[robotics-worldwide] [robotics-worldwide at usc.edu][jobs] PhD at Institut Pascal (France) - Optimal Multimodal Control for Asbestos Removal Robots

Juan Antonio Corrales Ramón Juan-Antonio.Corrales-Ramon at ifma.fr
Fri Dec 18 10:05:01 PST 2015

*Context of PhD: Bots2Rec Project*

Many houses and industrial buildings are no more healthy due to the
presence of asbestos. Nowadays, the removal tasks of asbestos in buildings
are performed by human operators wearing a hermetically sealed suit and
isolating the environment so that asbestos leaks are avoided. Obviously,
these conditions are not only uncomfortable for human operators and also
exhausting and dangerous so that each operator has to take a break after a
maximum of 2 hours of working. The European Project H2020 Bots2Rec has as
main goal the development of collaborative mobile manipulators that release
human operators for these dangerous and tough operations.

The consortium of the project is composed by an important construction
industry, 4 European SME and 2 research labs, including Institut Pascal at
Clermont-Ferrand where this PhD thesis will be developed.

*PhD Scientific Objectives*

The PhD candidate should concentrate in the development of the local
control framework of the mobile manipulator equipped with a special tool at
the end-effector for removing abestos from vertical walls. The control
approach will be based in a multi-modal strategy where the information
coming from several sensors (force sensors, vision sensors and radar) will
be combined to execute and monitor the cleaning task. This control approach
will be connected with a high-level planning developed by other members of
the consortium that will establish the global trajectory of the robot in
the environment in cooperation with other robots. This global trajectory
will be adapted locally by the control framework according to the status of
the cleaning task monitored from vision data while force data will
guarantee keeping the contact with the wall. The developed control approach
should also include safety procedures for humans entering in the working
area and data feedback for basic tele-operation by external human
operators. The developed control framework should also coordinate with the
planning level in order to update the plan of the environment for tracking
the areas of the environment that have been treated and the areas of the
environment that have to be treated next.

*Candidate's Profile*

In order to complete this project, we are looking for a candidate with the
next skills :

- Good theoretical basis in the control of robotic manipulators, visual
servoing, force control, computer vision and mobile robot localization.

- Good programming skills in C++.

- Previous knowledge in ROS, OpenCV and/or PCL will be highly appreciated.

- Previous experience with the integration of robotic manipulators (such as
Kuka LWR) in industrial projects will be desirable.

- Previous experience in SLAM techniques and localization in maps with
mobile robots will be desirable.

- Advanced knowledge of spoken and written English (C1 or C2 recommended).
Some knowledge of French will be appreciated.

- Previous research experience in the field of robotic manipulators control
and robotic manipulation is desirable.

The retained PhD student should be highly motivated for completing a PhD
thesis in 3 years, with high creativity, a solid theoretical basis in
robotics, dynamism for working in a research team of practical integration
but also autonomy for developing research.

*Candidature's instructions*

The candidates should send their candidatures via email at
*youcef.mezouar at ifma.fr
<youcef.mezouar at ifma.fr>*, *jcorralesr at ifma.fr <jcorralesr at ifma.fr>*
and *jean-christophe.fauroux at ifma.fr
<jean-christophe.fauroux at ifma.fr>*, including: (1) their CVs, (2)
motivation letter, (3) transcripts of previous university degrees
(unofficial are enough now), (4) previous research publications (if
existing) and (5) contact information of 2 or 3 people who could serve as

More information about the robotics-worldwide mailing list