Re: Robot O.S. and automation.
- --- In SeattleRobotics@yahoogroups.com, "dcwjobs2004" <dcwyland@...> wrote:
>Not the same story, but David.B put me onto Ron Goulart, circa
> Helen O'Loy is a robotic car from an Asimov (IIRC) short story. It was faithful to "her" owner/driver, to the point of "killing" a rogue robotic car that threatened him. After that, the owner/driver rejected "her" because "she" acted on "her" own, without human command.
1970s, who I had never heard of before.
The story in Screwloose called "Into the Shop" is about a so-called
**lawagon** which is detective, marshall, pursuit vehicle, judge,
jury, executioner, and cremator all rolled into one, and typically
with the last 4 functions finished off within 10-sec after the
criminal is apprehended.
Moral: whatever you do, don't let the robot be executioner.
> --- In SeattleRobotics@yahoogroups.com, "dan michaels" <oric_dan@> wrote:
> > --- In SeattleRobotics@yahoogroups.com, "David Buckley" <david@> wrote:
> > >
> > > They are getting on a bit but last year when I gave an invited talk at the annual Cybernetics Society (CybSoc)conference in London in 2009 I talked with a number who were still alive and most of them had no idea of the complexity of the tortoises'* system. On a simple cybernetics level he was doing one thing but you must remember that as well as being a neurologist, a namer of some brain-waves, he was also an expert electronics engineer writing papers (peer reviewed) in both fields. As well being an international expert on child development and a radio-star.
> > > *They are tortoises not turtles because - they taught us - It is to do with Alice.
> > > How they work is a bit like Helen O'Loy.
> > >
> > Hmmm, Helen O'Loy is a new one on me, but sounds a little
> > whimsical for my robots. But there is certainly room for more
> > than simple stimulus-response bots.
> > http://en.wikipedia.org/wiki/Helen_O'Loy
- --- In SeattleRobotics@yahoogroups.com, "Randy M. Dumse" <rmd@...> wrote:
>Randy, I glanced at the Crash Course article that Daniel cited.
> D. Daniel McGlothin said: Wednesday, February 09, 2011 11:45 PM
> > I've no proprietary interest in the QP offerings.
> > However, Samek's book on Practical Statecharts
> > significantly changed some aspects of how I
> > approach programming ...
> Daniel, do you understand Samek's hierarchical state machines? I
> never got the concept of hierarchical, and I suspect there's
> error in his thinking. I've got the book, Practical Statecharts
> in C/C++, but never to descend into it in detail. Looks like
> his idea of a state machine is a bunch of boxes and arrows, or
> roughly speaking a directed graphic chart. I wonder if he missed
> out on the overlying theory.
Re "concept of hierarchical", Samek's idea of hierarchy is really
a set of "nested" state machines, up to at least 4-levels deep
in his examples, where the inner ones "inherit" properties/control/info/etc from the outer ones. One of the
reasons to do it this way is so event entries from outside come
into the outermost SM and affect all of the inner SMs the same.
This saves on having to replicate the same response to the same
external event in each of the inner SMs individually. Rather, if
the outermost SM responds, then all of the inner ones simply
To a large extent, what he seems to be trying to do is merge
the ideas of OOP and FSMs, or alternatively, incorporating many
of the OOP ideas into the FSM world. Thus, you have encapsulation
and inheritance as his primary add-ons to SMs.
This view of hierarchy is a little different from the sort we
usually think about with Brooks' subsumption, of course.
I don't know how people normally implement SMs, but of course,
when we write regular code, we deal with matters of code-
replication simply by writing a callable-subroutine, rather
than replicating the same code over and over in all the places
where it is used.