[From Adam Matic, 2014.12.11]
from Rick Marken (2014.12.11.1430)
The beacon following robot is a first test of beacon following control
systems and the nice new wheels.Two top control systems are perceiving
the angle between the head and the beacon, and the distance from the
head to the beacon (the beacon is an IR light, the sensor is an IR
receiver).
RM: How is distance computed?
AM: The IR sensor is outputing two numbers, one is the angle of
displacement (of the beacon from the center line of the sensor), and
the other
is the distance, which is, I assume, computed from the intensity of
the signal. Works up to about two or three meters.
RM How does the robot know that the beacon is no longer visible and that
therefore the head should stay where it was when the beacon was last seen? I
know this probe let sounds like a stupid question but I'm puzzled because
you've got two control systems that must be controlling variables related to
the sensed amplitude of the beacon. So when there is no beacon the sensors
would go to zero which would presumably produce a large error in at least
one of the control systems, which would act cause the robot to correct the
error by moving the body or turning the head. But apparently when the IR
sensors go to zero (because the beacon is not visible) the head stays where
it was before the beacon disappeared, with the body maintaining alignment.
I see what happens at the end of the video. The head remains in the position
it was in when the beacon disappeared and the body keeps turning to maintain
alignment with the body, so when you straighten the head the body aligns
with the head. But why does the pursuit stop? Is there some code in the
robot that says stop moving if the amplitude of the beacon at both sensors
is 0?
AM:
It is not at all a stupid question, I was as surprised and puzzled
when the robot continued turning around after the beacon was turned
off. But it didn't always turn around, sometimes it just stopped. It
took me a few tries to figure out that speed of turning is related to
the angle of head to the body. And I programmed the bugger! As you
say, if the head is turned, and there is no beacon, there is a large
error in the head-body alignment control system, and it is causing the
body to turn with the speed of turning in proportion to the error.
There is a piece of code that says "if the angle of head to the beacon
is > 0, turn the head to the left (with speed proportional to the
angle), if it is > 0, turn the head to the right, and if it is zero,
don't turn the head." That last part is what allows for manual
adjustment of the head when there is no beacon. That is also an
unplanned behavior (a feature, not a bug!). The head can't be turned
manually if it is focused on a beacon, the motor resists any attempts.
The distance reducing control system is an integrator - the error is
increasing slowly when there is a beacon and a distance from it, and
that causes the robot to move forward. When there is no beacon, the
error stops growing, but the robot still continues to move until the
error goes to zero.
>
RM: Super!! You have restored my faith (and confidence)in the future of PCT.
Actually, now I would like to get myself a logo set and replicate your
balancing robot and beacon follower. Could you tell me (or us, on CSGNet)
how to do that? I think it would be fun and educational!! And I want that
beacon follower robot for my 1 year old granddaughter! It works great and
it's cute as can be!
They are terribly fun! The set is called Lego Mindstorms EV3 (the one
the beacon follower is made from), they sell it on Amazon and Lego
stores. To make working easier, you'd probably need some additional
stuff - a rechargeable battery and a small wifi dongle (Rupert, am I
missing something?). The gyro sensor is not a part of the set, but
can be bought separately. To program it in Java, you can use NetBeans
or Eclipse and the lejos software (http://www.lejos.org/)
Best,
Adam
···
Best,
Adam
On Wed, Dec 10, 2014 at 7:45 PM, Richard Marken >> <csgnet@lists.illinois.edu> wrote:
> [From Rick Marken (2014.12.10.1045)]
>
>> Rupert Young (2014.12.10 17.00)--
>>
>> Here's a couple of videos of some great work by Adam on the prototypes
>> we
>> are putting together.
>>
>> https://www.youtube.com/watch?v=_CsOqQE_X1Y
>>
>> https://www.youtube.com/watch?v=jQTuE1SyY3E
>>
>
> RM: Well, that was refreshing!! Great work you guys.
>
> RM: I would really like to know more about these robots. Do you have
> write
> ups on this that you could post to CSGNet? I'm particularly interested
> in
> two things. One is the lifting and roaming behavior of the "Two arm
> robot".
> Is there some kind of remote control telling it when to get up and where
> to
> move once it's balancing on two wheels? Or is it completely autonomous?
> The
> other is what the beacon following robot does when there is no beacon
> present? Does it just go towards the brightest part of the room? Or does
> it
> stop? I'm interested because I've never really understood how a control
> system is taken "off line" and it looks like the beacon follower might
> only
> do its following when there is a beacon around.
>
> RM: Of course, the other thing the beacon following robot illustrates
> quite
> nicely is control behavior. The person moving the beacon (I presume it's
> Adam) can make the robot move around the floor in any pattern desired.
> Moving the beacon relative to the robot is a disturbance to the
> perception(s) it is controlling (which I presume are the relative and
> total
> amount of light falling on two photo receptors) which the robot
> compensates
> for by turning and moving toward the beacon.
>
> RM: Again, great work guys. I'd love to see diagrams of the control
> systems
> involved in these robots! If you could make them diagrams that are
> understandable by electronics dummy's like me that would be really nice.
>
> Best
>
> Rick
> --
> Richard S. Marken, Ph.D.
> Author of Doing Research on Purpose.
> Now available from Amazon or Barnes & Noble
>
> In nature there's no blemish but the mind
> None can be called deformed but the unkind.
> Shakespeare, Twelfth Night
--
Richard S. Marken, Ph.D.
Author of Doing Research on Purpose.
Now available from Amazon or Barnes & Noble
In nature there's no blemish but the mind
None can be called deformed but the unkind.
Shakespeare, Twelfth Night