|
Controlling Self-Reconfiguration of Self-Reconfigurable Robots |
|
Self-reconfigurable robots are built from modules which are
autonomously able to change the way they are connected, thus changing
the overall shape of the robot. This process is difficult to control
because it involves the distributed coordination of large numbers of
identical modules connected in time-varying ways. We present an
approach to self-reconfiguration where the desired configuration is
grown from an initial seed module. Seeds produce growth by creating a
recruitment gradient, using local communication, which spare modules
climb to locate the seed. The growth is guided by a novel
representation of the desired configuration, which is automatically
generated from a 3D CAD model. This approach has two salient features:
(1) the representation is concise, with a size proportional to the
global shape rather than the number of modules and (2) there is a
clean separation between the goal and the local, goal independent
rules used by the modules. Download a video here, here, or alternatively go to the
publications section for the papers. |
Controlling Locomotion of Self-Reconfigurable Robots |
|
I investigate how to achieve global coordinated behavior of a
self-reconfigurable robot built from many independent modules. We use
a bottom-up approach where local interactions between the modules
produce the desired global behavior. We call our approach role
based control. See the publications
page and the CONRO homepage for
more details. Maybe you just want to see the videos |
|
Situated Communication |
|
I hypothesize that in multi-robot communication the physical signal that carries
a message is as important as the symbolic content of the
message. The physical signal alone contains information about the
location of the sender relative to the receiver. This information is
traditionally hard to come by and is very useful in many multi-robot
applications.
|
|
Player |
|
I participated in the initial development of Player.
Player is a device server for the ActiveMedia Pioneer 2-DX robot.
(more..).
|
|
Ant Inspired Robotics |
|
We look at how virtual trails can be used to coordinate the behavior of
multiple robots (a lot more..).
|
|
Robot Learning - a "Survivor" |
|
The goal was to develop as robust a robot as possible. The idea was
to develop a "survivor" robot that would continue to pursue its task
even if its sensors become severely damaged. We wanted to reach this
goal by using robot learning (more..).
|
