[Avery Andrews 950701.1130]
(Bill Powers (950630.1130 MDT))
>That [stuff about smart input functions producing ersatz perceptions]
>describes a result that would be useful; the problem is how to
>implement it. To "monitor some aspect of the activity of the effectors
>that are supposed to be affecting the controlled perception" would
>require sensors of a special kind (that can monitor activity of
>effectors from a non-involved point of view), and knowing that the
>effectors are supposed to be affecting a controlled perception requires
>something that knows about (senses) the effectors and understands their
>relationship to the perception in question. Then you have to add
Whether it's doable or not would depend a lot on the specific nature
of what's being done. We might for example have a perception of
`velocity' that was normally determined by visual flow of the
ground, but if this info is unavailable, computed from the
RPMs of the propellor: the RPM -> velocity function would be
calibrated whenever the ground was visible, but used when it
wasn't. This would of course fail if the currents changed while
the ground was invisible, but might be better than nothing (and
inertial sensors could deal with that, to at least some extent).
The presumed fact that this kind of arrangement will only work for
a small percentage of cases doesn't matter if those cases are
important: if you really need something in .5% of the situations
you find yourself in, you really need it (I think a standard theory
of why mammals play is that they are developing & honing skills
which aren't required often enough for adequate practice).
It's my belief that smart input functions would be pretty easy for
controlling locomotion, but pretty hopeless for, say, martial arts
(Tai Chi continues too look like a program for replacing all sorts
of crappy anticipatory strategies with good control systems).
On another topic, what's the relation between putting an integral
component into a controller and using slowing? Intuitively, it seems
to me like you might bet similar improvements in stability without
loss of precision of control by:
a) adding a slower
b) reducing the proportional component, and adding an integral
component.
Finally, the RARS stuff looks worth looking at, to me, not least
for the reason that other guys define the environmental physics
and the display, and all you have to do is produce a controller
(which can be written in plain C, if you don't want to learn C++).
And if you don't like some aspect of the physics, there's a
suggestion list to post your dissatisfactions on.
Avery.Andrews@anu.edu.au