulouse on a collaborative project with the Centre National d’Etudes Spatiales (CNES) on scheduling maneuvers and scientific experiments of the Rosetta/Philae spacecraft. Rosetta is a mission of the European Space Agency (ESA) and will involve the first ever landing on a Comet (67P / Churyumov - Gerasimenko) http://sci.esa.int/science-e/www/area/index.cfm?fareaid=13. The postdoc will start at the earliest in September 2011, and at the latest in January 2012.

=========== PROJECT ============

The project is concerned with developing algorithms for scheduling the various activities of the lander (Philae) in three phases. First, in the Separation, Descent and Landing (SDL) phase, the lander will be separated from the orbiter module and will land on the surface of the comet. Second, during the First Science Sequence (FSS), the main scientific experiments will be run using the onboard batteries as energy source. This phase should last three to four days, depending on the load of the batteries and on the quality of the schedule. Finally, during the Long Term Science phase, scientific activities will be resumed at a much slower pace, using the lander’s own solar panels to partially reload the batteries.

The experiments of the FSS and LTS – and to some extent the maneuvers of the SDL – need to be scheduled to satisfy a number of constraints involving the concurrent use of energy (batteries), and of the main CPUs as well as each instrument’s memory. Another important constraint concerns the transfer of data and the notion of “visibility”, that is, the availability of the orbiter to act as relay to transmit the data back to earth, which depends on its orbit and on the length of a cometal day. A tool developed using Ilog Scheduler and Solver (MOST) is currently used to model this problem and to generate feasible plans. Every instrument, subsystem and experiment has been modeled with this library, and it is therefore possible to verify solutions with a great degree of confidence of its feasibility. However, the complexity of the problem, in particular during the FSS phase, makes it difficult to find solutions in a reasonable time.

The role of the recruited researcher will be to explore ways of obtaining better solutions quicker, and to develop prototypes to assess the gain. This might be done by improving the model, the propagation algorithms and the search strategies currently employed, or by developing radically new models whose solutions can then be fed back to MOST.

=========== WORKING ENVIRONMENT ============

The project is expected to be carried out in collaboration with Pierre Lopez, Christian Artigues and Emmanuel Hebrard (MOGISA group: http://www.laas.fr/MOGISA-EN/), and in constant contact with the SONC team (http://smsc.cnes.fr/ROSETTA/GP_contacts.htm) of the CNES.

=========== APPLICATION ============

We are seeking applicants with an outstanding cv and a thesis related to scheduling and/or combinatorial optimization. A good experience of IBM/Ilog products would be a great plus, however it is not necessary.

Please send the following documents to Emmanuel Hebrard (hebrard@laas.fr) straightaway:

- A detailed curriculum vitae;

- A motivation letter;

- Abstract of the PhD thesis;

- The three most relevant publications;

- (optional) Reference letters from at most three referees.