LMR Diamond Mine Challenge
Duane Degn
Posts: 10,588
If any of you Parallax roboteers haven't seen the website, LetsMakeRobots.com, you really need to check it out.
Man, oh man, do they have robots.
I recently posted a couple of my robots on the site. I've looked through a bunch of their robots and there's a lot of really good stuff.
With my recent activity at LMR, I came across the "Diamond Mine" challenge.
The jest of the challenge, if I understand it correctly, is to have a robot drive a distance, dig for diamonds, inspect the diamonds to make sure they're not just rocks and bring the "diamonds" back to the starting position.
I realize my recent efforts using a compass might make undertaking this challenge not quite as a severe effort as it would have been previously.
It's a far cry from a figure 8 but I was hoping I could talk some of my fellow forum members into joining the fun.
I'm thinking of mounting my CrustCrawler AX-12 SmartArm on a compass guided robot. I'd probably use encoders to keep track of the distance and then have the robot arm randomly start scooping up dirt, rocks and hopefully diamonds.
I'd want the gripper on the arm to allow loose dirt to fall away and then bring the rock or diamond up to some sort of detector. I may use Phil's NTSC video capture, Hanno's video capture method or a CMUcam4 to look at the stones. I'm sure there are many low tech ways of determining if a stone is a rock or diamond. A simple photo resistor could be used to test for transparent diamonds vs opaque rocks.
The "diamonds" can be pretty much whatever you want them to be. I think this is a purposely introduced loophole to make the challenge easier to complete. (I think the "rocks" can also be of our own choosing though I don't know if RFID tagged rocks and diamonds count.)
I thought I'd document my efforts here, and hopefully encourage others to join in the challenge.
The deadline has been recently extended to the end of March.
Man, oh man, do they have robots.
I recently posted a couple of my robots on the site. I've looked through a bunch of their robots and there's a lot of really good stuff.
With my recent activity at LMR, I came across the "Diamond Mine" challenge.
The jest of the challenge, if I understand it correctly, is to have a robot drive a distance, dig for diamonds, inspect the diamonds to make sure they're not just rocks and bring the "diamonds" back to the starting position.
I realize my recent efforts using a compass might make undertaking this challenge not quite as a severe effort as it would have been previously.
It's a far cry from a figure 8 but I was hoping I could talk some of my fellow forum members into joining the fun.
I'm thinking of mounting my CrustCrawler AX-12 SmartArm on a compass guided robot. I'd probably use encoders to keep track of the distance and then have the robot arm randomly start scooping up dirt, rocks and hopefully diamonds.
I'd want the gripper on the arm to allow loose dirt to fall away and then bring the rock or diamond up to some sort of detector. I may use Phil's NTSC video capture, Hanno's video capture method or a CMUcam4 to look at the stones. I'm sure there are many low tech ways of determining if a stone is a rock or diamond. A simple photo resistor could be used to test for transparent diamonds vs opaque rocks.
The "diamonds" can be pretty much whatever you want them to be. I think this is a purposely introduced loophole to make the challenge easier to complete. (I think the "rocks" can also be of our own choosing though I don't know if RFID tagged rocks and diamonds count.)
I thought I'd document my efforts here, and hopefully encourage others to join in the challenge.
The deadline has been recently extended to the end of March.
Comments
Many diamonds will fluoresce under UV. But who's going to give you real diamonds to dig for?
Watch out for loopholes, especially the first time a contest is run. Someone will win on a technicality. Here's a rule on that site I would bend heavily:
"This challenge is not as hard as it first appears. Remember I said you can use your own objects to simulate the diamonds. That was an important HINT. You could use white gravel, black marbles, or whatever you want. If you use the same sort of sensor as in a line-follower, you can find the "diamonds" by sensing light versus dark."
Heck, I would choose neodymium magnets as diamonds that would jump onto anything metal that got anywhere close to them. I would "sense" them with a hall effect or even one of Emily's new reed switches, then scrape them off into a bin. Oh, and your magnetic compass would lead you right to them...
Too bad I'm travelling for most of March!
I'm thinking of having dark dirt, dark rock and white or clear "diamonds". I'm hoping to use your machine vision technique to identify white objects against a dark background.
I recently purchased a CMUcam4 which I may also try using to identify "diamonds".
My day job has recently been keeping me busy so I haven't had as much time to work on this project as I thought I'd have.
Boo Hoo for you. Now add twins and you have a glimpse into my world...