Robotics and Formal Methods

  • Posted on: 22 January 2021
  • By: smover
  • Updated on: 22 January 2021
Type recrutement: 
Sujet de thèse
Section: 
27 (Informatique)
71 (Sciences de l'information et de la communication)
Entité et lieu: 
Ecole Polytechnique, Palaiseau
Urgent ?: 
oui
Détails: 

We have a fully funded Ph.D. position to work at the intersection of Robotics and Formal Methods on the CoLeSlAw project.

Continuously Learning Complex Tasks via Symbolic Analysis (CoLeSlAw)
--------------------------------------------------------------------------------
Fully autonomous robots have the potential to impact real-life applications, like assisting elderly people. Autonomous robots must deal with uncertain and continuously changing environments, where it is not possible to program the robot tasks. Instead, the robot must continuously learn new tasks. The robot should further learn how to perform more complex tasks combining simpler ones (i.e., a task hierarchy). This problem is called lifelong learning of hierarchical tasks.

The existing learning algorithms for hierarchical tasks are limited in that: a) they require the robot to execute a large number of real actions to sample the continuous state space of observations, hence requiring a lot of time; b) they cannot deal with subspaces without continuous interpolation, as it is the case for a hierarchy of tasks.

The goal of the Ph.D. project is to explore the use of set-based and symbolic reasoning for the continuous space to tackle the above challenges (e.g., reducing the number of samples required to learn a hierarchy of tasks and allow for more effective planning of the robot tasks, further handling discontinuities in the task hierarchies).

The main outcome of the project will be an algorithmic framework to effectively explore task hierarchies and new reachability algorithms for data-oriented models, such as neural networks.

A detailed description of the project is available here

Candidate
--------------------------------------------------------------------------------
The ideal candidate will have a Master degree in Computer Science, Computer Engineering, or Robotics, with a strong background in at least one topic among learning algorithms, robotics, planning, and formal methods (e.g., abstract interpretation, model checking).

Deadlines
--------------------------------------------------------------------------------
The Ph.D. will start as soon as possible and not later than June 1st 2021. We encourage the candidates to contact us before February 28th 2021 to receive full consideration.

Work environment
--------------------------------------------------------------------------------
The Ph.D. will be carried out in the Laboratoire d’informatique de École Polytechnique (LIX), École Polytechnique, and in the Computer Science and Systems Engineering Laboratory (U2IS), ENSTA Paris, ENSTA Paris, under the supervision of Sergio Mover and Sylvie Putot from LIX, and Sao Mai Nguyen and Alexandre Chaputot from U2IS.

Doctoral School
--------------------------------------------------------------------------------
École Polytechnique and ENSTA Paris are part of the Institut Polytechnique de Paris (IPP) and the Ph.D. will be at in the IPP doctoral school.

Contacts and application
--------------------------------------------------------------------------------
For more information and how to apply contact Sergio Mover, Sao May Nguyen, Sylvie Putot, and Alexandre Chapoutot.

  • Sergio Mover, Cosynus Team, LIX and Ecole Polytechnique, sergio.mover at polytechnique.edu 
  • Sao Mai Nguyen, ASR Team, U2IS and ENSTA PariTech, sao-mai.nguyen@ensta-paris.fr 
  • Sylvie Putot, Cosynus Team, LIX and Ecole Polytechnique, sylvie.putot polytechnique.edu 
  • Alexandre Chapoutot, SSH Team, U2IS and ENSTA PariTech, alexandre.chapoutot ensta-paris.fr