J.M. Ferrà ndez
Robotics and Autonomous Systems, No.
58, 2010
Summary
The paper is an editorial introducing the linkage today
between robotics and neural science, going through a little bit of history of
robotics.
Autonomous robotics is
cited as the most complete paradigm of Artificial Intelligence, since it is
characterized by perception, planning and action.
Mobile Robotics, today
more an more under the attention of researches, is not only dealing with
complex program and non-analytic human knowledge, it is also managing to
uncertainties and complexity and error measurement, making it incorporated also
with typical problems more strictly related to robotics, in addition to
dependence on technology development.
In the sixties there has
been a great improvement in Autonomous Robots research, some principles of
robotics where applied during a time in which a “General Problem Solver” was
researched and the result at the time was “Shakey”, a mobile robot that used
memory and logical reasoning to solve problems and navigate in its environment.
“Shakey” at the time was a great leap forward, but it was still lacking in
completely being autonomous, in fact is was controlled by an off-board computer
and it wasn’t able to operate in high uncertain area. The seventies are
depicted by the author as a “technological
bottle-neck” for robotics, only in the mid-eighties there has been a come
back, mainly in the concept laying behind autonomous robotics, since the idea
of “modular computation” was introduced, with the aim of avoiding the problem
of huge knowledge databases trying to solve infinite problem cases, but
introducing modular programs capable of being called when required.
The idea that sensors and
effectors are coupled with perceptions and actions brought in the concept of
“conduct based” paradigm, denoting similarity between robots and ethology
studies.
Here is were the origins
of cybernetics are (“Embodiment of Mind” – McCulloh, Kilmer) and it developed
in being a science covering robotics, psychology and biology, making it in the
end a bioinspired robotics.
The XXI century is
denoting the challenge of both scientific disciplines: Autonomous Robotics and
Neuroscience and there is where we can see the attempt of controlling robotics
using biological neuroblatoma and EEG applied in the field of brain-computer
interface.
Studies on robotics team
working through social rules is another way of applying ethology studies in
robotics and studies are done also in observations in how humans select routes
in crowded environments.
The author reports that we
are still far from obtaining results, but that this is the direction of this
coming period.
Key
Concepts
Autonomous Robotics,
Artificial Intelligence.
Key Results
The paper results interesting under the aspect
of which are the key points that have been necessary to achieve the real
concept of autonomous robotics. As stated, we had to wait for the mid eighties
before entering the concept of modularity in programming languages and this
helped a lot in further steps, making it possible to improve control systems
and communication systems.
Interesting is the
parallelism between ethology and robotics, which is remarked also by inventions
and studies done in this direction, making robots as animals.
It would be interesting to
understand whether this concept is actually applied under the idea that the
robot is an external entity or if it can be viewed as a tool, not as it used to
be thought till the eighties, but as biological tool, part of the human being,
as it has been shown in studies regarding EEG applications in robotic control,
specially in the case of assisting people with sever disabilities.
No comments:
Post a Comment