[robotics-worldwide] [jobs] Postdoc in motion planning for drone detect & avoid at ENAC (Yannick Jestin)

classic Classic list List threaded Threaded
1 message Options
Reply | Threaded
Open this post in threaded view

[robotics-worldwide] [jobs] Postdoc in motion planning for drone detect & avoid at ENAC (Yannick Jestin)


-------- Message transféré --------
Sujet : Postdoc in motion planning for drone detect & avoid at ENAC
Date : Mon, 17 Sep 2018 17:43:24 +0200
De : guido <[hidden email]>
Pour : [hidden email]

The french National Civil Aviation University (ENAC) is seeking a
postdoctoral fellow in motion planning for drone detect and avoid.

To contribute to the integration of Unmanned Air Systems (UAS) in
commercial traffic, 2D and 3D Detect And Avoid (DAA) algorithms based on
Velocity Obstacles (VO) [1] have been studied [2]. This class of
algorithms is yet purely reactive over 10 seconds time steps, not taking
the past or future of trajectories into account, and ignores the flight
plan or goal of the UAS (and intruding aircraft).

To improve the performance of these algorithms, their horizon must be
extended to be able to plan more complex paths which can escape dire
conflicting situations or optimize the return to the UAS mission. To
this end, the candidate will explore the learning of trajectory
prediction based on data sets of collected real commercial traffic
trajectories to help built robust and efficient conflict-free path for
UASs. Additionally, the candidate will explore the integration of our
algorithm within a tree search algorithm to obtain good quality
solutions over a larger horizon. The optimization process will also take
into account intruding aircraft trajectory prediction, if available, to
improve the robustness of computed plans.

[1]  P. Fiorini and Z. Shiller, “Motion planning in dynamic environments
using velocity obstacles,” The International Journal of Robotics
Research, vol. 17, no. 7, pp. 760–772, 1998.

[2]  C. Allignol, N. Barnier, N. Durand, and E ́. Blond, “Detect &
Avoid, UAV Integra- tion in the Lower Airspace Traffic,” in ICRAT 2016,
7th International Conference on Research in Air Transportation, ICRAT
2016 Proceedings, (Philadelphia, United States), June 2016

The successful postdoc candidate will be offered a 12-month contract
with competitive salary, in an outstanding working environment.
Contact :* Yannick Jestin ([hidden email])

*Required Scientific background**: *machine learning, combinatorial
optimization, robustness, motion planning.

*Development environment: *OCaml, R, Python, GIT, Linux.

*Qualification:* PhD diploma

*Prior e**xperience:* none

*Starting date:* 01/11/18
**Deadline for application:* 15/10/18

*Duration:* 12 months

*Location:* ENAC, Toulouse, France.

*Salary: *competitive salary

robotics-worldwide mailing list
[hidden email]