Kaleojob

Experimentation with the AIPlan4EU Unified Planning Framework and various planning use cases

Capijobnew

Capijobnew

Domaine: Toutes nos offres

Région: Aude, Gard, Hérault, Lozére, Pyrénées-Orientales

Contrat: NC

Expérience: NC

Niveau d'étude: NC

Salaire: NC

Permis demandé: Permis NC

Niveau de qualification: NC

Description: Description: Experimentation with the AIPlan4EU Unified Planning Framework and various planning use casesLAAS CNRS, ToulouseTeam: RIS AIPlan4EU is a 3 years long European H2020 project whose goal is to develop and deploy a Unified Planning Framework (UPF) to host various planning technologies as to solve a large number of planning problems encountered in real life. After a first year of development, a number of planners have already been integrated in the UPF. A number of use cases in different use-cases from industrial partners have been specified and need tobe modeled. Use-cases include but are not limitied to problems related to warehouse logistics, underwater operations for autonous submarines and planning for space operations (more details the project website).The internship will take place in the context of this AIPlan4EU project for which the candidate is expected to specify in the UPF API various use cases, which will be identified with the advisors, and to experiment with theavailable planners as to solve the problem or identify the limitations of the tested planners.The planning approaches covered may include : ? temporal planning and scheduling, ? connection with acting/execution, ? connection with motion planning, ? hierarchical representations, ? robotic applicationsThe candidate will have to learn how to use the UPF, model a planning problem, specify how to select a planner to be used and run experiments. For planners developed at LAAS, possible changes or adaptation of the plannermay be developed.Required skills: task planning, motion planning, python, rust, gitProject website: https://www.aiplan4eu-project.eu/Russel & Norvig AI A Modern Approach, Chapter on Planning http://aima.cs.berkeley.edu/newchap11.pdf Contacts: Félix Ingrand (felix@laas.fr) & Arthur Bit-Monnot (abitmon@laas.fr)Keywords: action planning; motion planning; robotics; discrete decision/optimisationDiplôma: Master or IngénieurScholarship: yesDuration: 4-6 monthsIMPORTANT NOTICE:All applications must include a motivation letter which clearly shows that you carefully read the subject of the internship, you did a little digging checking referenced papers and websites, and explain why you believe your background (educational and prof experience), the courses you have followed and your personal interest/curiosity make you believe your are fit to apply. Any application without this letter will not be considered. Applications must be sent using: https://www.laas.fr/ost/node/167