Visit Your Local PBS Station PBS Home PBS Home Programs A-Z TV Schedules Watch Video Donate Shop PBS Search PBS
SAF Archives  search ask the scientists in the classroom cool science
scientists from previous shows
cool careers in science
ask the scientists
PREVIOUS SCIENTISTS

Photo of Rodney Brooks Almost Human

Is it possible to create a computer that mimics a human being? That's the goal of Rodney Brooks, who hopes his robot Cog will have the intelligence of a six-month-old human baby. Following the Frontiers special Robots Alive! , Rodney answered viewers' questions in an Ask the Scientists panel. Here are viewers' answers and his answers:



QOn the show I heard you say that robots being star fleet commanders was one of your fantasies. Did "Star Trek" or something similar inspire you to construct Cog?

HAL 9000 in 2001 was an early inspiration which came along when I was first trying to build intelligent machines as a boy and young teenager. Isaac Asimov's robots were more fodder for my imagination. Commander Data from Star Trek Generations is the fictional pinnacle of such developments, and so of course it is an ongoing inspiration.



QWhen do think that you will consider Cog to be finished or completed?

I don't expect to ever "complete" Cog. At some point it will get too clunky from too many patches on top of patches and we will decide it is better to build Cog II, rather than push on with the old Cog. I'll eventually feel we have succeeded if we ever get to the point where people feel bad about switching Cog X off.



QIt is my understanding that the components of the control system for a robot like Cog or It should be relatively simple by themselves, but complex in their interactions with each other. How do you design or decide the limits that are placed on each module in order to ensure that they are simple enough, yet will still yield interesting behavior when they interact with each other? Or are you required to tune each module after they are seen together?

We look at what is known about animal and human systems from neuroscience and cognitive science. This gives us some clues on how to decompose things. Then we think things through using the razor of "simplicity" in thinking about how the pieces should be designed. When we implement our design we often find things don't quite work as we expected. We patch things until it starts to work as we hoped, but often along the way we find ourselves travelling down a garden path, making some piece, or some interaction way too complex. Then we have to back off and rethink, and eventually repartition our design into simple pieces again.



QI'm curious to know what ideas you have for the future. Is Cog just a project to see if 'it' can be done, or do you see a use for this kind of technology? I think that what you are doing is cool and interesting.

Cog is meant to be the furtherest out project I could think of that would still be realistic enough for us to make some real progress. I don't expect to build anything like Commander Data anytime soon, although that would be the ultimate goal. When we started building our insect-like robots twelve years ago there were no direct applications for them, but now those technologies are in a NASA robot on its way to Mars, and in lots of robots that are being demonstrated for humanitarian mine clean up, search and rescue, and in various kinds of mining; and those are just the first most cost-effective applications while the technology is still relatively expensive compared to where it will be in another ten years. Likewise I expect projects like Cog (and there are now at least four such projects that have recently sprung up in Japan) to eventually spawn applications, although no particular application is driving the current research. But it is long term, and won't trickle into everyday life for at least twenty years.

A second reason for building Cog is that it provides a test bed for cognitive theories of how people think. It is turning out to be quite useful for this purpose, and so it may well have an effect on the questions cognitive scientists think to ask and study about people.



QI am interested in a career like yours, designing and building robots. What courses would I have to take in college? Do you have any other helpful information to help me get started in the field of robotics?

In high school (and college) make sure you take plenty of math -- it is the foundation for all good engineering. In college you could major in mechanical engineering, electrical engineering, aeronautics and astronautics, or in computer science. Robotics is very interdisciplinary and so except at a very few colleges there is not a major that is exactly fitted to robotics. While an undergraduate see if there are any robotics projects on your campus and see if there is any way to become an undergraduate research assistant on the project. Hands-on experience is the best way to learn about all the interdisciplinary aspects of robotics.

Whatever major you take, try to at least get the core courses in each of mechanical and electrical engineering, and in computer science. If majoring in mechanical or electrical engineering take some control theory courses. In computer science (or engineering) take courses in microprocessor control.



QI am very interested in building a robot. Do you have any information or hints about how to do this? Is it possible to build a robot without a lot of really hi-tech or expensive equipment?

There is an excellent book, written by one of the employees at my company and by a graduate student of mine:

Joseph L. Jones and Anita M. Flynn, Mobile Robots: Inspiration to Implementation, A K Peters, 1993, Wellesley, Massachusetts ISBN: 1568810113

If not available in your local library or bookstore, it is certainly available at www.amazon.com.

This book is aimed at the hobbyist, and is used in many high schools and colleges as a hands on text for how to build robots. Besides tutorial material on electronics, mechanics, software, and control, it includes plans for how to build certain robots, and lots of information about parts suppliers, and software suppliers. The publisher also sells a kit for one of the robots.

[I have no financial interest in this book or the kits. I just think it is the best thing around.]




 

Scientific American Frontiers
Fall 1990 to Spring 2000
Sponsored by GTE Corporation,
now a part of Verizon Communications Inc.