Reproduced with permission from TECH DIRECTIONS Magazine Jan
1994 Issue. (C) Copyright Prakken Publications, Inc.

/**********************************

*** TECH DIRECTIONS TEXT BEGINS ***
**********************************/

            SO, YOU WANT TO TEACH ROBOTICS...

     What can you teach with robotics in your classroom?
         Here's a short guide from a producer of an
   educational robotics line that shows how math, science,
          and even the nebulous field of "robotics"
                 fit into a tech ed program.


   TEACHERS can tell you their bottom line when it comes to modern
robotics: "What, besides robotics (whatever that is), will it
teach?"
   And Students want to know: "Can I play with it, or do I
mostly just watch?"
   Both teachers and students can benefit for a low-cost, easy-
to-use, flexible, and *fun* robotic construction system. Beyond
just "robotics,", such a system can help teach principles of
technology in physics, problem-solving research and development,
trigonometry, quadratic equations, and machine control. What's
more, the magic of instant robots today can catch the fancy of
the students who will one day give us the smart machines of the
future.

   IGNITING INTEREST. Photo 1 shows a throwing robot built of
three coordinate servo motors. If you know to do it,
construction time for the thrower is about seven minutes, and
programming will take another five. This exercise, more
qualitative than quantitative, is useful for igniting students'
interest and letting them know this will be a hands-on unit.
   Work groups of three or four students have the assignment of
building a machine capable of throwing a table tennis ball one
meter into a cup. The teach can offer them graduated hints:
   Use the spoon on an arm of the servo.
   Use more than one servo in a chain.
   Try different starting and ending positions for the motors to
maximize range.
   Student constructions might surprise you - a machine that
bats the ball instead of throwing it; a thrower that gets the
ball in the cup on one bounce; or an arm with a reach extended
by a thin wooden dowel and rubber band, or other materials
available in your classroom.
   The qualitative (though not yet quantitative) lessons in the
automated throwing exercise are substantial. Perhaps foremost,
students engage in a creative team effort to solve a problem for
which there are several solutions. Their efforts can yield
results in a single class period, presenting a snapshot of real-
life R&D, right down to competing with other teams trying to
solve the same problem.
   Further mining the "real-world R&D" theme, you can isolate
teams around the classroom so they can't see each others'
ongoing "research." This may even generate more original ideas,
since each group knows it will get sole credit for its own work.
In a sense, this is what patents are all about.
   Other qualitative lessons involve getting first tastes of
design optimization, low/middle/high trajectory, machine control
and repeatability, and the potential for machine hazards
presented in a generally low-hazard format.

   QUANTITATIVE CONCEPTS.  Photo 2 shows a ladling robot that
demonstrates many *quantitative* concepts while still providing
an engrossing experience. The problem is to move rice from one
bowl to another as quickly as possible without spilling any in,
say, two minutes.
   Qualitatively, students work in detail on an optimization
problem, where their variation of process parameters (speed,
acceleration, angles, and so on) will result in different
yields. They may need to develop some "techniques," for
instance, making the robot shake the spoon slightly to dislodge
any hanging rice grains, since they must perform the task
without spilling.
   Quantitatively, this exercise touches many areas of
technology: In physics and mechanics, students see the results
of differing accelerations first hand. As an exercise, they can
compute how long it takes to reach a speed of S m/sec from a
standing start, given an acceleration of A m/sec/sec. Teachers
can assign advanced students to solve for time of transit, given
an acceleration, a maximum speed, and a distance to be covered.
This problem involves only quadratics, but the complete solution
has some subtleties.
   In the area of quantitative experimental design, student
groups may prepare written reports, brief or extended, detailing
what structure and accelerations they tried, which settings
spilled rice, which did not, and so on.
   Finally, in the area of "robotics" itself, students will see
that it takes three joints (or DOF's = degrees of freedom) to
reach points in a three-dimensional region. You might inspire a
spirited discussion using the human arm as a model.
   The shoulder, being a ball joint, has two DOF's, and the
elbow adds one more. These alone would allow the hand to reach
any point in a three-dimensional area. The rotation of the upper
arm and forearm, plus the "flop" of the wrist provide three more
DOF's. These six DOF's allow you to put your hand anywhere in a
three-dimensional (x,y,z) region to any angular orientation
(a,b,c). The slight side-to-side "rock" of the wrist provides a
redundant "fine control" DOF.
  You can end with a question for thought: "What would an arm
look like that could reach into a vending machine through the
bottom chute?" Your students' "standard" arms can't do it, but
some automotive spray-painting robots can.

   DEMONSTRATING TORQUE. Photo 3 shows an elementary "torque"
demonstration. The servo motor can lift only half as much if the
arm length is doubled. You may then want to show an exploded
diagram of a servo motor to point out the gear train as a means
of torque amplification, providing higher torque but at lower
speed.
  The "closed form" of the torque demonstration, coming after
the competitive and free-form thrower and spooner exercises, can
provide a refreshing return to classroom normalcy.

   TRIGONOMETRY. Photo 4 shows a pick-and-place robot arm. By
now your students will know that they can get this arm to move
small objects around. By itself, a robot arm is not necessarily
interesting; its trigonometry, however, is.
   If we know all the joint angles and lengths of the arm, how
do we compute where in space (x,y,z) the gripper is, and which
way (u,v,w) it is facing? The full answer to this question is
beyond most college students, but fortunately, you can focus on
a small and interesting piece of the whole problem, as follows:
   "If I have a single servo motor and a straight member that it
swings, then can you tell me where the end of the straight
member is, in x and y coordinates, if I give you the length and
swing angle?...
   "Right. It is cosine times length, sine times length. Very
good.
   "Now what happens if we put a second servo on the end of the
first straight piece, rotating in the same plane? Where is the
end now?...
   "Right again..."
   The foregoing, of course, is a dream sequence. But having
gone this far down the robotic path, some of your students might
want to figure out where the griper is. And if some of them can
tell you where the flat, two-joint arm ends up, then you can
give them the real problem... "For that same flat arm, given the
x and y coordinates of the arms end, solve for the joint
angels." Very tough - and very real.
   Trigonometry aside, the robotic arm is useful in other ways.
The effect of gear "backlash," known to all machinists, can be
illustrated by moving the arm to the same end point, say, where
it is about to pick up a marble, from different starting points.
Much off the variation of where the gripper ends up comes from
the "slack" in the servo gear trains. This slack is taken up in
a direction that depends on the last motion. Any of your
students' machine shop experience will help you here.
   Photo 5 shows a pair of robotic fingers twiddling a table
tennis ball. This is pure robotic fun, and takes only a few
minutes to assemble and program. You can use it as a quick
"treat" project or an eye-catching demo as students file into
class.
   REAL RESEARCH. Photo 6 shows an exploratory robot arm in use
at a General Electric R&D lab in Schenectady, NY. The robot was
used in a demonstration-of-concept project for automating a
section of an analytical chemistry lab. The scientist in charge
said it took him just 15 minutes to program the art to pick the
vial from the rack, maneuver it through the weight scale shield,
put it on the scale, move the vial to a stirring machine, and
finally return it to the rack. He commented, "The Robix system
performed this task 50 times (all we did) without spilling a
drop."
   Photo 7 shows the Robix(tm) RCS-6 Robotic Construction
System, removed from its storage transport case. The RCS-6 was
designed specifically for educators, students, and industrial
modelers to provide sophisticated but easy-to-use exploratory
robotics at low cost. The robots in Photos 1-5 were created
using the standard kit. The GE lab added a higher-torque servo
for more lifting power, at a cost of about $50.
   The Robix(tm) RCS-6 is available for $550 directly from its
maker, Advanced Design, Inc.

/********************************
*** TECH DIRECTIONS TEXT ENDS ***
********************************/