Proposal for the Artificial Intelligence Research Paper
By Margit Lang
October 22, 1999
ROBOTICS
Robots in Science Fiction and Reality
The topic of my research paper is Robotics, or better to say robots. I want to write about robots (or androids) and analyze them. I want to figure out what robots we build now are able to do in comparison to androids from popular films like C3-PO, R2-D2 from Star Wars or LdtCmtr. Data from Star Trek, Starship Enterprise The Next Generation. And what will our robots be able to do in the near future? Is it possible that there will be such androids in anytime far from now?
I want to pick up some points like walking on two feet, which has turned out to be a bigger problem than scientists have thought. Or it was even hard to build robots, which find the right way through a labyrinth, without pumping against the walls. All the things we might think are trivial to do for living beings are quite complex to do for computers and robots.
At the Carnegie Mellon University there are a lot of projects on their way. One project is about a helicopter, which should fly on its own. It measures all the factors, like the strength of the wind and the direction, watches out for any handicap that might appear, and so on.
In Japan some scientists build a mechanic dog. (It might be available aver there at Christmas!) This dog looks like a small puppy. It has sensors on its head, where you can stroke it. It can do anything you would expect form a new Japanese toy. But it also can play a sort of soccer. It is "trained" to run after a ball and kick it into a goal. Well it is really a nice toy, but the knowledge behind that is immense.
It’s hard to learn the right connections between perception and action. For robots it is still difficult to see the real environment. A robot cannot get to know what is really going on.
It just can focus on a part of the surrounding, and so it misses a lot of information about what all around it.
Another aspect I want to mention in my research paper is thinking. Is a machine able or will it be to think for itself, make its own decision. Where are the difficulties to "teach" a robot how to think? Will robots ever be able to have feelings? A major part of the android Data from the Enterprise is to learn to be human. How can you simulate feelings anyway? And is there really the danger that robots will overtake the world in the future? Can this ever happen?
Unfortunately most of the books in the library are not up to date. So will concentrate my research on articles in the Internet. Especially I will use publications from the Robotic Institute of the Carnegie Mellon University about their research and projects.
References:
Robotic Institute Seminar, Carnegie Mellon University
http://www.ri.cmu.edu/seminarDeken, Joseph: Silicio Sapiens, The Fundamentals and Future of Robots
1985, Bantam Book
ISBN: 0-553-25365-4
Hirose, Shigeo: Biologically Inspired Robots, Snake-Like Locomotors and Manipulators,
1993, Oxford University Press
ISBN: 0-19-856261-6
Caudill, Maureen: In Our Own Image, Building an Artificial Person
1992, Oxford University Press
ISBN: 0-19-507338-X
Cohen, John: Human Robots in Myth and Science
1967, Barnes and Company
Simons, Geoff: Is man a Robot?
1986,
ISBN: 0-471-911062
Asimov, Isaac and Frenkel, Karen A.: Robots Machines in Man’s Image
1985, Harmony Books / New York
ISBN: 0-517-55110-1
Marrs, Texe: The Personal Robot Book
1985, TAB Books Inc.
ISBN: 0-8306-0896-6