Robonova avoiding obstacles

[ozfiddler]

My son and I have decided to each make a robot that can avoid obstacles. We are going to set up a course in the (removed by spam filter) and see which one can get through the fastest.
I'm working on modifying my Robonova and have now fitted him with a distance sensor in a moving head. This is the results so far:

phpBB [media]

Unfortunately he has no sense of direction! I'm going to have to get a compass module so that he makes the correct choice when turning.
Otherwise he might just end up back at the starting line.

There's a few more details here: http://simplycomputing.com.au/archives/164

[PedroR]

Hi ozfiddler

Your idea sounds ver interesting and we're lookiong forward to see the results of your "competition".

I do whish to make a note though about using a compass: please bear in mind that the servos create a lot of magnetic fields. The feedback I've received from people here on the forum who've tried to use one is that you don't really get very accurate results because of these magnetic fields.

Positioning the sensor in different places (especially away from electronics) seems to help. (you may also want to add an LCD for debugging purposes btw if you don't have one yet as it may help you debug the readings of the compass as the robot moves.)

Regards
Pedro

[ozfiddler]

Ah... thanks Pedro. I was wondering about that. Might have to try and come up with an alternative solution. Any ideas? We were thinking for my son's tracked robot that he could use 2 light sensors - one on either side. There' s a window down one side of the room so that might give enough light difference to use as a reference (and lots of scope for bad jokes... "Stay away from the dark side Luke.")

By the way, my order of half-price Robo extras arrived today thanks... very speedy, considering it was going half way around the world. Great stuff. Lots of new toys to play with!

Cheers.

[PedroR]

Hi ozfiddler

I'm glad to hear you got your order and you have new projects lined up with the new material!

As for alternatives it is indeed a matter of ingenuity.

I could suggest you grab a GPS micro mini (http://robosavvy.com/store/gps) although it may be a bit too much for the job. And since you'll be indoors, the signal will probably be bad.

With regards to the light sensor that is probably a very good idea.

Another alternative (which is used in some (expensive) commercial systems) is placing some markers on the ceiling and having some sensors pointing to them to do triangulation.

Maybe you could combine the light sensors with light emitters (or use some IR strategically placed emitters instead)

Another idea would be using a colour sensor. We have a few in stock from inex and they are analog which makes them easy to interface with.
You could for example mar different areas of the maze with different colours so using the sensor you could know in which zone you are.

There is a scanned datasheet for this device here http://docs.google.com/viewer?a=v&q=cac ... aqiA&pli=1

and we have them in store here http://robosavvy.com/store/product_info ... cts_id/371

They're on a discount right now and they are compatible with Robonova as they are 0-5V sensors.

Anyway we'll be looking forward to hearing more from your project!

Best Regards
Pedro.

[ozfiddler]

Thanks for the ideas Pedro... I can see I'm going to have to get creative about this. I'll keep you posted on the progress.
Cheers, Alan

[ozfiddler]

Wow... I had no idea this would be so difficult. We spent all day on it yesterday and got nowhere. It wasn't even that hard a course - just had to get through the door at the end of the room:



The light sensor idea didn't work and trying to get some sort of alternating left turn/right turn rule going was also not any use.

Hmmm. I think the compass may be the only option. Has anyone tried a compass module with Robo with any success? Are some brands less prone to servo magnetic interference than others?

[PedroR]

Hi Ozfiddler

I was trying to find some information from previous projects with magnetometers and Robonova on our forum.

Compasses (or "magnetometers") have already been done with Robonova here on the forum so it's matter of doing a search for "compass" here http://robosavvy.com/forum/search.php

The best result I found is this http://robosavvy.com/forum/viewtopic.ph ... sc&start=0

It is a bit of a long discussion but you may find a lot of useful information here about using magnetometers (they're called compasses in previous projects but they're essentially the same).

Regards
Pedro.

[ozfiddler]

Thanks Pedro - the CMPS03 Devantech was the one I was thinking about and they seem to have got it working OK without too much interference. I suppose I can also make sure Robo is not moving at all when the reading is taken. If necessary I can mount the unit up on a long aerial! Right... that's the next step then.
Cheers,
Alan