that worked. The final design used longer shafts on
the wheels/output gears, and even though the legs
themselves would be unsupported on the outside,
at least they didn't collide with the shaft for the
motors and input gears.
Legs are the New Wheels
When we were designing the legs for our bots,
we tried to implement as many of the lessons from
the RHex robot as possible. We wanted to emulate
the hook design characteristic of the newer
incarnations of the RHex robot, hoping that we too
would be able to impart some bouncy efficiencies
to our bots. Fortunately, our VEX kit had the ideal
parts to turn into robot legs: four long metal frame
The two pieces were long enough to cut in half for a
total of four legs, and the metal was ductile enough to
bend into the shape we needed. The frame pieces had
holes in them that aligned perfectly with other VEX parts,
meaning they would be easy to mount. With screws
picking up the mounting holes, the wide surface of the
frame pieces was rotated 90 degrees from where we
This was easily corrected with a vice and a monkey
wrench — we simply put a 90 degree twist in the metal.
This allowed us to pick up the mounting holes on one
plane, while also contacting the ground with the wider
surface of the metal in an orthogonal plane.
We next put the hook-like curve into the leg that
would hopefully put a spring in the bot's step. To put the
same radius into each leg, we used a large socket as a
template and bent each of the legs around it. The metal
was ductile enough to be bent by hand, which caused us
some concern as to whether the legs would be able to
flex without bending when moving the robot.
To add some much needed friction to the legs, we
attached some adhesive backed foam to the ends of each
foot. We also wanted to enable the Scribbler to be
equipped with the legs, so we drilled some holes into the
Scribbler's wheels to mount to. Now, we were ready to
take another run at the obstacle course.
Having each leg on the VEX robot powered
independently allowed us to position the legs as we
pleased. We set the wheels at four different increments
of 90 degrees, hoping this would give the bot the best
chance at a stable gait. Before tackling calculus again, we
wanted to see how the bot fared on flat terrain. Did it
have the torque to spin the legs? Would the legs impart
some springy efficiency, or would they buckle under
We fired up the robot and moved the radio joystick
with some trepidation, but any doubts about the
effectiveness of our design were quickly dispelled as the
SERVO 08.2014 73
GETTING A LEG UP.
SCRIBBLER GETS SOME HAND-ME-DOWNS.