Robotics Reference
In-Depth Information
Figure 56.
Dog-to-sofa reconfiguration (Courtesy of Keith Kotay, Dartmouth Robotics
Lab (formerly Rus Robotics Lab at Dartmouth))
Self-reconfigurable robots are able to adapt their shape without hu-
man assistance. They are made up of a set of identical robotic modules
that can make decisions according to their circumstances. Think of one
of these modules as a kind of Lego
TM
brick that incorporates electronics,
mechanical parts and intelligence. Instead of a human putting two or
more of these bricks together in order to build something, imagine that
each brick has the intelligence necessary to decide what shape it wants
to build out of a whole collection of “Lego bricks”, and how it proposes
to build it. And furthermore, imagine that each brick can move towards
and away from other bricks, and stretch out mechanical and electronic
connectors to neighbouring bricks so they can bond together, until such
time as the whole structure decides to reconfigure itself again. You are
now thinking of something that is conceptually very close to one of the
robot modules developed by Daniela Rus and her group at MIT.
Demonstrations of how Rus' modules reconfigure themselves can be
foundonherlab'sWebsite,
13
including one showing a robot change
showing a robot climbing stairs by repeatedly changing shape. This Web
site is most definitely worth a visit. Take a look at the video clips and
imagine what will be possible when the successors to these modules can
be miniaturized and manufactured in very high volume with a low unit
cost. Reducing the length, width and height of each module to one-
hundredth of their current size, a level of miniaturization that we are
quite likely to see during the next 20 to 30 years, would result in robotic
13
Visit http://groups.csail.mit.edu/robotics/modular robots/molecule/simulation.html.
