Missions

The engineer will join the Robotics and Interactions (RIS) group at LAAS-CNRS, whose activities lie in the application of AI techniques to autonomous robots.
In this context, the team works on an automated planner to enable autonomous decision making of one or multiple robots in complex situations : Aries (https://github.com/arbimo/aries). The planner relies on the composition of various techniques from AI (automated planning, constraint programming) and automated reasoning (SAT/SMT).

The work of the engineer will be on the improvement of the planner as well as on its integration both with LAAS' robotic platforms (eg. PR2 robot) as well as in an automated planning library developed within the AIPlan4EU project (https://aiplan4eu.fbk.eu/).

Activities

The work of the candidate will revolve around the Aries planner and will cover several aspects within the duration of the contract.

- Direct contribution to the Aries planner. The objective is to improve the solver which can be done with algorithmic contributions (depending on the candidate's experience in combinatorial techniques) or peripheral functionalities, (multi-threading support, parsing, incremental operation, ...). Involved skills: Rust, Planning/Scheduling, SAT/CP.

- Integration of the solver with the AIPlan4EU project. Integration will require exposing python bindings of Aries. Direct contributions to Aries and to the python library developed within the AIPlan4EU project will be necessary to extend its operating modes. Involved skills: Python/Rust, gRPC, AI planning.

- Support for integrating Aries and related tools within a robotic architecture. Integration of automated decision tools within a "Sense-Plan-Act" loop for LAAS' robotic platforms. Involved skills: robotics, ROS, Gazebo.

This assignments cover a large spectrum and can be focused on particular aspects to match the experience and aspirations of the candidate.

Aries is an open-source planner developed in Rust. Hence Rust will be the primarily used programming language and it is expected that the candidate already masters it or is willing to learn it.

Skills

- Master or engineering degree in Computer Science.
- Knowledge in algorithmic, combinatorial optimization or Artificial Intelligence.
- Experienced in programming, ideally with the Rust programming language. Previous knowledge in Python or C++ would be very valuable as well.
- Knowledge of classical robotic tools (ROS, Gazebo) and software engineering (CI, docker, gRPC, ...)
- Good working knowledge in english (written/spoken)

Work Context

The candidate will join the Robotics and Interactions team at LAAS-CNRS. The team positions itself at the intersection of AI and robotics to treat emerging problems in the decisional autonomy of robotic systems, with a particular emphasis on those involving the interactions with other robots or humans.
The team notably has a long history and strong contributions in the application of automated planning systems to real robotic systems and in their integration within their decision architecture (which notably involve motion planning and control, perception, control and execution).

The engineer will work on site in close interaction with the team's researchers and doctoral students and will be brought to interact with the academic and industrial partners of the AIPlan4EU project.

Constraints and risks

On site work.

Additional Information

For any question regarding the position, please feel free to informally contact Arthur Bit-Monnot before sending an application.

We talk about it on Twitter!

Software engineer with applications to AI for robotics (TOULOUSE) https://bit.ly/3huGfQl #Emploi #OffreEmploi #Recrutement

— EmploiCNRS (@EmploiCNRS) Thursday, 08 July, 21