Subject:
|
Re: Following the Line Follower
|
Newsgroups:
|
lugnet.robotics
|
Date:
|
Fri, 27 Sep 2002 01:37:12 GMT
|
Viewed:
|
1186 times
|
| |
|
|
Steve Hassenplug wrote:
>
> Very interesting (and good) work.
Thanks. I couldn't have done it without those ubiquitous felt pens.
> It may be useful for you to post the program used, because (as you said) it
> can make a difference in the results.
The bots used fairly standard programs (written in ROBOLAB).
For the single light sensor bot: turn right when it sees bright; turn
left when it sees dark.
For the bot with 2 light sensors: according to the matrix
L sensor | R sensor | Action
Bright | Bright | Forward
Bright | Dark | Turn right
Dark | Bright | Turn left
Dark | Dark | T-junction, action is programmable
At a T-junction, the bot can be programmed to stop (end of line), change
direction or do some task such as picking up or dropping off cargo, etc.
> However, I have to disagree with part of the conclusion:
> "A line follower using a single light sensor will not be able to negotiate a
> sharp 90 degree bend in both directions (clockwise and anticlockwise). You
> need a line follower with 2 light sensors for that. "
>
> This problem can be solved by either changing the construction of the robot
> OR writing a slightly more complex program.
You're probably right. I was confining myself to only using the standard
programs that kids come up with.
> For example, on the robot you used, you had the one light sensor in front,
> and between the wheels. If you move the light sensor to one side (I believe
> the left side, for the program you used) so it is right in front of the
> wheel, (like on your 2-sensor robot) I think you'll find it can actually
> make it through the 90-degree turn in both directions.
>
> I can elaborate, if anyone is interested.
Ok, I'll just take off or simply disregard the right sensor on my 2
sensor bot. That should do it.
> Coincidently, this will shift the robot so it will drive more on-top of the
> line, instead of on the edge. But that's a side effect.
I'll give it a try. It would certainly be nice to have a bot with a
single light sensor negotiate a sharp right angle bend both ways. So
far, I've only seen it done by bots with two light sensors, using fairly
simple programming.
> There's a contest coming up in Seattle in a couple weeks, where this
> information could be very useful.
>
> http://www.workshop3d.com/rcx/index.htm
>
> However, if you have a really good robot design or program, please DON'T
> enter it. I'm hoping to win some more cool prizes.
Fortunately, I'm too far away from Seattle. Good luck!
--
C S Soh
CSSoh's Lego Pneumatics
http://www.geocities.com/cssoh1
... where air is power!
|
|
Message has 1 Reply:
 | | Re: Following the Line Follower
|
| (...) You can improve the single sensor program considerably by having it go straight when it sees an intermediate value, which is what it would see when it's right on the edge. So, you have two cutoffs, LOW and HIGH. If light value is less than (...) (22 years ago, 27-Sep-02, to lugnet.robotics)
|
Message is in Reply To:
| | Re: Following the Line Follower
|
| Very interesting (and good) work. It may be useful for you to post the program used, because (as you said) it can make a difference in the results. However, I have to disagree with part of the conclusion: "A line follower using a single light sensor (...) (22 years ago, 26-Sep-02, to lugnet.robotics)
|
5 Messages in This Thread:
 
- Entire Thread on One Page:
- Nested:
All | Brief | Compact | Dots
Linear:
All | Brief | Compact
This Message and its Replies on One Page:
- Nested:
All | Brief | Compact | Dots
Linear:
All | Brief | Compact
|
|
|
|
