Hard to tell on the video but check out rule and team Q&A. Robot is designed to trap stacked cones against the wall to minimize inconsistencies. Cones on taped line. GS6a Control / Possession Limit, one cone. (Minor for each cone + Minor for each cone for each 5 seconds)
Thanks for checking on us, we already had a talk with refs about it at our first meet. As long as it doesn’t give use an advantage or the dirstance isn’t extreme like last season with the shared alliances hubs it’s fine.
Do you use any special programs such as roadrunner? And if not and you are only using drive encoders, how are you able to be consistant every single time?
We had a member design it, then we used a laser cutter to engrave the number into it. We actually created it last year and just took it from our old robot.
Hello! Do you use pid control or feedforward tuning? I try to use roadrunner using only drive encoders but i don t know which one would be easier to tune and use
PID isn’t that hard to do, but we will use free forward for our next robot. I personally think Roadrunner is easier to do, but free forward is more flexible for us.
Did you use opencv for your detection? Also do you use a camera to identify cups against the walls, or is it pre programmed instructions? I’m just curious, it’s cool either way.
We use a custom model tensor flow because we didn’t want to learn open cv. We do not use a camera to identify the cups against the wall. We just use pre programmed instructions and have our robot tuned pretty well.
@@ftcmicrocosm that’s really smart actually. I totally understand the pain of learning opencv, but if you want to learn, I highly recommend looking at team wolfcorps’ tutorial on UA-cam. That’s how I learned. Very very fun. Do you have any good tutorials for tensor flow
Honestly, it would be better to just use Java and not blocks. As a team we hav en ever used blocks so we don’t really have any tips. Honestly the hardest part about coding is connecting the robot and figuring out how to set it up, not actually the coding part.
Hard to tell on the video but check out rule and team Q&A. Robot is designed to trap stacked cones against the wall to minimize inconsistencies. Cones on taped line. GS6a Control / Possession Limit, one cone. (Minor for each cone + Minor for each cone for each 5 seconds)
Thanks for checking on us, we already had a talk with refs about it at our first meet. As long as it doesn’t give use an advantage or the dirstance isn’t extreme like last season with the shared alliances hubs it’s fine.
Do you use any special programs such as roadrunner? And if not and you are only using drive encoders, how are you able to be consistant every single time?
We were using road runner in the video. We are only using drive encoders as well.
Hey team, how can you make the robot encoders speed up and then slow down?
We use roadrunner
@@ftcmicrocosm Is ur team just using the built in encoders on the drive or use are you guys using a dead wheel?
We use the encoders on our motors, not dead wheels. We might add some dead wheels by the end of the season.
This is so good! I've been trying to use drive encoders with PID as well but have been having issues. Do you have your code posted on github?
I will make sure our programmers do that on Tuesday. We do not have our current code on GitHub I think.
The thing on the side of your robot that has your number did you make it or did you design then order because it’s cool
We had a member design it, then we used a laser cutter to engrave the number into it. We actually created it last year and just took it from our old robot.
Hello! Do you use pid control or feedforward tuning? I try to use roadrunner using only drive encoders but i don t know which one would be easier to tune and use
We use PID for it
PID isn’t that hard to do, but we will use free forward for our next robot. I personally think Roadrunner is easier to do, but free forward is more flexible for us.
🎉🎉🎉
Did you use opencv for your detection? Also do you use a camera to identify cups against the walls, or is it pre programmed instructions? I’m just curious, it’s cool either way.
We use a custom model tensor flow because we didn’t want to learn open cv. We do not use a camera to identify the cups against the wall. We just use pre programmed instructions and have our robot tuned pretty well.
@@ftcmicrocosm that’s really smart actually. I totally understand the pain of learning opencv, but if you want to learn, I highly recommend looking at team wolfcorps’ tutorial on UA-cam. That’s how I learned. Very very fun. Do you have any good tutorials for tensor flow
Not really, we are thing about creating one ourselves
Nice
Thanks
Did you use blocks or java? And did you use motor encoders?
Java, and we used motor encoders.
Any chance you would have tips on programming a mecanum wheel autonomous in blocks using encoders?
Honestly, it would be better to just use Java and not blocks. As a team we hav en ever used blocks so we don’t really have any tips. Honestly the hardest part about coding is connecting the robot and figuring out how to set it up, not actually the coding part.
How did you make it so perfect. Did you use any sensors to detect where it goes?
Nope, we only used encoders from our wheel motors. There were no odometry wheels or sensors.
Do you guys use PID and IMU?
PID
@@ftcmicrocosm Yo also did yall use roadrunner or did you guys just make your own PID? lol
We used roadrunner for that robot.