On Thu, March 15, 2007 7:57 pm, belinda thom said:
> Hi,
>
> I've read thru the Sensors module on the Wiki, which provides a lot
> of useful information. However, I'm still uncertain what the answer
> is to the question that originally led me there, so I thought I'd try
> the list.
>
> If I do something like:
>
> robot.range[0].distance()
>
> and then
>
> robot.range[0].value
>
> I often get very similar answers (even calling distance or value
> multiple times gives different answers, even if the robot is not
> moving, presumably b/c the simulation models noise).
>
> Is there a difference between these two? When would one use one
> versus the other?

Hi Belinda,

They are the same. We originally had the distance method for all similar
sensors, but then realized that it isn't always a distance---sometimes it
is just a value. For example, you wouldn't want to call the reading from
robot.light[0] a distance.

BTW, the noise can also be adjusted on simulated sensors. For example, if
you:

>>> robot.light[0]._noise
[0.05, 0.05]
>>> robot.light[0]._noise = [0, 0]

you can turn noise off for all sensors. You could also change the amount
of noise per sensor. (This is a very simple noise model, an area that
could be enhanced. But it is better than no noise.)

> Also, is the dead reckoning (e.g. robot.x, robot.y, robot.theta)
> considered "global" information, or is this "local enough" to the
> robot to be used by a layer in a subsumption architecture?

It should be a local estimation of global position. However, for the
Pyrobot simulator, it does not contain any noise, so it is also perfect
global information. At least 0,0 is the position that each robot begins
in. (Making these values less than perfect is another area where Pyro
could be enhanced. There are some comments in the
pyrobot/simulator/pysim.py (look for FIXME if interested).)

> As always, thanks,

You're welcome; hope this helps!

-Doug

> --b
>
> p.s. I don't have access to robots, so when I ask questions on this
> list, its always wrt simulated robots.


_______________________________________________
Pyro-users mailing list
[email protected]
http://emergent.brynmawr.edu/mailman/listinfo/pyro-users

Reply via email to