Tweet
@Paul23: I have 4 driven motors with wheels, and 2 non-driven wheels with encoders. This should allow pretty decent dead reckoning. I want to try two things. 1st method: I will get a very accurate initial reading at base control. Base control has GPS not moving. Long averaging. As I move the rover, I'm updating my positioning via dead reckoning. 2nd method: I will pause the rover for every 24 scans of the boom, updating position. Once scans are done I move the rover again via dead reckoning to next spot (12" later), process starts over. I want to see how these two methods compare.
@Tepco: I could use this for mine detection. That's not what I'm shooting for though. I am building this platform to test both my robotics experiments, and my MD hobby. This rover is semi-automated. Not fully automated. Semi automated makes remote control easier, but still leaves major control in my hands. Sweeping will be automated. Rover movement except for turns will be automated. A turn can even be semi automated. My perspective from a hundred feet away will be skewed. The rover will turn the precise amount to ready itself for another pass. I will simply hit a single button, rather than steering manually. Same thing goes for movement. I will have a button for "shift +" or "shift -". This will turn the rover 45deg move forward a bit, then turn rover back 45 deg. All with a single button click. This is semi automated. You DONT want fully automated. Too many variables to control, and it won't work, since over time those unknown variables will stack until it's completely out of whack. Keep in mind, I will have a map on my operator station of the rover's movement track. I'll see my scans and the boom sweep. The boom sweep area will be color coded one color, and the rover movement track will be another.
Of course, if I want to steer or turn manually, I will be able to do so if desired.
Hopefully, the MD portion, via the small particle gold detecting I'm planning on doing, will pay for the project. It's not necessary though. Just doing this to have fun. I've been building machines like this for years at my regular job, and I want to do something on my own with my own budget. The sky's the limit with a $500K+ budget. Last system I built, for around $1M saved a customer about $30M. This technology is very marketable in certain industries.
@Midas: That's what I'm thinking GPS method #2 will get me.
@Davor: You know that's actually not a bad idea. The Phidget GPS uses an external antenna, I may be able to exchange for one better suited.
Here is my take on GPS. It doesn't have to be perfect. The repeatability I will have with GPS couple with dead reckoning off of non-slipping wheels is the main thing. It will allow me to see big patterns in a scan field not possible with single point measuring. As long as the targets are referenced off one point, that's my main goal. That will allow me to see the invisible dry wash with the gold targets along its path. Stuff just not possible with hand digging.
My MD coil will also make a big difference in my opinion. By doing little bitty scans and not missing a square centimeter, I bet you I'll be able to find stuff that is missed. I bet I could even go to a known gold location and find stuff that has been missed a hundred times.
As to my part receipt progress. I received in my last 3 motor amplifiers for the main wheels. I also received my rotation axis with motor. I'm still waiting on my 24V power supplies from Hong Kong, and my 1.85KW generator to come in.
Started laying out the frame this weekend. I built a raised work bench approx 4ft x 8ft and 30" high for laying out the frame.
I made good progress on the frame. I have the two sections cleaned up quite a bit, and have the connection bar ready to couple on both sides.
Attached are pics of the frame and work bench.
@Tepco: I could use this for mine detection. That's not what I'm shooting for though. I am building this platform to test both my robotics experiments, and my MD hobby. This rover is semi-automated. Not fully automated. Semi automated makes remote control easier, but still leaves major control in my hands. Sweeping will be automated. Rover movement except for turns will be automated. A turn can even be semi automated. My perspective from a hundred feet away will be skewed. The rover will turn the precise amount to ready itself for another pass. I will simply hit a single button, rather than steering manually. Same thing goes for movement. I will have a button for "shift +" or "shift -". This will turn the rover 45deg move forward a bit, then turn rover back 45 deg. All with a single button click. This is semi automated. You DONT want fully automated. Too many variables to control, and it won't work, since over time those unknown variables will stack until it's completely out of whack. Keep in mind, I will have a map on my operator station of the rover's movement track. I'll see my scans and the boom sweep. The boom sweep area will be color coded one color, and the rover movement track will be another.
Of course, if I want to steer or turn manually, I will be able to do so if desired.
Hopefully, the MD portion, via the small particle gold detecting I'm planning on doing, will pay for the project. It's not necessary though. Just doing this to have fun. I've been building machines like this for years at my regular job, and I want to do something on my own with my own budget. The sky's the limit with a $500K+ budget. Last system I built, for around $1M saved a customer about $30M. This technology is very marketable in certain industries.
@Midas: That's what I'm thinking GPS method #2 will get me.
@Davor: You know that's actually not a bad idea. The Phidget GPS uses an external antenna, I may be able to exchange for one better suited.
Here is my take on GPS. It doesn't have to be perfect. The repeatability I will have with GPS couple with dead reckoning off of non-slipping wheels is the main thing. It will allow me to see big patterns in a scan field not possible with single point measuring. As long as the targets are referenced off one point, that's my main goal. That will allow me to see the invisible dry wash with the gold targets along its path. Stuff just not possible with hand digging.
My MD coil will also make a big difference in my opinion. By doing little bitty scans and not missing a square centimeter, I bet you I'll be able to find stuff that is missed. I bet I could even go to a known gold location and find stuff that has been missed a hundred times.
As to my part receipt progress. I received in my last 3 motor amplifiers for the main wheels. I also received my rotation axis with motor. I'm still waiting on my 24V power supplies from Hong Kong, and my 1.85KW generator to come in.
Started laying out the frame this weekend. I built a raised work bench approx 4ft x 8ft and 30" high for laying out the frame.
I made good progress on the frame. I have the two sections cleaned up quite a bit, and have the connection bar ready to couple on both sides.
Attached are pics of the frame and work bench.
Comment