Tapping is a mode of manipulation well suited to certain tasks such as
micropositioning. If you want to make a small adjustment to the
position of an object, you can give it a small tap. The basic physics
of tapping an object involve impact followed by sliding. In our
research, we are developing algorithms that incorporate an
understanding of the mechanics and underlying strategies for tapping.
We are creating a robotic positioning system based on this mode of
manipulation. Suppose we have a special "tapping actuator" that can
release a rod that strikes an object. We are building a number of
such actuators that can deliver taps of different strengths and then
automatically reset themselves.
Imagine a "tray" which is surrounded by several tapping actuators. We
can then put a part in this tray, measure its position, and then fire
one of the tapping actuators. By repeating this process, we can
position that part very accurately on this tray.
The research questions in this project are how to devise an
algorithm that will position parts with the fewest number of taps.
There are also problems related to the configuration of the
micropositioning cell. For example, where is the best place to put
the tapping actuators? Or even: given a part, how many tapping
actuators are needed?
This work is supported by the National Science Foundation through
award IIS-9977562.
Wes Huang /
whuang@cs.rpi.edu