Yasemin Erden on BBC

AISB Committee member, and Philosophy Programme Director and Lecturer, Dr Yasemin J. Erden interviewed for the BBC on 29 October 2013. Speaking on the Today programme for BBC Radio 4, as well as the Business Report for BBC world N...


Read More...

Mark Bishop on BBC ...

Mark Bishop, Chair of the Study of Artificial Intelligence and the Simulation of Behaviour, appeared on Newsnight to discuss the ethics of ‘killer robots’. He was approached to give his view on a report raising questions on the et...


Read More...

AISB YouTube Channel

The AISB has launched a YouTube channel: http://www.youtube.com/user/AISBTube (http://www.youtube.com/user/AISBTube). The channel currently holds a number of videos from the AISB 2010 Convention. Videos include the AISB round t...


Read More...

Lighthill Debates

The Lighthill debates from 1973 are now available on YouTube. You need to a flashplayer enabled browser to view this YouTube video  


Read More...
0123

Notice

AISB opportunities Bulletin Item

Phd Position at LASMEA/Cemagref Clermont-Ferrand France


We are offering a Phd position at LASMEA in collaboration  with CEMAGREF.

Subject of phd : Obstacle avoidance algorithm in dynamic environment

Contact :
Roland Lenain : roland.lenain@clermont.cemagref.fr
Benoit THUILOT : Benoit.THUILOT@lasmea.univ-bpclermont.fr
Philippe Martinet : Philippe.Martinet@lasmea.univ-bpclermont.fr

Send your :
Curriculum
One Motivation Letter
Notes of MASTER and previous diploma

Deadline : 15th november 2008

*****************************************************************



Title: Obstacle avoidance algorithm in dynamic environment

General background
This subject takes place in the context of researches drives on mobile robots
navigation under uncertainty in the frame of TIMS (Technology and Information
support systems for Mobility and Safety) Research Federation in
Clermont-Ferrand, France. More particularly, the thesis position will take
place between LASMEA (Research Theme Rosace ) and Cemagref (Research Theme MOST
) and will aim at increasing the autonomy of Unmanned Vehicle with respect to
uncertainty linked to the robots environment, thanks to the development of
algorithms enabling dynamic obstacles avoidance. Numerous works have indeed
been completed in this research context, in particular related to mobile robot
localization and control with uncertain dynamics, but neglecting yet the
interaction with obstacles. The algorithm to be developed will then take
benefit of the algorithms already available and is expected to be based on a
global approach including:

* The mobile robot localization in a known environment. Related to the previous
and on going works, several techniques can be used with respect to the vehicle
state. Using absolute and known 3D maps, or based on a previous learning and
updating map, these various approached are based on several and various sensors
and often attached to a control approach.
* Obstacle detection, description and tracking. This detection has to be focused
on motionless obstacle (building, sidewalk, trees, hole) and dynamic one as well
(such as pedestrian, animals). The detection has then to account for obstacle
motion and dimensions, but also the capability of mobile robots with respect to
its estimated state issued from the monitoring.
* Monitoring of the vehicle state and its interaction with environment. The
navigation under uncertain dynamics (in particular in an off-road context)
requires the estimation and reconstruction of both the interaction behaviour
between vehicle and its environment (such as adherence, perturbations,
stability) but also the internal robot operation (such as fault detection).
* Reference path and corridor generation based on known map. The motion control
of mobile robots requires in an assumed to be known environment requires the
knowledge of a nominal path to be followed including margin (corridor) pending
on environment and robot capability.
* Navigation strategy and control of vehicle motion for obstacle avoidance.
Several control strategies are available for obstacle avoidance accounting for
the robot motion control, the efficiency of which depends on the environment
context and robot state. The selection of an appropriate strategy with respect
to the estimated robot configuration and respecting the navigation goal (path
tracking, velocity preservation) is a key issue of this subject.

Work objectives and methodology
The proposed thesis will contribute to the design of a navigation system
dedicated to mobile robots subject to uncertain dynamics. Several strategies
for obstacle avoidance will then be proposed pending on the robot
configuration, state and its interaction with the environment. Based on works
related to the localization, motion control and detection already in
development. More particularly, the work will be focused on the modelling of
robot interaction with its environment and potential obstacle allowing to
derive a dynamic obstacle definition. Coupled with the motion control, several
techniques (such as path deformation or potential field) will be studied,
improved and evaluated to allow the robot navigation ensuring its integrity.
The interaction with trajectory tracking control laws will also be study to
preserve motion accuracy with respect to a desired path, preliminary defined.

Theoretical developments will be tested in full scale experiments thanks to the
experimental platforms available in the Research Federation. The electrical
autonomous mobile robots AROCO (for natural environment) and Cycab (dedicated
to urban environment) equipped with numerous exteroceptive and proprioceptive
sensor will be more particularly used to show the capabilities of proposed
algorithm.


Contact
The thesis will take place jointly in LASMEA and Cemagref, in Clermont-Ferrand,
France. The thesis is expected to start in September 2008 for a 3 year period.

Thesis Director:
Philippe Martinet, Full Professor LASMEA
martinet@lasmea.univ-bpclermont.fr

Supervisors
Roland Lenain, Research fellow, Cemagref
roland.Lenain@cemagref.fr

Benoit Thuilot, Associate Professor, LASMEA
thuilot@lasmea.univ-bpclermont.fr