[The 'o's are the articulation points.]
The fastest gait (walking pattern) for a hexapod is the 'alternating tripod'
where the weight is supported on legs 1-4-5 for the first step, then on legs
2-3-6 for the second step, then back to legs 1-4-5 for the next step, and so
on cycling between the two 'tripod' sets of legs for alternate steps.
As Flik starts up, it rises to its full height with all six legs fully extended.
To walk, it lifts the tripod that will be free for this step (1-4-5, say) leaving
the support tripod (2-3-6) fully extended and twists the articulation points to
take a step. At the end of the step Flik fully extends the free legs to form
a stable hexapod with the fully extended support legs, and then lifts
the support legs for the next step. The advantage
of this is that, once the hexapod is at its full height, little energy or time
is required to shift from one set of legs to the other. The disadvantage is
that each leg requires a motor (two leg motors in each section).
By contrast, Hexapod_I uses a rack through the center of each section to shift
from one leg to another -- only one leg motor is required in each
section. However, it also means that the bot must 'sit down' in the
middle of each step while it is shifting weight from one leg to another. The
result is that it must 'climb' back up from a gravitationally stable position
each time it shifts weight from one tripod to another, and this takes more
time and energy than the method used in Flik's design.
Note on walking
The ground covered by an articulated hexapod design such as Hexapod_I is
proportional (all other things being equal) to the angle through which the
articulation points move. If you are contemplating this kind of design you should
- design for the tips of adjacent pairs of legs to pretty
much meet when the articulation is twisted at its maximum angle.
- design that the distance from the articulation point to the point
where legs exit the robot is the same for each segment.
What's nice about Hexapod_I is that it is a *big* robot with a big turn angle
on the articulation points -- it's pretty menacing-looking when it's walking
around the room.
Note on turning
All in-line hexapod robots have difficult turning, and Hexapod_I is no exception.
The turning radius is large and uses a fairly ugly 'skid-steer' approach.
CONTROL
Control was primitive -- two 'whiskers' at the front led to touch sensors.
The program walked forward until the whiskers detected an obstacle and then
the robot backed and turned to try and walk around it.