Searching \ for '[PIC]: line follower' in subject line. ()
Make payments with PayPal - it's fast, free and secure! Help us get a faster server
FAQ page: www.piclist.com/techref/microchip/devices.htm?key=pic
Search entire site for: 'line follower'.

Exact match. Not showing close matches.
PICList Thread
'[PIC]: line follower'
2001\10\26@011724 by Tony Nixon

flavicon
picon face
Hi all,

I got lucky with another "cream" project.

Does anyone have info, or links to info, on real vehicles following
lines or tracking to a target point? The distance is going to be about
600 meters at speeds upto 110KPH.

I'm not sure how to attempt it because the operator is supposed to
simply "press a button" and different vehicle types may be used, each
having different operating characteristics.

I hope it's within the PIC's capabillities.

--
Best regards

Tony

mICros
http://www.bubblesoftonline.com
spam_OUTsalesTakeThisOuTspambubblesoftonline.com

--
http://www.piclist.com hint: The PICList is archived three different
ways.  See http://www.piclist.com/#archives for details.


2001\10\26@120223 by John Ferrell

flavicon
face
Automated warehouse systems and office mail delivery carts have been using
this scheme for years, but not at that speed!

Velocity & turn radius specs will make it interesting...

John Ferrell
6241 Phillippi Rd
Julian NC 27283
Phone: (336)685-9606
Dixie Competition Products
NSRCA 479 AMA 4190  W8CCW
"My Competition is Not My Enemy"



{Original Message removed}

2001\10\27@183604 by M. Adam Davis

flavicon
face
You might consider looking at
1) Car Automation - many universities and other entities are working on
such systems
2) Rockets - I've seen rocket articles where hobbyists outfit their
rockets with some basic guidance

Both ought to give you an idea of the controls and algorithms you'll
have to study to keep your vehicle on course.

Is this an extension of the project where you remotely drove cars into
telephone poles and such (or am I confusing you with another piclister?)

There were a series of messages about one piclister in particular who
was trying to get his rocket to track a balloon.

-Adam

Tony Nixon wrote:

>Hi all,
>
>I got lucky with another "cream" project.
>
>Does anyone have info, or links to info, on real vehicles following
>lines or tracking to a target point? The distance is going to be about
>600 meters at speeds upto 110KPH.
>
>I'm not sure how to attempt it because the operator is supposed to
>simply "press a button" and different vehicle types may be used, each
>having different operating characteristics.
>
>I hope it's within the PIC's capabillities.
>
>--
>Best regards
>
>Tony
>
>mICros
>http://www.bubblesoftonline.com
>.....salesKILLspamspam@spam@bubblesoftonline.com
>
>--
>http://www.piclist.com hint: The PICList is archived three different
>ways.  See http://www.piclist.com/#archives for details.
>
>
>
>
>

--
http://www.piclist.com#nomail Going offline? Don't AutoReply us!
email listservspamKILLspammitvma.mit.edu with SET PICList DIGEST in the body


2001\10\28@145940 by Brandon Fosdick

flavicon
face
Tony Nixon wrote:
>I got lucky with another "cream" project.
>
>Does anyone have info, or links to info, on real vehicles following
>lines or tracking to a target point? The distance is going to be about
>600 meters at speeds upto 110KPH.
>
>I'm not sure how to attempt it because the operator is supposed to
>simply "press a button" and different vehicle types may be used, each
>having different operating characteristics.
>
>I hope it's within the PIC's capabillities.

Some quick numbers:

110KPH = 30.56 m/s

20MHz PIC = 4 MIPS = 2.5e-7 sec/instruction

30.56 m/s * 2.5e-7 sec/instr = 7.64e-6 m/instr

Assume 100 instructions per controller cycle:
7.64e-6 m/instr * 100 instr/cycle = 0.764 mm/cycle

So, as a rough guess, you could assume that a decent control law would
generate tracking errors on the order of millimeters or centimeters. If
that's sufficient then a PIC should be able to handle it, depending on
how long your control loop is.

--
http://www.piclist.com hint: To leave the PICList
.....piclist-unsubscribe-requestKILLspamspam.....mitvma.mit.edu


2001\10\28@153353 by Josh Koffman

flavicon
face
What kinds of lines are they? Are they the painted road lines, or can
you spec the lines? If you can spec them yourself, it shouldn't be too
hard, OWI (I think) has had a robot that traces lines for a long time.
However if you must adapt to other lines, I don't know. Plus, my concern
would be how to control the degree of steering if the line doesn't stay
straight. You can't just have the controller say, the line is to the
left, steer hard left. perhaps if you had a few sensors on either side
of the line so you can have a percentage of how far off you are, and the
controller could act appropriately. however, what if you drift off to
the left, and then the line turns to the right...how would you deal with
total loss of the line?

Just some thoughts...I'm really interested in how you solve this.

Josh Koffman


{Quote hidden}

--
http://www.piclist.com hint: To leave the PICList
EraseMEpiclist-unsubscribe-requestspam_OUTspamTakeThisOuTmitvma.mit.edu


2001\10\28@171322 by Tony Nixon

flavicon
picon face
"M. Adam Davis" wrote:

> Is this an extension of the project where you remotely drove cars into
> telephone poles and such (or am I confusing you with another piclister?)

Yes. That system requires manual input. It works well, but they want
full automation on this one. The two things it must do is steer to a
target point and be at (X)kph when it gets there. The big problem is
that the system will never know the vehicle type until it 'drives' it.

--
Best regards

Tony

mICros
http://www.bubblesoftonline.com
salesspamspam_OUTbubblesoftonline.com

--
http://www.piclist.com hint: To leave the PICList
@spam@piclist-unsubscribe-requestKILLspamspammitvma.mit.edu


2001\10\28@171336 by Tony Nixon

flavicon
picon face
Josh Koffman wrote:
>
> What kinds of lines are they? Are they the painted road lines, or can
> you spec the lines? If you can spec them yourself, it shouldn't be too
> hard, OWI (I think) has had a robot that traces lines for a long time.
> However if you must adapt to other lines, I don't know. Plus, my concern
> would be how to control the degree of steering if the line doesn't stay
> straight. You can't just have the controller say, the line is to the
> left, steer hard left. perhaps if you had a few sensors on either side
> of the line so you can have a percentage of how far off you are, and the
> controller could act appropriately. however, what if you drift off to
> the left, and then the line turns to the right...how would you deal with
> total loss of the line?
>
> Just some thoughts...I'm really interested in how you solve this.

Me too :-)

I was thinking video camera to 'see' the line.

--
Best regards

Tony

mICros
http://www.bubblesoftonline.com
KILLspamsalesKILLspamspambubblesoftonline.com

--
http://www.piclist.com hint: To leave the PICList
RemoveMEpiclist-unsubscribe-requestTakeThisOuTspammitvma.mit.edu


2001\10\28@191642 by Josh Koffman

flavicon
face
How will you steer the vehicle and operate the brake/accelerator?

Josh

Tony Nixon wrote:
{Quote hidden}

--
http://www.piclist.com hint: To leave the PICList
RemoveMEpiclist-unsubscribe-requestspamTakeThisOuTmitvma.mit.edu


2001\10\28@193204 by Tony Nixon

flavicon
picon face
Josh Koffman wrote:
>
> How will you steer the vehicle and operate the brake/accelerator?

The mechanics is already completed, just the software to do.

--
Best regards

Tony

mICros
http://www.bubblesoftonline.com
salesEraseMEspam.....bubblesoftonline.com

--
http://www.piclist.com hint: To leave the PICList
EraseMEpiclist-unsubscribe-requestspammitvma.mit.edu


2001\10\28@221036 by Josh Koffman

flavicon
face
So is it that you know what vehicle is to be driven, but the PIC
doesn't?

Josh

Tony Nixon wrote:
{Quote hidden}

--
http://www.piclist.com hint: To leave the PICList
RemoveMEpiclist-unsubscribe-requestTakeThisOuTspamspammitvma.mit.edu


2001\10\28@223224 by Tony Nixon

flavicon
picon face
Josh Koffman wrote:
>
> So is it that you know what vehicle is to be driven, but the PIC
> doesn't?
>
> Josh

At the moment, we set the vehilce up, and do some simple calibration
runs. This info tells the PIC all it needs to know to control the
velocity. I steer the vehicle via a wireless data/video link and a
virtual headset using a dummy steering wheel. The PIC also controls the
steer mechanism via my input.

I never know what vehicle will be used until it gets delivered, but with
the above system, it doesn't matter much. With a totally automated
system, (ie. just push a button) it will matter because the system won't
know anything about the vehicle's driving characteristics.

I'm wondering, and this is probably far fetched, that given some
'general' characteristics that apply to all vehicles, and from data I
have already gained from previous use, that the system can learn and
compensate en-route. It doesn't have to go around corners, just follow a
straight line.

Testing this in the real world is going to be fun. The PIC has already
scared the pants of a "co-driver" colleague :-)

--
Best regards

Tony

mICros
http://www.bubblesoftonline.com
EraseMEsalesspamspamspamBeGonebubblesoftonline.com

--
http://www.piclist.com hint: To leave the PICList
RemoveMEpiclist-unsubscribe-requestKILLspamspammitvma.mit.edu


2001\10\28@224852 by Jeff DeMaagd

flavicon
face
----- Original Message -----
From: Tony Nixon <Tony.NixonSTOPspamspamspam_OUTENG.MONASH.EDU.AU>

> > Just some thoughts...I'm really interested in how you solve this.
>
> Me too :-)
>
> I was thinking video camera to 'see' the line.

I would think that processing video data this quickly is generally beyond
the capabilities of a microcontroller, particularly in rejecting false
markings and all that.  Maybe if you had a PIC dedicated to just watching
some sort of very low res / high refresh rate camera communicating with
another PIC you _might_ be able to do it safely.

The speed involved seems to be radically swift, I sure hope humans aren't
allowed near these things in operation.

Jeff

--
http://www.piclist.com hint: To leave the PICList
spamBeGonepiclist-unsubscribe-requestSTOPspamspamEraseMEmitvma.mit.edu


2001\10\28@232459 by Tony Nixon

flavicon
picon face
Jeff DeMaagd wrote:

> > I was thinking video camera to 'see' the line.
>
> I would think that processing video data this quickly is generally beyond
> the capabilities of a microcontroller, particularly in rejecting false
> markings and all that.  Maybe if you had a PIC dedicated to just watching
> some sort of very low res / high refresh rate camera communicating with
> another PIC you _might_ be able to do it safely.
>

What I had in mind was to use the contrast of the road surface and a
bright colored line, and perhaps in it's simplest form, may only need to
be a window comparator to monitor the voltage change on the video lines.

The line will appear to change angle from the cameras perspective, but
that may mean just monitoring a few frames.

If the line needs special paint, and the use of lighting that highlights
the line close to the vehicle, then so be it.

> The speed involved seems to be radically swift, I sure hope humans aren't
> allowed near these things in operation.

I had one guy wanting to stand in the line of trajectory once just so he
could get a better camera shot, and of course I wouldn't run it lest he
get a closeup photo of the bonnet emblem just before he got hit. It's
amazing how much trust people put in some things.


--
Best regards

Tony

mICros
http://www.bubblesoftonline.com
KILLspamsalesspamBeGonespambubblesoftonline.com

--
http://www.piclist.com hint: To leave the PICList
EraseMEpiclist-unsubscribe-requestspamEraseMEmitvma.mit.edu


2001\10\28@234552 by Gennette, Bruce

flavicon
face
Seems to be a bit of a novice camera operator to *NOT* know about using
mirrors for any dangerous shots.  (A totally destroyed $3 mirror is much
better than 3 months in hospital).

Bye.

{Original Message removed}

2001\10\29@035316 by Tom Handley

picon face
  Tony, since you can define your own line on a dedicated track, how
about using a simple array of emitter/detectors mounted on a bar that
can clamp underneath a vehicle. This would eliminate problems with
sunlight as well. A few detectors would be easy to manage compared to
video and contrast matching. It seems to me that you may need a DSP to
respond fast enough using that method.

  - Tom

At 15:23 29-10-01 +1100, Tony Nixon wrote:
{Quote hidden}

------------------------------------------------------------------------
Tom Handley
New Age Communications
Since '75 before "New Age" and no one around here is waiting for UFOs ;-)

--
http://www.piclist.com hint: PICList Posts must start with ONE topic:
[PIC]:,[SX]:,[AVR]: ->uP ONLY! [EE]:,[OT]: ->Other [BUY]:,[AD]: ->Ads


2001\10\29@045628 by Alan B. Pearce

face picon face
>I had one guy wanting to stand in the line of trajectory once just so he
>could get a better camera shot, and of course I wouldn't run it lest he
>get a closeup photo of the bonnet emblem just before he got hit. It's
>amazing how much trust people put in some things.

Sounds like you got yourself a nice position as a stunt organizer for a
movie company ;))

--
http://www.piclist.com hint: PICList Posts must start with ONE topic:
[PIC]:,[SX]:,[AVR]: ->uP ONLY! [EE]:,[OT]: ->Other [BUY]:,[AD]: ->Ads


2001\10\29@122811 by Peter L. Peres

picon face
> Yes. That system requires manual input. It works well, but they want
> full automation on this one. The two things it must do is steer to a
> target point and be at (X)kph when it gets there. The big problem is
> that the system will never know the vehicle type until it 'drives' it.

This sounds fairly dangerous. I.e. I think that it has other applications.
I hope that you considered this.

Peter

--
http://www.piclist.com hint: PICList Posts must start with ONE topic:
[PIC]:,[SX]:,[AVR]: ->uP ONLY! [EE]:,[OT]: ->Other [BUY]:,[AD]: ->Ads


2001\10\29@122815 by Peter L. Peres

picon face
The problem is not the error from the control algorythm, the problem is
the error in slip in the vehicle. You'd need much more than a center line
follower unit to track at 110mph imho.

Peter

--
http://www.piclist.com hint: PICList Posts must start with ONE topic:
[PIC]:,[SX]:,[AVR]: ->uP ONLY! [EE]:,[OT]: ->Other [BUY]:,[AD]: ->Ads


2001\10\29@164751 by Tony Nixon

flavicon
picon face
"Alan B. Pearce" wrote:
>
> Sounds like you got yourself a nice position as a stunt organizer for a
> movie company ;))

Some people from WB did come here once from a stunt team looking for
info. When told about the system I created, one of them dismissed it as
not being able to work, but after lots of runs, I think it's proven
itself.

--
Best regards

Tony

mICros
http://www.bubblesoftonline.com
@spam@sales@spam@spamspam_OUTbubblesoftonline.com

--
http://www.piclist.com hint: PICList Posts must start with ONE topic:
[PIC]:,[SX]:,[AVR]: ->uP ONLY! [EE]:,[OT]: ->Other [BUY]:,[AD]: ->Ads


2001\10\29@165142 by Tony Nixon

flavicon
picon face
"Peter L. Peres" wrote:
>
> > Yes. That system requires manual input. It works well, but they want
> > full automation on this one. The two things it must do is steer to a
> > target point and be at (X)kph when it gets there. The big problem is
> > that the system will never know the vehicle type until it 'drives' it.
>
> This sounds fairly dangerous. I.e. I think that it has other applications.
> I hope that you considered this.

If I'm on the same wavelength, a nasty thought. I hope it would be too
much hassle to set up, and unless you tracked a short straight line, the
video link would be useless unless you had clear overhead reception.

--
Best regards

Tony

mICros
http://www.bubblesoftonline.com
spamBeGonesalesspamKILLspambubblesoftonline.com

--
http://www.piclist.com hint: PICList Posts must start with ONE topic:
[PIC]:,[SX]:,[AVR]: ->uP ONLY! [EE]:,[OT]: ->Other [BUY]:,[AD]: ->Ads


2001\10\30@091555 by Douglas Butler

flavicon
face
How about creating your "line" with a pair of spinning lasers?  If they
spin in opposite directions and are synchronized by a timing belt or
simillar, a detector would see both lasers simultainously only on a line
passing between the lasers.  Left or right of that line it would see one
laser before the other.  You would have to modulate the lasers to tell
them apart.  I think a PIC would be able to interpret that signal.

Doing anything other than a straight line would require modulating the
spin rate of the lasers, though but possible.

Sherpa Doug

> {Original Message removed}

2001\10\31@125237 by Olin Lathrop

face picon face
> How about creating your "line" with a pair of spinning lasers?  If they
> spin in opposite directions and are synchronized by a timing belt or
> simillar, a detector would see both lasers simultainously only on a line
> passing between the lasers.  Left or right of that line it would see one
> laser before the other.  You would have to modulate the lasers to tell
> them apart.  I think a PIC would be able to interpret that signal.
>
> Doing anything other than a straight line would require modulating the
> spin rate of the lasers, though but possible.

If you're willing to do all that you could use the same principle as LORAN.
LORAN is a series of radio beacons intended for aircraft navigation.  Each
beacon has a rotating narrow angle constant transmitter and a pulse
transmitter.  The pulse fires once per cycle of the rotating transmitter at
a known angle.  You can determine your angle from the beacon by determining
when you receive the pulse from the rotating transmitter with respect to the
fixed pulses.  You can only tell where you are along a line from each
beacon, but you can tell position from multiple beacons.


********************************************************************
Olin Lathrop, embedded systems consultant in Littleton Massachusetts
(978) 742-9014, .....olinspam_OUTspamembedinc.com, http://www.embedinc.com

--
http://www.piclist.com hint: The PICList is archived three different
ways.  See http://www.piclist.com/#archives for details.


2001\10\31@125714 by Chris Carr

flavicon
face
Was this problem not resolved by the Romans culminating in the highest speed
on the 3rd July 1938 ?

Regards

Chris Carr

--
http://www.piclist.com hint: The PICList is archived three different
ways.  See http://www.piclist.com/#archives for details.



'[PIC]: line follower'
2001\11\01@192226 by M. Adam Davis
flavicon
face
If you chose to use video, and you knew the line was straight, then it
would be a simple process to extract the first and last lines of video,
and from that determine the heading of the vehicle as long as the line
doesn't fall off the sides of the video.

The positive for this is that you have a much finer array for such
little cost than putting sensors on the front bumper.

Use reflective road paint, and make sure the vehicles lights are on.  As
long as the test track isn't wet you'll have a very strong contrast,
using a B&W camera, to sense the line in both strips of video.  If you
don't have enough contrast, then flag an error and don't start the engine.

As far as vehicle dynamics, if I'm thinking through this clearly, you
only have two things you need to worry about, steering and speed
control.  Since there is no one in the vehicle, you can just slam on the
brakes, though that could be a third control to worry about as well.

I hope you will have a chance to test the system on each car once or
twice before doing the actual run.  If so, you can run a small algorithm
that finds how quickly the car accelerates and brakes, and how much
steering is required for a certian amount of correction.

I suspect the thing you fear is making an oscillating system, where the
vehicle starts to swerve back and forth, oversteering.

This can, of course, all be done dynamically, using some sort of PID
feedback system.  Am I overcorrecting?  The decrement the register that
tells me how much to correct for a slight veer, etc.

As far as speed control, that should be relatively straightforward.  If
it's not, get them to give you the code for their cruise control.  I was
surprised to find mine would increase/decrease my speed by exactly one
MPH for each click, so I imagine it's controlled by a micro.

-Adam

Tony Nixon wrote:

{Quote hidden}

--
http://www.piclist.com#nomail Going offline? Don't AutoReply us!
email TakeThisOuTlistservKILLspamspamspammitvma.mit.edu with SET PICList DIGEST in the body


More... (looser matching)
- Last day of these posts
- In 2001 , 2002 only
- Today
- New search...