Pioneer P3-AT mobile robot (rightmost photos: config-1 at ISR/IST)
MSc dissertation proposal 2008/2009
Vision-based
Motion Control of a
Introduction:
"Although a few of the robots of tomorrow may resemble the anthropomorphic
devices seen in Star Wars, most will look nothing like the humanoid C-3PO. In
fact, as mobile peripheral devices become more and more common, it may be
increasingly difficult to say exactly what a robot is. Because the new machines
will be so specialized and ubiquitous--and look so little like the two-legged
automatons of science fiction--we probably will not even call them robots. But
as these devices become affordable to consumers, they could have just as
profound an impact on the way we work, communicate, learn and entertain
ourselves as the PC has had over the past 30 years."
Excerpt of A Robot in Every Home, Bill Gates,
ScientificAmerican.com, January 2007.
Future robots will be more than mere tools doing isolated tasks: they
will be quasi-team members whose tasks have to be integrated with those of
humans or other robots [HRI2007]. One of the key aspects of the integration is
collision free motion among the various robots and people working in the environment.
This work is therefore mainly concerned with designing motion controllers
capable of responding efficiently, while considering the own dynamic
limitations of the robots and the constraints to motion imposed by other robots
or people.
Objectives:
The objectives of this work are twofold: (i)
Visual tracking of patterns by their colours and/or shapes; (ii) Controlling a
unicycle-type robot in order to reach the visually tracked target-pattern. At
the end is expected to demonstrate a point to point navigation capability by a
mobile robot under the view of vision sensors.
Detailed description:
Controlling the motion of a mobile robot is still a very active
research-and-development topic mainly because the large variety of applications
and the available locomotion methods and sensors. In this work-proposal the
robots are of the very common unicycle-type, i.e. carts or cars moving in a
two-dimensional world having two parallel driven wheels, one mounted on each
side of their centre, and one or two offset castors to maintain balance. These
vehicles allow simultaneous arbitrary rotation and translation being only
constrained in the sideways motion (the sideways motion cannot be instantaneous).
The main sensors to use are computer vision and odometry.
More specifically, the work-proposal consists of controlling one
unicycle robot to navigate to a target location following a reference path. The
target location is defined by an off-board video
camera. The path following, or tracking, can be aided by both the off-board
and on-board sensors. The controller must consider the dynamic
constraints of the mobile robot and re-adjust
the motion according to unexpected obstacles
(e.g. other robots or people). In other words, the on-board sensors will be
allowed to a certain extent to redefine the path to follow, based on the
characteristics of the robot and the state of the environment. The control
methodologies may involve some visual path following [Gaspar00], visual servoing [VISP], or dynamic control [Aguiar04] as considered to be the most adequate for the tasks
at hand.
The work is therefore organized in the following main steps:
1) visual tracking of patterns marking the robot
and the target, based on colours and/or shapes
2) combining the sensors information in common
reference frames, using information fusion filters
3) controller design for the unicycle type
robot, in order to move it towards the target
References:
[HRI2007] - 2nd ACM/IEEE International Conference on Human-Robot
Interaction, http://hri2007.org/
[Aguiar04] "Pose Estimation of Autonomous
Vehicles using Visual Information: A Minimum-Energy Estimator Approach", A.
Pedro Aguiar and João P. Hespanha,
in Proc. of IAV2004 - 5th IFAC/EURON Symposium on Intelligent
Autonomous Vehicles, Lisbon, Portugal, Jul. 2004.
[Gaspar00] Vision-based Navigation and Environmental Representations
with an Omnidirectional Camera, José Gaspar, Niall Winters, José Santos-Victor, IEEE Transaction on Robotics and Automation, Vol 16, number 6, December 2000
[ISR-galery] Some images of unicycle type
robots at ISR: see labmate, scouts, pioneers, ... in "Mini galeria
de fotografias de projectos
@ IST/ISR", http://users.isr.ist.utl.pt/~jag/infoforum/isr_galeria/
[OpenCV] Open Computer Vision Library, http://sourceforge.net/projects/opencv/
[VISP] Visual servoing videos, http://www.irisa.fr/lagadic/visp/video.html
Expected results:
At the end of the work the students will have enriched their knowledge
in:
* Computer vision
* Control of unicycle type vehicles
Examples of expected demonstrations in simulated and/or real
environments:
* remotely controlling the unicycle-type robot
by pointing to one reference target-object to be reached by the robot
* guiding the robot towards a target, while the robot negotiates its own
navigation in order to avoid collisions with other robots
Observations:
--
More MSc dissertation
proposals on Computer and Robot Vision in:
http://omni.isr.ist.utl.pt/~jag