| This week's evening hobby fun time was mainly
spent assembling Bing's lower legs and designing and starting
construction of little bits for Bing's upper legs. While
this is the main task, there is not really much to write
about, as the procedure is now almost habit. Four or
five more weeks of this, and then (with luck) I'll reach the
main first goal:
Turn Bing into a software problem.
Of course, I fear that various pieces of Bing will need to
be reconstructed due to oversights in the design, but the
basic idea is to have a platform that would be capable of
walking if only it knew how. I have two fears about
reaching that goal, neither of which will be answered until
Bing is constructed: 1) Will the servos provide the
power and speed necessary for walking? It will do me no
good if poor little Bing isn't strong enough to lift its legs
off of the ground. If that happens, I'll have to rebuild
it with larger servos and making the deadline would become
highly dangerous. 2) I'm not entirely certain that
Bing's basic design is completely sound. My particular
fear is that the attempt to lift a leg will result instead of
Bing flopping forward on its "face". If I'd
taken some engineering classes between parties in school I
could do a better job validating the design concept, but now
it seems far too much like work to learn statics and dynamics
and do vector sums. We'll just see.
Even as just a software problem, it is becoming clear that
it might be a difficult software problem (who would
have guessed?) At a minimum to start I'll at least need
to monitor what is going on in Bing better than I can do
watching text debugging output in hyperterm. I had been
meaning, for some projects at work, to get a little bit of
practice with DirectX anyway, so I spent some late night hobby
hours working on a host interface for Bing development.
It's a nice side benefit if I can apply some of what I'm
learning at home to my job. Here is a screenshot of what
I have so far (and as usual, click on it for a larger
image). It looks more impressive than it is, right now
it is more of a mockup than anything and it does not
actually talk to the microcontroller.
In other news, I got a "MiniRoboMind", a
neat-seeming little board with a Motorola 68332 controller on
it that I wanted to use for image and sound data gathering and
processing. Unfortunately, after a couple hours of use
the board zonked out. I sent it in to have its
manufacturer look at it and see what's wrong with it.
Now I'm wondering if that is even the route I will take...
managing two different controller architectures might be more
work than it is worth. So once again I am still
undecided what to do about Bing's vision. But I don't
really need an answer at least until Bing's legs are done, as
I have no time to play with it now anyway.
A recommendation: earlier today I broke my last 1/8
inch end mill while attempting to make a part; since the local
shops don't carry much in this vein, I ordered a couple from Discount
Tools, from whom I have ordered some other stuff. An
hour later (on Sunday night!) I got a call from them to
clarify part of the order because they were ready to pack it
up. Cool!
Finally, I have mused a bit on ideas for Bing II -- but I
certainly don't want to fall into the pattern I see so often
in hobby robot projects: robots never really finished or
working right because they are abandoned for new projects.
Happy Thanksgiving everybody! |
