RP-able robots

From: Jim Mallos (jmallos@wizard.net)
Date: Mon Feb 07 2000 - 18:45:46 EET

Will we ever see a robot emerging full-formed from a stereolithography
vat? This is a distant prospect if we imagine making the same robots as
are now made.

Suppose instead we just look for technologies that are easy to
stereolithograph and see if we can make a robot from them? Our quest
then becomes a matter of creative robot design and some new applications
of today's stereolithography.

Here are some easy-to-stereolithograph technologies that can form a


Mark Tilden at the Los Alamos National Laboratory has shown that
interesting robotic behavior can be attained with very simple control
systems (for example a walking robot with just 12 transistors.) His
approach uses force feedback to create a chaotic system comprising both
the robot and its environment, see
http://www.patents.ibm.com/details?&pn=US05325031__ .


Fluidic. The simple controls needed by Tilden robots can be provided by
no-moving-part air fluidics. This is a developed technology, we just
need to learn how to make the components by stereolithography.


Pneumatic via flexible passageways.


Biomorphic, probably walking. Wheels and wheel-like elements such as
gears are banished from the design---they need too much precision.


Flex joints. Journal bearings are banished for the same reason as


Flexible pneumatic actuators. Figure 1 of U.S. patent 3981528,
http://www.patents.ibm.com/details?&pn=US03981528__ , shows a simple
shape that combines flex joint and actuator in one.


Flexible pneumatic actuators---they can be used passively as force and
contact sensors.


Any semi-flexible material with a wide temperature range. A
semi-flexible material can be programmed to be fully rigid (with ribs)
or fully flexible (with convolutions.) We don't need anything else---the
air in the pneumatics does the hard part.


Thermal storage, either hot or cryogenic. (Storage temperatures limited
by the tolerance of the construction material.)


Thermodynamic. Perhaps the easiest conversion method is simply to fill
the robot's little belly with dry ice. Capturing the subliming carbon
dioxide at a modest pressure can supply the fluidics and actuators with
a steady stream pressurized gas.

A more elegant technique for the long term is the Stirling amplifier. A
Stirling amplifier is simply an air-filled chamber that has a hot wall
on one side and a cold wall on the other. A porous, insulating divider
called the regenerator separates the two sides of the chamber. If the
divider is pushed all the way to the cold side, most of the air will be
hot, and the air pressure in the chamber will be increased as a result.
Similarly, the air pressure decreases when the divider is pushed back
over to the hot side. Given a big enough temperature difference between
the walls, the air pressure changes can do more work than it takes to
move the divider. If the divider is itself actuated by pressure, this
forms a signal amplifier that can perform control functions, or amplify
a pneumatic output signal to drive an actuator. (Notice that this system
works with reversing AC flows instead of a DC stream of gas.) The
Stirling amplifier can be energy efficient because most of the
tranferred heat comes from the regenerator, not the walls which are
maintained at temperature using the thermal storage.

How long do you think it would take to engineer such a robot?

Jim Mallos

For more information about the rp-ml, see http://ltk.hut.fi/rp-ml/

This archive was generated by hypermail 2.1.2 : Tue Jun 05 2001 - 23:02:47 EEST