One of the principal researchers, Prof. David Wang, electrical and computer engineering, is a member of an interdisciplinary research group -- the Construct Group. It includes Profs. Glenn Heppler, systems design engineering; Farid Golnaraghi, mechanical engineering; and 15 to 20 graduate students. Other UW faculty members are also involved, as associates, including mathematicians and physicists.

Wang's main contributions are in "control," which involves designing computer algorithms that will permit robots to do the things people would like them to be able to do and getting them to do these things better. Control problems are highly mathematical. "When we seek to control a mechanism (robot), a lot of the mathematical modeling is incredibly complex, involving some of the most difficult math problems one could conceive of."

One of the problems in designing robots that are smaller, lighter weight and more compact is controlling vibration. The larger and heavier the robot, the less impact vibration is apt to have on it. As well, "friction effects" start to become more important as robots get smaller.

"You try to get around the problems by designing robots with sensors that can detect vibration when they are in operation, and actuators that are useful in damping it down," he says. "You quickly find you must use a computer to control these vibrations, however, because a robot has to be able to react to the high frequency vibrations so very quickly. The challenge is to design for new levels of flexibility."

Wang's current research interests lie in three main areas: lightweight robots and how to design them; "haptic" applications (devices connected to computer to give sensations); and designing a robotic (pilotless) helicopter.

In the first area, he notes that the original Canadarm, built by Spar Aerospace for NASA -- and essentially a robot -- was so large it took several minutes to settle it down whenever it started to vibrate. He became involved in the search for techniques that would keep the arm from starting to vibrate, or that would shorten the duration of the vibrations if they did start.

Wang notes that typically, an industrial robot designed to lift a six-pound weight will in itself have to weigh about 400 pounds. This is because it has to be sufficiently rigid so it won't deflect under the strain of having to lift six pounds. A human arm weighs, of course, only a fraction of the weight of a robot, and it will lift a six-pound weight handily. The idea, therefore, is to get robots to be as efficient as humans.

The advantage of robots is that they can lift a six-pound weight over a very long period of time and they can do it 24 hours a day for days or weeks on end, without ever getting tired.

A truly lightweight robot could be much more energy efficient than anything that it currently on the market and consequently would be much cheaper to operate.

"Making robots smaller, lighter and stronger is all based on finding ways to deal with the problem of vibration," Wang concludes. "People have been thinking about this for a long time. But today the speed of our computers makes it realistic to expect to be able to deal with the problem better. We can also use the computer to find new ways to counteract the effects of high frequency vibrations caused by running robots at high speeds."

He is skeptical about the possibility of solving such problems simply by relying on some of the newer lightweight composite materials that are available, partly because these materials tend to be much more expensive than aluminum or steel; also, the consequences can be catastrophic if one of the composites does break.

As for "haptic" applications, the area involves devices one can connect to a computer to get sensations that you can actually feel with your fingers on your computer mouse. Wang calls it a "virtual reality mouse."

A mouse tends to be very small and lightweight, because the (human) computer operator needs to be able to move it around easily. Recent UW research in this area has already resulted in the formation of a spin-off company, Alliance Technologies Corp., that is currently much concerned with developing new ways to make it easier for blind people to use computers.

Alliance has developed a special way to move the mouse over data displayed on a computer screen, so that the fingers of the user can sense a little "bump" as they move the mouse across a line of text (printing), as well as across a border or icon. Wang likens it to going over a speed bump in a car.

Thus, a blind computer operator can sense each line of the text on the computer screen, as he or she scrolls down it, and can then tell a speech synthesizer (reading machine) to read each line aloud each time he or she senses such a bump.

"We have built some additional features into the software," Wang notes, "including gravity wells that make it easier for a sightless person to issue commands to the screen, and warning signals if the operator should get the mouse too close to the top or bottom of the screen. In other words, people who are blind can now use a computer quite satisfactorily. Working on this particular application has been very gratifying." He says American pop music star Stevie Wonder bought one and has used it to surf the web.

Because the application was developed to make it easier for blind people to use computers, it had to be user-friendly, meaning, in this case, that it had to be lightweight. "Certainly we had to get rid of friction effects such as one could get from a mouse sliding over the desk," Wang notes.

He says that haptic interfaces utilizing the Internet promise to make it possible, some day, to not only see a visible situation but also to feel it by picking up an "object" and sense how smooth it feels or how heavy it is. And all this could be done in real time. However, the unpredictability of time delays during such communication presents a challenge.

"We will have to compensate for the time delay between the moment we send the data across the network and the moment it gets there ... or from the moment they send it back to us and the moment we in turn feel it here," Wang says. "But we've had students working on this problem and they have come up with a way to compensate for a time delay of a tenth of a second. Preliminary work is now completed on taking into account a somewhat longer delay."

Wang foresees the day when such haptic devices will permit a doctor to interact much more effectively with patient in a remote northern location in Canada, though he admits "the technique isn't something I'd want to rely on if I ever had to undergo brain surgery."

As for robotic helicopter research, this area is being developed primarily by a group of undergraduate and graduate students who come together as the Waterloo Aerial Robotics Group. The students have won design awards for two years running in competition with other students in other universities.

They are currently involved in an international competition, seeking to create a robotic unmanned helicopter that will fly by itself into a disaster area in search of human survivors, including those who are badly injured. The helicopter has to be able to take off, fly, search, identify the disaster area, fly back to the base and land. It has to be able to deal with clouds, fire, smoke, and still fly safely into an area and identify survivors. The problems the students have to deal with include how to control such a flight and how to program a computer so it can reliably take the place of a human pilot.

Wang says that the project has to deal with sensors, cameras, strain gauges and other instrumentation. The student researchers try to "stay away from stuff that is too exotic, because we want to look into things that can be put into practice. But we're pretty sophisticated at the same time. On the helicopter project, for instance, we are using a GPS unit that gives us accuracy of within one centimetre."

He says the helicopter they are working on is a fairly large model with a wingspan of about five feet. When it takes to the air, it carries a substantial equipment load including gyroscopes, camera systems, very sophisticated sensors and -- a unique feature -- its own high-speed Pentium computer to handle all the data from the sensors.

Tech Space - SpaceDaily Special Report

Community
Email This Article
Comment On This Article

Related Links
Space

Memory Foam Mattress Review

SPACE MEDIA NETWORK PROMOTIONS Solar Energy Solutions Tempur-Pedic Mattress Comparison ... Newsletters :: SpaceDaily :: SpaceWar :: TerraDaily :: Energy Daily XML Feeds :: Space News :: Earth News :: War News :: Solar Energy News

SPACEDAILY

Satellite Launch To Boost DTH In India
Calcutta, India (SPX) Dec 28, 2005
The successful launch Thursday of India's heaviest satellite from spaceport of Kourou in French Guyana may have boosted the country's space research efforts to yet another level, but it has also lifted the spirits of at least three Direct-To-Home televisions broadcasters, one of which has been waiting for years to launch its services in India.