Quote:
Originally Posted by VeqIR
We encountered the same question after our first district and spent a lot of time analyzing how we were spending time for each stack. After our first district, we found we were taking a lot of time adding the noodle to our stack, aligning to the chute and intaking totes.
To fix alignment time, we added sensors to allow us to hit a button to automatically align to the tote chute (saved ~8s/stack). To improve noodling time, we moved from using the chute to going over the wall (saved ~6s/stack). To improve tote intake time, we tuned our intake system so that the tote always stops at the same spot before being added to the stacker (~10s/stack). It should be noted that our robot has a built-in ramp.
At this point, we are looking at around 3s/tote and ~35-40s per stack when we are actively loading from the HP (without using the 254/1114-style ramp).
If you have video of your competitions, I recommend timing each element of your stacking and find out what actions take the most amount of time. Once you have that info, you can more accurately target optimizations for your stacking time. You may find out that 4-5s/tote is completely tolerable if you optimize another element of your stacking (aligning, noodling, outputting).
Best of luck to you!
|
This is all great advice.
We started the season by guessing/prototyping how long various actions would take and building a model of how we thought the scoring would work. Of course, some of our assumptions turned out to be incorrect, and the robot as built at our first competition was not quite finished in a few areas.
After our first competition at CVR, we had video of each of our matches. We went through and analyzed precisely how long different actions took (replacing our estimated numbers), compared them to the "best of breed" in FRC (1114, 118, 987, 148, etc.), made adjustments where we found room for improvement, and were able to just about double our scoring output by SVR.
Let that be a lesson. A video camera is not very expensive. Any team should take video of their own matches and go through the exercise of identifying optimizations. There is typically a lot of "low hanging fruit" that requires no or minimal mechanical changes. This exercise can literally double how effective your robot is, no matter whether you are a rookie or a longtime vet.