Page 75 - Modern Robotics Building Versatile Macines

REAL-WORLD ROBOTS 55

people in ordinary homes, not graduate students interested in eso-
teric aspects of human psychology.”
In the following years, a number of other companies also decided
that robot toys and “pets,” such as Sony’s Aibo dog, were the wave
of the future. iRobot and Hasbro responded with a walking dino-
saur robot for the 2002 season.

Household Robots: A Different Approach

As Angle, Greiner, and Brooks looked for new robot applica-
tions, one that soon came to mind was building a smarter vacuum
cleaner. While dolls were something of a niche market, virtually
everyone owned a vacuum cleaner and faced the regular, tedious
chore of pushing it around the house in the never-ending battle to
eliminate dust.
The question was how to build a robotic vacuum cleaner that
could do its job virtually unattended. While at MIT, Angle, Greiner,
Brooks and other researchers built generally very complex experi-
mental robots. As Angle told Robotics Trends,

When I was at MIT, I built a robot that had everything you could
dream of in a robot, it was small and had 23 motors, 11 computers,
150 sensors, everything. It was over the top, but it could never work
for more than an hour at a time.
After I finished it I thought, what’s next? I had built the robot
of my dreams, but I was not satisfied. The answer was that it was
a research project. It was not a robot that would touch people in
their daily lives. The visions portrayed in science fiction novels had
not yet been achieved. I wanted a robot that could clean my house.
I wanted to do something other than making interesting “one off”
robots.

Angle and Greiner learned that it is not easy to turn high-fly-
ing robot ideas into viable products. First, there is the difference
between the design of a research robot and a product that has to
be cheap enough to find a consumer market. As Angle noted to
Business Week Online: