Basically, what would you put on your “dream” robot (still keeping with the other robot restrictions, i.e., motors, etc.)?
On mine, I’d like to use a LOT more sensors. I’m talking pressure transducers in my pneumatics, thermocouples monitoring motors and their controllers, strain gages on my structures.
I would like to be able to disengage a motor completely, but also change the overall gear ratio to keep the same top speed (or acceleration, haven’t decided) sans a motor. That would mean I could drop down from 3 motors to 2 motors on the fly so if I’m drawing too much current, I won’t trip my breaker.
I’d like to fully characterize all the motors, batteries and everything in between used so PIDs/LQRs can be even better designed.
I’d like a full IMU suite complete with constant-drift gyros, 6-axis accelerometers, magnetometers, and be able to pull in quadrature encoders from the wheels, then characterize during all modes of robot operation (so motor magnetic fields can be compensated for). I’d also like to include fiducials from the vision system. Then I’d fuse everything together with a properly designed Kalman Filter (preferably all implemented in an FPGA).
Oh to dream…in reality, I’d just like 2 weeks of driving practice. Is that too much to ask for?
A quantum singularity powered matter transporter a la Star Trek.
At the start of the match it would beam all the competitors unharmed into the pattern buffer.
Then let the rest of your alliance play unchallenged and finally beam all the rest of the game pieces to the maximum scoring positions just before it returns all the other robots to the field safe and sound.
I am just not sure if it would need wheels. Maybe it could beam itself to the pits.
I am bald so I can get away with this…‘Make it so’.
Probably a two-speed swerve drive with PTO and ield centric gyros and gyros for collision detection/ avoidance, a localized AI to get around defense, full holonomic motion, all controlled via an android driver who has perfect reflexes. Robot should be circular to spin past defense.
8 cim/minicim drive or 6 individual modules with a cim each.
Well I have been able to get accurate to within a few inches.
Using 2 cameras on a Lego or servo pan and basic geometry.
If FIRST were willing to put cameras on the field and retro-reflective tape on the robot bumpers this does not seem like a request that is really all that hard.
Might be handy for FIRST to have that information as well for the purpose of diagnostics.
Another way would be to weave wires into, or under the field rugs and induce a current into the grid with the robots.
This trick would work like a keyboard matrix. Then the accuracy would depend on the grid spacing.