Help with Slave Robot

[fate]

thanks a lot guys.. another alternative we're testing right now is using the same IR transceivers for direction to know the distance of the user. Is it possible that we take out the received IR signal and assume distance through its RSSI?? i mean we won't need really precise distance from the user, all we need is at least to have a 0.7-1meter allowance for the user to freely move.  Is it possible to use the RSSI even on the moment when the robot is properly oriented at the user's back as distance detection ?? super thanks.. im also one of the pursuers of this project.. i guess we'll be bugging you guys for the whole year .. ) 

[Soeren]

RSSI?

[necxz]

Its seems it will be really hard to implement too many sensors. I have read many studies and most of you are suggesting that it is really hard. What if we remove the range finder instead. From what I've read if we only embed an anti-collision sensor(ultrasonic) and an IR beacon they will work without interfering with each other and thus solving our problem. Am i correct?  

[fate]

Received signal strength indication - we can determine distance of an object through the strength of the return signal.. like for our case, IR.. the attenuation of signal by the distance it traveled assuming no other interference.. this was the suggestion of our adviser. wherein the returned modulated frequency will be used to assume distance.. that the robot will stop if the calculated distance(considering noise here) is like for example less than 1m.

[Soeren]

Hi,

Received signal strength indication -

Aha, it helps writing it out in full the firsst time when usin self invented acronyms.

we can determine distance of an object through the strength of the return signal.. like for our case, IR.. the attenuation of signal by the distance it traveled assuming no other interference.. this was the suggestion of our adviser. wherein the returned modulated frequency will be used to assume distance.. that the robot will stop if the calculated distance(considering noise here) is like for example less than 1m.

That will be very variable depending on the target. If you get some useable results wearing blue jeans, don't wear white or black when you have to demonstrate it, or you'll be disappointed with your grades.

[madsci1016]

Aha, it helps writing it out in full the firsst time when usin self invented acronyms.

http://en.wikipedia.org/wiki/Received_signal_strength_indication

And, since you will no doubt mock wikipedia

http://www.google.com/search?q=RSSI
http://acronyms.thefreedictionary.com/RSSI



So... not so self invented. Either that or 'fate' help write the indicator panels on all the radios with RSSI indicators I have ever seen. 


I think you guys are way over complicating this problem. You had a good plan with using the IR beacon for bearing, a ultrasonic for following distance and two side ultrasonics for avoidance, and some good programming to determine if it's an obstacle or target in front of you.

Heck, you can even just have bumpers that would feel for obstacles and try to go around.

[Soeren]

So... not so self invented. Either that or 'fate' help write the indicator panels on all the radios with RSSI indicators I have ever seen. 

Apparently not, just obscure enough that it ought to have been spelled out, so as not to waste peoples time Gogling stuff.
Glad you enjoyed your time spend searching and your "field day" 

[madsci1016]

Apparently not, just obscure enough that it ought to have been spelled out, so as not to waste peoples time Gogling stuff.
Glad you enjoyed your time spend searching and your "field day"  

Miss my chance to poke fun on our resident dinosaur?

It's not obscure, in fact it's written on the two radios that happen to be on my bench right now. It would have been nice not to have to spend 30 seconds searching, but since you didn't bother someone had too. I'll be sure to spell out all basic engineering acronyms in my posts for you from now on.  

(This is in jest, fyi)

[fate]

hey madsci and soeren.. pls dont make this such a big fuss.. i'm really glad your both helping us.. it really means a lot if you must know.. :)we're just discovering other options to make our project worth defending during our paper presentation and you both did really helped a lot. thanks.. neczx and i are going crazy right now, i do get both of your points though.. that is not to make our project complex when it could have been easy.. (tap on my head)

we've decided to include in our limitations that the robot wouldnt be able to detect if its the user or an obstacle is up in front. it will avoid whatever it is as long as it gets near the robot when it is moving forward.   meaning it is not the user. then we could include a park mode in the user's transceiver so that he/she can get near the vehicle. did we decide right?? thanks guys, we are planning to include you both and this site to our acknowledgements..

[madsci1016]

Don't worry fate, Soeren is a wise old man and I had to get my jab in when I could, as the opportunities come few and far between.

the robot wouldnt be able to detect if its the user or an obstacle is up in front

I wouldn't say that is such definite terms, because like i said before, you could get fancy with your code such that if there's an object in front and the IR beacon is being received, object = user. If there's an object in front and beacon isn't being received, object = obstacle.

[fate]

Don't worry fate, Soeren is a wise old man and I had to get my jab in when I could, as the opportunities come few and far between.

 

the robot wouldnt be able to detect if its the user or an obstacle is up in front

I wouldn't say that is such definite terms, because like i said before, you could get fancy with your code such that if there's an object in front and the IR beacon is being received, object = user. If there's an object in front and beacon isn't being received, object = obstacle.

thanks.. got that..

[knossos]

What do you mean either sides?

The red is the IR beacons and path, the blue is a forward looking ultrasonic that is used to maintain distance, and the yellow are side looking ultrasonics to avoid hitting someone.

what if b/w the robot and the user suddenly a person blocks his path?

There is no simple answer to this problem. Any range sensor won't be able to make out a obstacle versus the person it is suppose to be following.

The only thing I can think of, is if that IR beacon tells you it lost the other beacon completely. If the IR beacon lost signal, and there's something close in front of the robot, you can assume it's an obstacle and not the person to follow.

Instead of using IR Rangefinders, you could use an IR emitter on the master (the person who the robot should follow) and a IR camera on the slave.  The Nintendo Wii Remote is a bluetooth enabled IR camera.  Only drawback is you would need to interface through a computer, but its an option that you could try readily.  See this example of someone using the Wiimote to make a multitouch whiteboard:

Low-Cost Multi-touch Whiteboard using the Wiimote

[madsci1016]

Instead of using IR Rangefinders, you could use an IR emitter on the master (the person who the robot should follow) and a IR camera on the slave.  

You are comparing apple to oranges here. First, they are using an IR beacon for bearing, not an IR rangefinder. Second, the Wii camera will be ok for telling you the bearing to the IR, but will be tricky to get distance. It would require two IR sources on the person, and a computer algorithm to measure the distance between the two. Even then, it may not be as accurate as an IR rangefinder.

The Wii remote is good for many things, but way too impractical for this.

[fate]

Instead of using IR Rangefinders, you could use an IR emitter on the master (the person who the robot should follow) and a IR camera on the slave.  

actually, we have thought about wiimote, but had given up the idea. its way too complex for our bearing where we could could use a simpler device, quite impractical.

[fate]

are IR really used in line-of-sight communications only? that blocking the signal could mean a signal loss? it's quite a puzzle to me where in this video, obstacles are present (such as the sofa) and still the beacons were able to detect each other.  I mean did i miss something here?

Beacon Robot

[madsci1016]

are IR really used in line-of-sight communications only? that blocking the signal could mean a signal loss? it's quite a puzzle to me where in this video, obstacles are present (such as the sofa) and still the beacons were able to detect each other.  I mean did i miss something here?

Beacon Robot

It is mostly line of sight, and won't go through a couch. I'd have to know more about that robot to answer your question. I have a feeling it just avoid obstacles until it gets a signal from the beacon.

[knossos]

You are comparing apple to oranges here. First, they are using an IR beacon for bearing, not an IR rangefinder.

Sorry it was late when I made the post, and I completely misread that.

And yes the wii remote would be impractical, I meant it more as an informational/proof of concept idea.