Subject:
|
RE: articulation points?
|
Newsgroups:
|
lugnet.robotics
|
Date:
|
Wed, 20 Nov 2002 23:11:13 GMT
|
Original-From:
|
Rob Limbaugh <rlimbaugh@^AvoidSpam^greenfieldgroup.com>
|
Viewed:
|
739 times
|
| |
|
|
> I think you've hit the nail on the head here Eric. To me, the
> most logical
> way of looking at a robot arm would be from the point of view
> of the end
> effector, i.e. what it can and cannot do. The way the arm
> moves to get there
> may be of less concern, or it may not in some cases, I suppose.
That, to me (and according to all definitions I've seen) is "range of motion".
However, without DOF's, an effector couldn't be positioned.
> Stating n DOF (for large n) seems to be somehow cheating,
> marketing speak, a
> bit like 6 head video recorders! You could have a robot arm
> where the end
> effector only has 4 DOF, yet according to some measures the
> arm has 6 or
> more DOF. To state it is a 6 DOF machine seems misleading, so
> what then is
> the true measure?
The more DOF's, the better the overall control and positioning *could* be. In
the case of the video recorders, you can tell the differences quite easily in 2,
4 and 6 head systems by pausing videos; or displaying the video during
frame-advance, fast-forward, or rewind; or recording something on a tape in LP
or SLP mode. I've seen 2-head VCR's that had better video than cheap 4-head
VCRs, but that is due to overall implementation of the system.
Mechanically, the contrast could be made by putting your leg in a cast. You can
still function and move around within your fully extended ranges of motion, but
doing things requiring better control will be harder... specifically, going up
and down a staircase... knee and ankle flexibility is greatly missed! In this
example, your maximum extended ROM hasn't changed, but your degree of resolution
to move your foot within that range has been greatly reduced.
So, in my opinion, they should list maximum/minimum range of effector motion and
maximum/minimum ranges of the control arms an effector is attached to
(clearance?) along with the number of "joints" (linear or otherwise) in the
system.
> How many DOF does a normal tracked vehicle have? What about a Killough
> platform?
I would think in both cases that the platform (as a whole) has 1DOF (z-axis).
On the tracked vehicle, each drive sprocket has 1DOF that encompasses 360
degrees. For the Killough, each of the three wheel-sets has 3DOF (acts like a
ball), each for 360 degrees. However, in both cases, assuming no obstacles, the
range of motion for the platforms is essentially infinite in a plane except for
the space the platform itself occupies.
Building on that idea, a bipedal walker would have a similar overall total range
of motion for the platform, but how well it can be positioned within that plane
will be determined by how well it can step within it's own total footprint. In
which case, it may have more total joints (more DOF's), but be harder to
position and control (such as pointing the bot in specific directions) at any
given time.
Just like the number of heads in a VCR, having all those DOF's doesn't
necessarily mean you've gained anything useful to accomplish the task the system
is designed for.
- Rob
|
|
Message has 1 Reply:
 | | Re: articulation points?
|
| (...) I was under the impression that a 6 head video recorder was a marketing concept, i.e. it has 4 video heads and 2 audio heads, so in this case (all things being equal) it would be exactly the same as the 4 head one. This is really what I've (...) (22 years ago, 21-Nov-02, to lugnet.robotics)
|
2 Messages in This Thread:
 
- Entire Thread on One Page:
- Nested:
All | Brief | Compact | Dots
Linear:
All | Brief | Compact
|
|
|
|
