Robotics Reference
In-Depth Information
-8-
Things to Do for Robots
T
his chapter is intended as a kind of pivot, between the “how com-
puters have become intelligent” of the preceding part of the topic,
and the “why and how robots will become much more intelli-
gent” in the subsequent chapters. Here then is a summary state-of-the-art
survey, my personal selection of remarkable aspects of robot technology
and intelligence which, I feel, typify where we are today (2005). Most of
them reveal some of the achievements of AI researchers during the past
50 years, and any of the robots described in this chapter could also be
designed and programmed to perform any or all of the other tasks de-
scribed hitherto in this topic. The exception is a novel technology, which
promises eventually to enable robots to be self-sufficient in terms of the
electrical power they consume.
There is no intention here to discuss the nuts and bolts of robotics,
of the artificial muscles that enable the limbs and other parts of a robot
to move, or of the electro-mechanics of the moving parts themselves.
Instead, and with the exception of the section on gastrobots, we confine
our interest to those capabilities of robots that require intelligence.
Robot Soccer
Since the mid-1990s robot Soccer has become one of the most highly
competitive topics in AI development, demonstrating how intelligent ro-
bots can work together in a co-ordinated manner, communicating with
each other, agreeing on how to solve a problem (scoring more goals than
the other side in a Soccer match) and then splitting a task amongst them
The first MiroSot World Cup Soccer Tournament took place in Ko-
rea in November 1996, with 23 teams from ten countries, and was won
by Newton Research Laboratories of Seattle. This particular event and
the organisational infrastructure that created it quickly blossomed into a
fully-fledged international association called FIRA, which organises an-
nual world championships. There are various categories of robot Soccer
