Subject:
|
Re: Robot That Sorts Lego Pieces
|
Newsgroups:
|
lugnet.robotics.vc
|
Date:
|
Sun, 13 Apr 2003 23:16:02 GMT
|
Viewed:
|
6480 times
|
| |
|
|
In lugnet.robotics.vc, John Shearing writes:
> Hi Chris,
>
> It seems to me that for the first time in history, kids could have a robot
> that can do the thing they hate most. Putting their toys away. This could be
> a very effective way to show kids the importance of robotics.
>
> Clearly, you have given this a lot of thought. I am glad that someone else
> is interested in this problem. By the way, I really like your caliper device.
>
> There is a way to accomplish part identification without real recognition
> software. It would involve a stage for viewing the pieces that would have a
> white background. Then I could have the camera move its way around the part
> in effect measuring it in much the same way as a coordinate measuring
> machine. This combined with weighing the parts could determine their
> identities. But this process would be time consuming and so would not be
> practical. So now I am wondering about real object recognition software for
> use with the Vision Command system.
>
> If you or anyone else knows of any books that discuss object recognition I
> would like to have your recommendation.
>
> Also, if anyone out there is working in the field of vision recognition,
> please chime in.
>
> And lastly, do any of you think this is a good challenge for the First Lego
> League?
>
> Thanks, John
I don't want to discourage your efforts, but I view this as more of an
academic exercise than anything that is likely to result in a practical
brick sorting machine. There is so much variation in size and shape of the
pieces in a typical mound of LEGO that it will not be easy simply building
reliable, jam-proof mechanisms to move those pieces to where you want them,
let alone figuring out what each piece is in the first place.
As I mentioned in my other message, my caliper chamber is only one of the
three complex subsystems that would be required to make this work at all.
And unfortunately, it is probably the easiest one to build. (That's why I
built it first.)
As far as FLL goes, I think this would be well beyond the level of
difficulty that the average elementary school student could handle.
Although I wouldn't rule out the possibility that some child could surprise
us all, I wouldn't pose it as a challenge to FLL unless I was able to solve
it myself, or at least reasonably sure I could solve it in a finite amount
of time. I think the best challenges for that kind of setting are the ones
that are relatively simple, yet which leave plenty of room for creative
solutions. You want a large percentage of the kids to succeed at making a
working robot, and then the group gets to compare/contrast the various
approaches taken with individual entries.
You could always constrain the problem in ways that would make it somewhat
easier to solve. For example, only require the robot to sort regular brick,
and then only a few sizes like: 1x1, 1x2, 1x3, 1x4, 2x2, 2x3, 2x4. Even
that might be too many sizes. I think the challenge would be just as
valuable if you only allowed a few brick sizes, as any workable solution
could hopefully be extended. Extra credit for creative handling of
"misfits" that aren't on the list, ie: put them all into a separate reject bin.
As far as suggested reading for image processing, I haven't really studied
this in over ten years, and the field has advanced quite a bit in that time.
So any book I've read and could recommend would be sadly out-of-date. Start
with a few Internet searches on phrases like "image recognition" or "pattern
matching" and you should find a lot of free information. Or search Amazon
for titles in print.
Again, I don't mean to be a nay-sayer. I would really like to see someone
build a decent brick sorter, and I'm sure that it can be done. I just think
it might be a bit much to ask of a child. I also think that if you set out
to build this thinking that it will save you time in the long run, you're
probably setting yourself up for some disappointment.
- Chris.
|
|
Message has 1 Reply:
 | | Re: Robot That Sorts Lego Pieces
|
| Hi Chris, (...) That is why I am looking for a software solution. (...) This is very insightful. You are no doubt correct. If I were to present this challenge to children, I would only be setting them up for failure. A good teacher will carefully (...) (21 years ago, 15-Apr-03, to lugnet.robotics.vc)
|
Message is in Reply To:
| | Re: Robot That Sorts Lego Pieces
|
| Hi Chris, It seems to me that for the first time in history, kids could have a robot that can do the thing they hate most. Putting their toys away. This could be a very effective way to show kids the importance of robotics. Clearly, you have given (...) (21 years ago, 13-Apr-03, to lugnet.robotics.vc)
|
8 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
|
|
|
|
