Einstein Can-Race Compilation Video

[Chris Hapstack]

I recorded most (twelve) of the can races on Einstein and compiled the clips into a video here (http://youtu.be/CYEv_uu3rtY). Slow-mo of all clips is also included.

Edit: This is original video shot by me, not from the webcast.

[Kevin Leonard]

This is the coolest thing to come out of this year, byfar.
Fantastic.

[Ichlieberoboter]

That's awesome. And wow, that airplane thing really got out of hand...

[Govind Girish]

Team 118 would keep grabbing the containers at the exact same time as 987 and both would get stuck, but 118 also programmed the robot to let go during autonomous, allowing 987 to grab the containers

[Caleb Sykes]

Team 118 would keep grabbing the containers at the exact same time as 987 and both would get stuck, but 118 also programmed the robot to let go during autonomous, allowing 987 to grab the containers

118 did win once when they ripped off (https://www.youtube.com/watch?v=CYEv_uu3rtY&feature=youtu.be#t=2m16s) one of 987's hooks.

[Chad987]

118 did win once when they ripped off (https://www.youtube.com/watch?v=CYEv_uu3rtY&feature=youtu.be#t=2m16s) one of 987's hooks.

Yep. They were designed to have a single ziptie holding them on to avoid larger damage. I believe we started running 2 zipties as a do or die method after that

[Lil' Lavery]

Looks like 1678 won all of their races. However, I don't they matched up against the fastest tier of burglers in any of their races. Curious to see how 1678 would have done against 987/2056/330.

[Electronica1]

Ok, so watch this video muted while playing this (https://www.youtube.com/watch?v=EAwWPadFsOA) in the background. It actually works pretty well. ;)

[Chad987]

Looks like 1678 won all of their races. However, I don't they matched up against the fastest tier of burglers in any of their races. Curious to see how 1678 would have done against 987/2056/330.

2512 was running the exact same can burgalers as us So I'm going to guess pretty similarly, though we begin driving forward almost instantly in auto. 1678s grabbers are awesomely fast.

[Bennett548]

548 is proud to have had some canburglars on Einstein!
1711 developed their own version of the canburglars that we cheesecaked with them at MSC. 1089 had a set that 548 and 1023 helped them get going just before and during Einstein.

Unfortunately, there were electrical or programming issues with the set on 1089, so they weren't going full speed.

548 ourselves had an even faster set on our robot, but had some bad matches on Hopper and went out in the semis.

I think it would be really cool to have a canburglar drag-race off to the side of the practice area at IRI, with an ultra high-speed camera to catch the action and find out the fastest.

[George Nishimura]

Looks like 1678 won all of their races. However, I don't they matched up against the fastest tier of burglers in any of their races. Curious to see how 1678 would have done against 987/2056/330.

Did they ever face up against 3310 or 4522 who seemed to have pretty fast canburglars on Newton? From watching them I didn't think 3310 could be beaten...

[Kevin Leonard]

548 is proud to have had some canburglars on Einstein!
1711 developed their own version of the canburglars that we cheesecaked with them at MSC. 1089 had a set that 548 and 1023 helped them get going just before and during Einstein.

Unfortunately, there were electrical or programming issues with the set on 1089, so they weren't going full speed.

548 ourselves had an even faster set on our robot, but had some bad matches on Hopper and went out in the semis.

I think it would be really cool to have a canburglar drag-race off to the side of the practice area at IRI, with an ultra high-speed camera to catch the action and find out the fastest.
Well thank you very much for 1711's canburglars!

[John Retkowski]

One thing I also noticed about 1678's canburgler was that they had the robot driving forward even before the arms fell into the cans. I think this really helped them get the cans off the step and out of the way of the opposite bot, so that there wouldn't even have to be a tug of war had.

[Anteprefix]

Did they ever face up against 3310 or 4522 who seemed to have pretty fast canburglars on Newton? From watching them I didn't think 3310 could be beaten...

1678 faced off directly against 3310 in Semifinal 2 on Newton and won the can battle.

[stens987]

This video was great! Thanks for sharing!

[barn34]

We were hoping to get the opportunity to test ours against the best this weekend. Unfortunately, that opportunity will have to apparently wait until IRI.

https://m.youtube.com/watch?feature=youtu.be&v=G3o_VpobYFs

[Louis_]

Ok, so watch this video muted while playing this (https://www.youtube.com/watch?v=EAwWPadFsOA) in the background. It actually works pretty well. ;)

+1 edit this into the song.

[trumpthero786]

Thank you for this great video!

[Chris Hapstack]

Thanks, everyone! I figured video was the only way I was going to clearly see the races from the upper deck.

I think 548 embodied the best parts of the whole cheesecakeing trend by continuing to help teams even when their personal interests were no longer at stake. I was definitely hoping for 548 to get to Einstein so we could see a 548 v. 1711 matchup, kind of an ironic master/pupil showdown.

Not getting to see 1114's harpoons in action was unfortunate, and I'm convinced that the burglars 254/973 debuted in the Carson tournament could have beaten anything we saw on Einstein. According to 254, they could snag the cans in 120ms. They were shockingly fast in person.

[JB987]

Just to add to the discussion...
987's bot was programmed to start back driving at .07 seconds and as you can see from the slow-mo our weakness was the flexibility (which we went with to minimize damage) of our arms. We lost some precious time reduction for our claw to the can as a result. Kudos and congratulations to the winning alliance! As so many of us were saying all along...it's about the cans man (mostly;) )

[seanthompson]

1678 faced off directly against 3310 in Semifinal 2 on Newton and won the can battle.

Pretty sure they agreed to each take two of the cans, and never actually faced off.

[Anteprefix]

Pretty sure they agreed to each take two of the cans, and never actually faced off.

I seem to recall 3310 losing the race and then grabbing 2 from the other side during teleop. I could be mistaken though. Did anyone take video of Newton eliminations?

[DampRobot]

I was surprised how exciting and fun this aspect of the game was (even though it really only came into play at CMP elims and Einstien). It's also impressive how dominant 1678s grabbers were, I'd really like to take a look at them one day.

[Bennett548]

Don't thank just us! The Raptors somehow built those in two weeks with no unbag time. It took us three to build our first set, and we had some unbag time.

I also forgot to say that team 67 had some canburglars that seem to have been partly inspired by 548's. At msc they used polycarbonate arms, but I hear that at World's they came over to the dark side and used carbon fiber.

I'm going to try to get the students to put together a whitepaper about how those canburglars were designed and built.

[orangelight]

Don't thank just us! The Raptors somehow built those in two weeks with no unbag time. It took us three to build our first set, and we had some unbag time.

I also forgot to say that team 67 had some canburglars that seem to have been partly inspired by 548's. At msc they used polycarbonate arms, but I hear that at World's they came over to the dark side and used carbon fiber.

I'm going to try to get the students to put together a whitepaper about how those canburglars were designed and built.

Great we're going to write a paper...

[TJP123]

I think it would be really cool to have a canburglar drag-race off to the side of the practice area at IRI, with an ultra high-speed camera to catch the action and find out the fastest.


We had the same thought. That might be more fun than the full event.

[tindleroot]

Correct me if I'm wrong, but I noticed that there were not any teams on Einstein that were featured MAINLY for their canburgling (e.g. 548, 3310, 503, etc.) While the canburglars on Einstein were lightning fast, they were not THE fastest compared to some of these other teams.

I think the reason this tends to happen is because many of the top seeds decided to pick the top teams for stacking ability in addition to canburgling, and teams like 118, 987, 1114, etc. had both fast burglars and superior stacking ability (548 does too). When our team was making our pick list Friday night, we decided to favor 68 just ahead of 503 since 68 could make more stacks in addition to their canburglar, even though 503's canburglar was proven and fast. I'm sure a lot of alliance captains without canburglars had similar discussions as well.

[orangelight]

We had the same thought. That might be more fun than the full event.
2016 Game: Crazy Canburglars

[Brian Selle]

Thanks for this video, great job.

1678 faced off directly against 3310 in Semifinal 2 on Newton and won the can battle.

This is not true. We agreed prior to take opposite sides. Based on slow motion video 1678 gets to the can about 20-30ms before 3310 but that wouldn't have mattered as we would have locked in the hole before the cans moved. It would have been an epic tug-of-war.

[Kevin Leonard]

Correct me if I'm wrong, but I noticed that there were not any teams on Einstein that were featured MAINLY for their canburgling (e.g. 548, 3310, 503, etc.) While the canburglars on Einstein were lightning fast, they were not THE fastest compared to some of these other teams.

I think the reason this tends to happen is because many of the top seeds decided to pick the top teams for stacking ability in addition to canburgling, and teams like 118, 987, 1114, etc. had both fast burglars and superior stacking ability (548 does too). When our team was making our pick list Friday night, we decided to favor 68 just ahead of 503 since 68 could make more stacks in addition to their canburglar, even though 503's canburglar was proven and fast. I'm sure a lot of alliance captains without canburglars had similar discussions as well.

1711's job on the 1325 alliance was to right containers and burgle cans. And if there was none of that, they'd pull totes 1 by 1 from the landfill

It wasn't a glamorous job, but it was better than 20 doing the same job for sure. Our canburglars were much slower, and much less consistent.

[galewind]

548 is proud to have had some canburglars on Einstein!
1711 developed their own version of the canburglars that we cheesecaked with them at MSC. 1089 had a set that 548 and 1023 helped them get going just before and during Einstein.

Unfortunately, there were electrical or programming issues with the set on 1089, so they weren't going full speed.

548 ourselves had an even faster set on our robot, but had some bad matches on Hopper and went out in the semis.

I think it would be really cool to have a canburglar drag-race off to the side of the practice area at IRI, with an ultra high-speed camera to catch the action and find out the fastest.

In a frantic race to get them running, we had to drop the PID out and run straight state machine (Drop arms at 100% until one pot reached a set value). We noted some hardware hitching when we were testing them at 30% in the pits (The one side with the broken pot was off speed).

I'm not sure if our auton selection from the smartdash took too much time to initialize, but we powered the motors at 100%. Unfortunately we spent so much time in the morning with the first attempt that we didn't get enough time to diagnose issues with the second.

That said, we're incredibly thankful that you guys gave us the opportunity to be cheesecaked. Absolutely worked our tails off to get not one, but two different solutions implemented throughout the morning and into the evening. Another story for another post :).

[rpatel3001]

Absolutely thank you so much to 548 for that wicked fast cheesecake! If I remember right, Mr. G, it was the left pot we were reading and the left gearbox that had problems. (My arm is indeed still hurting after being thwacked at 30%)

After thinking it through I'm pretty sure I know what that initialization delay is and how to optimize it. We should get an instance of SendableChooser.getSelected() in autonomousInit() instead of making a request every time(that's silly idk why we did that) or even hard code it instead of allowing selection if it comes down to it. Other than that, I'm sure there's a couple milliseconds of delay due to the structure of our state machine. (Maybe we'll get a rematch with 118 at IRI ;) )

[Abhishek R]

Besides their speed, it looks like 1678's canburglar was designed to tilt the cans towards their alliance immediately upon impact, increasing the probability that they pull away with the cans in hand even if the opposing robot managed to get into the hole.

It's too bad I was pretty tied up this weekend, it's a shame I wasn't able to see the Citrus Circuits machine before Saturday (definitely wanted to check out that feeder intake system too). Congratulations to them and their alliance!

[Chris Hapstack]

The tilting issue was one thing that made me think 1114's harpoon guns could still be beaten by standard arms. As seen in the video that 1114 posted here (https://instagram.com/p/16UeQ_hnGK/), the harpoons tilt the cans towards the opposing side pretty significantly when they hit.

[Electronica1]

The tilting issue was one thing that made me think 1114's harpoon guns could still be beaten by standard arms. As seen in the video that 1114 posted here (https://instagram.com/p/16UeQ_hnGK/), the harpoons tilt the cans towards the opposing side pretty significantly when they hit.

Makes you wonder, if two opposing robots were able to get their hooks in the cans, would the two robots be able to tip 1114's towers/ pull them onto their side of the field?

[RoboChair]

Thanks, everyone! I figured video was the only way I was going to clearly see the races from the upper deck.

I think 548 embodied the best parts of the whole cheesecakeing trend by continuing to help teams even when their personal interests were no longer at stake. I was definitely hoping for 548 to get to Einstein so we could see a 548 v. 1711 matchup, kind of an ironic master/pupil showdown.

Not getting to see 1114's harpoons in action was unfortunate, and I'm convinced that the burglars 254/973 debuted in the Carson tournament could have beaten anything we saw on Einstein. According to 254, they could snag the cans in 120ms. They were shockingly fast in person.

I would like to mention that our can grabbers were only operating at 80% power the whole event. We could have run as much as 120% power if we really wanted the cans. Our can grabbers really could dial it to 11, but it would be brutal on the driving mechanism and we deemed the wear and tear to not be worth it and turned out to not need it.

[RoboChair]

Besides their speed, it looks like 1678's canburglar was designed to tilt the cans towards their alliance immediately upon impact, increasing the probability that they pull away with the cans in hand even if the opposing robot managed to get into the hole.

It's too bad I was pretty tied up this weekend, it's a shame I wasn't able to see the Citrus Circuits machine before Saturday (definitely wanted to check out that feeder intake system too). Congratulations to them and their alliance!

You are completely correct about our grabbers pitching the cans forward on impact. We might release some of our test footage later this year showing our full power tests.

[Jared]

You are completely correct about our grabbers pitching the cans forward on impact. We might release some of our test footage later this year showing our full power tests.

That would be really awesome if you guys did. Those can grabbers were insanely fast and scary to watch.

[Paul Copioli]

1678 faced off directly against 3310 in Semifinal 2 on Newton and won the can battle.

This is not true.

We lined up on our side and they lined up on theirs. We did not battle each other but I do have slow motion video showing it to be a virtual tie if we did line up against each other (they beat us to the can by about 20 milliseconds).

In that match we got our 2 cans and 1678 got their two cans. There was no battle.

[Paul Copioli]

I would like to mention that our can grabbers were only operating at 80% power the whole event. We could have run as much as 120% power if we really wanted the cans. Our can grabbers really could dial it to 11, but it would be brutal on the driving mechanism and we deemed the wear and tear to not be worth it and turned out to not need it.

No, you can't dial it to 120% power. By definition that is 100% power.

There are students on this forum that may not understand this statement and it completely misleads them.

[Anteprefix]

Pretty sure they agreed to each take two of the cans, and never actually faced off.
Thanks for this video, great job.

This is not true. We agreed prior to take opposite sides. Based on slow motion video 1678 gets to the can about 20-30ms before 3310 but that wouldn't have mattered as we would have locked in the hole before the cans moved. It would have been an epic tug-of-war.

This is not true.

We lined up on our side and they lined up on theirs. We did not battle each other but I do have slow motion video showing it to be a virtual tie if we did line up against each other (they beat us to the can by about 20 milliseconds).

In that match we got our 2 cans and 1678 got their two cans. There was no battle.

Looks like someone was suffering from memory loss. In hindsight, waking up at 5 in the morning to watch alliance selections after 3 hours of sleep might not have been the best choice.

[RoboChair]

No, you can't dial it to 120% power. By definition that is 100% power.

There are students on this forum that may not understand this statement and it completely misleads them.

There is nothing wrong with my statement. It is exactly like over-volting an electric motor for more power. The motor is designed to take 12 volts(100%) but we run it at 9.6 volts(80%) to reduce strain, but could run it at 14.4 volts(120%) to make it run faster. Yes we can run a system with more than 100% power. We were using surgical tubing and could keep adding more and more past what it was designed for.

[Paul Copioli]

There is nothing wrong with my statement. It is exactly like over-volting an electric motor for more power. The motor is designed to take 12 volts(100%) but we were running at 9.6 volts(80%) to reduce strain, but could run it at 14.4 volts(120%) to make it run faster. Yes we can run a system with more than 100% power. We were using surgical tubing and could keep adding more and more past what it was designed for.

For the sake of sparing everyone in this forum, we'll just agree to disagree on this one.

[RoboChair]

For the sake of sparing everyone in this forum, we'll just agree to disagree on this one.

I am fine with that as well, meant no disrespect.

[Adam Freeman]

Don't thank just us! The Raptors somehow built those in two weeks with no unbag time. It took us three to build our first set, and we had some unbag time.

I also forgot to say that team 67 had some canburglars that seem to have been partly inspired by 548's. At msc they used polycarbonate arms, but I hear that at World's they came over to the dark side and used carbon fiber.

I'm going to try to get the students to put together a whitepaper about how those canburglars were designed and built.

Yep, our's were inspired 100% by 548's design that we initally saw when we paired with them at Livonia. We started designing a set to be used at MSC (poly carb version) a couple days later. They were 0.5s to the can due to the flexibility of the poly. After that we re-designed with carbon fiber, testing last Monday for the first time. Our testing showed them to be no slower than 250ms and much more repeatable. We knew this would be pretty fast, but feared we might be designing for an old target based on discussions with others.

Unfortunately we did not design for the can battle stalemate very well. Due to the nature of having them as part of our transport configuration required us to be able to attach them on the field very quickly. We did not have a robust attachment stategy which ultimately resulted in our arms faliure in the division finals. If we would have made it to Einstein our arms probably would have been ripped off the robot after the first tie. Which seemed pretty inevitable since we were very close to the same speed as most of the teams out there.

[Jaywalker1711]

Don't thank just us! The Raptors somehow built those in two weeks with no unbag time. It took us three to build our first set, and we had some unbag time.



To be completely honest, even I was shocked that we had them built and working in time. It was a lot more late nights in the build room and CAD lab.

It wasn't easy, we had to change our A-frame to accommodate the double burglars and then add mounts onto our drive base (and hack away a but of aluminum).

I just wish that we had more time on the practice field. They were never running at full power (because we were never consistent that way) and it took us all of quals to even get them working decently. I was ecstatic when they finally started working well in Carson elims.


And while 548 wasn't on Einsteins, 4 of their canburglars were. How cool is that.

[TJP123]

Unfortunately we did not design for the can battle stalemate very well. Due to the nature of having them as part of our transport configuration required us to be able to attach them on the field very quickly. We did not have a robust attachment stategy which ultimately resulted in our arms faliure in the division finals. If we would have made it to Einstein our arms probably would have been ripped off the robot after the first tie. Which seemed pretty inevitable since we were very close to the same speed as most of the teams out there.

Our end grabbers were about $1.50 worth of PVC held on with cotter pins. I can't find video yet, but on Archimedes QF6(?-our second match) and all three SF matches we got into some epic battles due to not wanting to back up and risk ruining a 20-point auton.

We got to the cans first, (maybe not against 1538, they were pretty fast) but pretty much let the other side catch up, then the fun began. We never lost a can that we hit, even in our first SF match where we didn't put the pins in. 15 feet of stretched surgical tubing later, we even ended up with that can. For about 30 seconds (x4), this game was as fun as any other.

I've got to find those videos.

[chapman1]

When reading through this thread and understanding how 548 Robostangs inspired and facilitated so many competing teams in the Finals by openly sharing their grabber technology, one gets a glimpse of the true meaning of co-opertition.