Straight line Robot navegation ( high accuracy navegation) 300 feet navigation.

markuster

Hi,

I want to use my BoeBot with my basic stamp and I need an
straight line navigation.

I don't know which is better to add :

A- Parallax GPS alternative.
B- Parallax HMC6352 Compass Module alternative.


Other alternatives I thought were:
-- Using Infrared sensors but I want 300 feet ( Parallax IR are about 20 feet)
-- Using QTI sensors could be good idea , but I don't want to draw a line.

What do you think is better?
I want just to turn on the BoeBot and go in a straight line.

Thanks, Mark

kf4ixm

In my opinion, i would go with the compass module, there would be too much error i think with gps to go as straight as possible.

erco

Wheel encoders for indoor, flat floor use. Neither your A or B choices will work as well as simple encoders·to maintain a straight line. You will need encoders to measure distance accurately anyway, so might as well get double duty and use them to keep your wheels synchronized.

Suggest you move this post to the robotics forum.

▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
·"If you build it, they will come."

Post Edited (erco) : 1/6/2010 7:22:28 PM GMT

SRLM

Another solution would be to shine a laser beam along the path that you want to take, and simply follow that. Like line following without the line.

▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
Powered by enthusiasm

Ken Gracey

Between the compass and GPS, I'd take the compass to travel "straight". Much easier to implement. Just keep the heading within a few degrees.

To know how far you've traveled, GPS + the Boe-Bot Encoder Kit could help approximate the distance.

300' is pretty far for a Boe-Bot. Depending on the surface (sounds like you are outside) it may hop around a bit.

Ken

markuster

Thanks,


Yes, it is an outside navigation.

I think the laser could be a good alternative too.

I could fix a laser 300 feet away from the boebot but how
could follow the laser , maybe at night could be a little easy
but daytime is difficult.

If the laser is 300 feet away from the boebot it is difficult to follow specially daytime.

I saw a video from parallax with LDR (light-sensitive resistance)
but my problem is that the laser emitter is 300 feet away. Please see:
http://www.parallax.com/Portals/0/Downloads/mm/video/boebot/lightfollower.mpg

Any help please ??

Thanks Mark.

$WMc%

markuster

If the laser is pointed at the Boe Bot on a level field, Then 3 IR tranys or detectors could be used. The center IR would need to be set-up on the laser,If the Boe went right then the right IR would see the laser and correct movement to the left and vise versa. Both servos would run FWD when the center IR was lit by the laser. A sun light hood would help for daylight use.

▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
The Truth is out there············___$WMc%___···························· BoogerWoods, FL. USA

erco

Outdoors in broad sunlight, IR detectors will be saturated, even a shaded hood will be of little use.

Thought about a gyro?

▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
·"If you build it, they will come."

$WMc%

markuster

Lasers and IR detectors will work in direct sun light. Take a look at laser speed detection. Like the ones the COPs have to catch you speeding.

What Your after is not easy. The use of more than one tracking method is often used.This makes error checking easier.

Don't forget Google

▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
The Truth is out there············___$WMc%___···························· BoogerWoods, FL. USA

Bill Chennault

Mark--

I've played a lot with lasers, since my other hobby is shooting. I believe you could do 300 feet in broad daylight. But first, you need to do a simple test. Go out at night to the site and see if you can spot something at 300 feet with the laser. If so, anchor the laser where it is and run over to whatever you spotted. If you can still see the laser on it, move it to the left and then to the right and watch the spot. (Gee. I hope the thing you spotted wasn't a TREE!)

As you are looking at the spot, look at its diameter. It is not going to be that little tiny dot that left the lens of the laser. It may be huge (for a laser). It may be too large for your Boe-Bot application. If the spot is not too large, then set up the laser as before in the dark, but easily and accurately aimed at the now ready Boe-Bot. When it becomes dayling crank her up and see if the sensitivity of your Boe-Bot sensors are sufficient to see the light.

If so, don't worry. Be happy.

If not, then Ken's GPS/Compass technique would be my fall back. If the traveling surface is really smooth, with no chance of wheel slippage·and no high speed is required (that might cause a wheel to bounce and spin), you COULD use wheel encoders. As long as your software counts and adjusts when the counts are different, it should work.

I have a tracked machine that works quite well like that. He's been out in the grass and snow and rocks, even.

--Bill

▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
You are what you write.

markuster

Thanks Bill and all the people here:

I think that laser is a good idea too.

I have attached a picture with my idea.

Perhaps I have to add more sensors, maybe ten
instead of three.

Do you think that this could work better than the encoder
or better than the Compass ?

Thanks, Mark

Bill Chennault

Mark--

The laser thing may not work at all at that distance due to beam spread. I don't have a clue as to how wide the beam will become at 300 feet. I only know that the spread is easily detectable at 300 feet. If it is "too" much, then your detectors may have to be spread over several feet. (Just a guess.)

Do a simple test with a laser pointer at night. You don't need anything fancy other than the side of someone's house about a hundred yards away. It would be great if you had a co-conspirator at the neighbor's house (I guess it could be YOUR house, but where's the fun in that?) to measure the beam width on the side of the house. Then, you will know for certain about the placement of the sensors on the front of your robot.

I sure would like to hear about the results of your test.

--Bill

▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
You are what you write.

Phil Pilgrim (PhiPi)

I'd be more inclined to have an XBee broadcasting at the target location, with a direction-sensitive Xbee receiver on the bot. The receiver assembly would be mounted on a servo tha sweeps from side to side. The receiver's output would be the RSSI pin, which measures signal strength. By keeping th strongest signal frontmost, you would be taking the most direct path to the target.

Anything optical that's attempted in daylight over that distance is going to be a problem.

-Phil