816 - 90% Complete (Video inside)

[thefro526]

Here's a short video of one of our first test runs with our 2011 Robot.

http://www.youtube.com/watch?v=d7WoW7F-svo

We're working on making everything better, faster, stronger, more reliable, etc from here.

[Craig]

Very impressive! Nice maneuverability

Hope your gripper can take a few licks in combat!

[JesseK]

For more iterations and better practice, put that 12" wall up at the edges of your lanes. We found that it's somewhat difficult to see the tubes on the floor if you're in the lane diagonal from your driver's station. It will also get your drivers used to exiting the lanes where they're supposed to. For bonus practice, get your 2010 robot out to play some blocking defense.

Nice & Simple -- I like it.

[thefro526]

For more iterations and better practice, put that 12" wall up at the edges of your lanes. We found that it's somewhat difficult to see the tubes on the floor if you're in the lane diagonal from your driver's station. It will also get your drivers used to exiting the lanes where they're supposed to. For bonus practice, get your 2010 robot out to play some blocking defense.

Nice & Simple -- I like it.

Thanks!

We actually added the lane dividers to our field the night after this was taken and the driver didn't seem to have all that much difficulty with them. We train him to not hit the wooden walls. (messes our bumpers up)

[Andrew Remmers]

hm... Looks like a familiar robot. Reminds me of Einstein :p

[nahstobor]

Very simple design, I can't wait to see it in action. It's scary to think that this machine is only going to get faster :eek:

[Ryan Dognaux]

Thanks for posting this video. It took ~30 seconds to travel across the field, grab the tube, travel back across the field and place the tube; I see this being very realistic for a lot of teams, especially during week 1. Human players throwing tubes to midfield will significantly reduce that time though.

[Brandon Holley]

Thanks for posting a video of a most completed robot in an unawkward angle/crappy resolution!

Looks great. With some practice I'm sure you will find some success this year!

-Brando

[thefro526]

Thanks for posting this video. It took ~30 seconds to travel across the field, grab the tube, travel back across the field and place the tube; I see this being very realistic for a lot of teams, especially during week 1. Human players throwing tubes to midfield will significantly reduce that time though.

An approximately ~30 second cycle time will probably be a good benchmark for an "average" robots. Robots that are above average will probably get to around a ~20 second cycle time, which actually doesn't seem to be all that unfeasible. We'll have a few more teams coming to our practice field throughout the week, so we'll be able to get some good measure of how long it takes us to score with obstructions. (Though I won't be able to speak for other teams)

Also, we've found that floor pickup is the way to go, since it allows you to shorten the distance that you HAVE to travel to get the tubes.

[Tetraman]

Oh boy. The one thing I forgot about when working on a team, is that feeling you get when you see a robot that does pretty much exactly what your team's robot does, but does it about 150% better - expecially when it comes to the construction and build. We need to start learning how to slim down our designs and make them less bulky.

Very nicely done. And wowzah on the full field. Good luck this year!

[robodude03]

Also, we've found that floor pickup is the way to go, since it allows you to shorten the distance that you HAVE to travel to get the tubes.

Our team is definitely leaning towards the same thought. The cycle time will be very important in terms of maximizing the number of tubes that are scored. Anyhow your bot is looking really good Dustin, I'm really liking the simplicity in design!

[rcmolloy]

When you showed me the video yesterday, I was heavily intimidated. Although, we should be up to your status by weeks end. :)

Nice job man. Really, I am very excited to play with you guys in Philly and Jersey!

[Dancin103]

816, it looks awesome as usual! I can not wait! Good luck in the 2011 season. :)

Cass

[steelerborn]

Yeah a great robot will be able to place one complete logo alone on the top row, and still have time to be positioned for mini-bot deployment.

Also have you guys put your arm through some serious testing?
I noticed that you raise your arm outside of your safe zone, could you arm and bot take a really hard side hit in that zone?
I also see a lot of swaying back and forth when you are moving in that zone. A robot that only plays defense may torque that arm way out of line, perhaps making the arm a little stronger, or bringing a good supply of spares may be an option (so long as they are light).

And one more point, when you were driving and making some sharp turns your claw seemed to also sway side to side, making a nice "cradle" where your arm could rest in could easilly solve this problem.

We have played with the arm lengths and wrist lengths so that we can raise and lower the arm with ease while completly in the safe zone, to make sure we stay functional.

Just a thought, but looks pretty good so far.

[thefro526]

Yeah a great robot will be able to place one complete logo alone on the top row, and still have time to be positioned for mini-bot deployment.

Also have you guys put your arm through some serious testing?
I noticed that you raise your arm outside of your safe zone, could you arm and bot take a really hard side hit in that zone?
I also see a lot of swaying back and forth when you are moving in that zone. A robot that only plays defense may torque that arm way out of line, perhaps making the arm a little stronger, or bringing a good supply of spares may be an option (so long as they are light).

And one more point, when you were driving and making some sharp turns your claw seemed to also sway side to side, making a nice "cradle" where your arm could rest in could easilly solve this problem.

We have played with the arm lengths and wrist lengths so that we can raise and lower the arm with ease while completly in the safe zone, to make sure we stay functional.

Just a thought, but looks pretty good so far.

Thanks for the input.

The lower link has been replaced with an aluminum unit, which has made the arm substantially more rigid as it travels up and down. The swaying left to right has been taken care of by adding more support to tie the top of the arm super structure in with the base. We've also added a guide of sorts that the arm falls into while in the down position which limits the sway of the claw to approximately a half inch.

As far as the arm being raised outside of the safe zone, it's not really something we're worried about - yet. When at it's lowest point and at the top scoring point the arm is contained completely within the frame and bumper perimeter - and the arm can be raised from it's lowest point to it's highest point while remaining completely in the safe zone. More often than not, if we're approaching the scoring grid unobstructed then we will raise our arm a bit early, but if someone is waiting to play defense on us then we will wait until we're in the safe zone to raise it.

[Madison]

Thanks for the input.

The lower link has been replaced with an aluminum unit, which has made the arm substantially more rigid as it travels up and down. The swaying left to right has been taken care of by adding more support to tie the top of the arm super structure in with the base. We've also added a guide of sorts that the arm falls into while in the down position which limits the sway of the claw to approximately a half inch.

As far as the arm being raised outside of the safe zone, it's not really something we're worried about - yet. When at it's lowest point and at the top scoring point the arm is contained completely within the frame and bumper perimeter - and the arm can be raised from it's lowest point to it's highest point while remaining completely in the safe zone. More often than not, if we're approaching the scoring grid unobstructed then we will raise our arm a bit early, but if someone is waiting to play defense on us then we will wait until we're in the safe zone to raise it.

What was the lower link previously? PVC?

Our machine shows a lot of the same characteristics that yours does and it seems like we've taken a lot of the same steps to mitigate the problems. I've been considering replacing the lower link with aluminum or fiberglass, but we see a lot of elastic deformation in that member and I'm worried that the replacement will plasticly deform instead of stiffen things up.

[Akash Rastogi]

Dustin, this robot is god awful...It has nowhere near the respectable amounts of orange and blue yet. Get on that.

[camtunkpa]

looks good 816. I could have sworn I saw a similar gripper somewhere around our shop. I can only hope ours works as well as yours does. Practice it up and add some more orange and blue and you should be good to go this season!

[J_Miles]

I think cycle time will be a big deal, but I also think that there could be serious benefits with creating a system that easily obtains game pieces from the human players:

1. There is not contention for the tubes dispensed by the human players

2. Those tubes are ALWAYS in the same location, with little to no variance, and can be fed directly into a robot's gripper with a significant amount of precision

I feel like with a large number of teams, there might be something in taking the extra few seconds to go to the human player's station because it has the capacity to save lots of time over trying to acquire a game piece that is on the floor with the drivers having only an obstructed view of it. At the same time, I'm fully aware that there will be teams with a human player who can throw game pieces consistently to a spot, and those robots will be able to immediately pick up and place the tubes. I think, though, that most teams will find some balance right in the middle, between slow floor pickup and reliance on the human player.

[Daniel_LaFleur]

Thanks for posting this video. It took ~30 seconds to travel across the field, grab the tube, travel back across the field and place the tube; I see this being very realistic for a lot of teams, especially during week 1. Human players throwing tubes to midfield will significantly reduce that time though.

I see 30 seconds as unrealistic as there will be 5 more robots on the field and some of those will be playing defense. I believe 1 minute cycletimes will be closer to the norm.

I also believe that throwing in the tubes will only entice your opponents to take the tubes your human player has 'provided'

[J_Miles]

I see 30 seconds as unrealistic as there will be 5 more robots on the field and some of those will be playing defense. I believe 1 minute cycletimes will be closer to the norm.

Here I disagree with you. I think there will be a good number of teams who will be able to hit that 20-40 second range easily.

I also believe that throwing in the tubes will only entice your opponents to take the tubes your human player has 'provided'

Here, though, I entirely agree. I think that throwing tubes onto the playing field will be a brilliant strategy...sometimes. At other times, I think that it would benefit teams more to deliver directly to the robots.

[thefro526]

What was the lower link previously? PVC?

Our machine shows a lot of the same characteristics that yours does and it seems like we've taken a lot of the same steps to mitigate the problems. I've been considering replacing the lower link with aluminum or fiberglass, but we see a lot of elastic deformation in that member and I'm worried that the replacement will plasticly deform instead of stiffen things up.

Madison, the original lower link was 1-1/4" Sch 40 PVC. We noticed a lot of elastic deformation under load, and our operator complained about the amount of "bounce" while lifting and moving. We replaced the lower link with a 1-5/8 OD Aluminum tube with a .109 wall and that has fixed all of our issues.

I noticed that the PVC arm had approximately 1/2" of an outward bow in the center after it was removed and showed that it had begun to stress and discolor. There is no doubt in my mind that the link would have broken at some point in our season, I'm just not sure when.

Dustin, this robot is god awful...It has nowhere near the respectable amounts of orange and blue yet. Get on that.

Orange and Blue will come eventually, I'm more concerned with Mini-Bots at the moment. Lol.

[Barry Bonzack]

Best of luck to 816, its fantastic that you all have a week to practice driving and programming. Thanks for posting a non-cryptic video.

[Daniel_LaFleur]

Here I disagree with you. I think there will be a good number of teams who will be able to hit that 20-40 second range easily.




I guess we'll just have to wait and see.

By my numbers an alliance will average 3-4 tubes per match, by your numbers an alliance will average 10 tubes per match.

Lets see who's closer ;)

[thefro526]

Best of luck to 816, its fantastic that you all have a week to practice driving and programming. Thanks for posting a non-cryptic video.

Thanks Barry. This video is actually from last Wednesday or Tuesday night, so we've been up an running for nearly a week now. We learned in 2008 that having at least a week or two to debug is extremely important if a practice bot isn't available after ship so we work extremely hard to get our design finalized early.

Oh, and to answer your question posted on the video, we were inspired by the original Orange 4-bar. (1902 in 2007) :D We can only hope to have the same success that 1902 had that year.

Also, I'm really glad that people found this video so helpful for strategy discussion and to get a feel for the game. We have more videos that will be up later this week, we just can't exactly reveal the completed robot before the sponsors and team families see it. ;D

[Basel A]

816, looks like you're most of the way to being a very competitive robot! I envy that you'll have a full week for programming/driving practice/etc. That could really be the difference.

I guess we'll just have to wait and see.

By my numbers an alliance will average 3-4 tubes per match, by your numbers an alliance will average 10 tubes per match.

Lets see who's closer ;)

Without defense, it looks like there'll be many teams that can do a sub-30 second run without defense (816 looks capable). With defense, I don't think this would balloon up past 45. For some teams it could even still be under 30!

[MrFixIt2719]

Looks good like your tube intake

[Madison]

Out of curiosity -- did the P60 failure you experienced happen before or after you replaced your link with aluminum?

Our gearbox is, so far, doing okay, but we're also investigating an aluminum replacement and I am wary that the added rigidity might cause or accelerate a failure.

[Ryan Dognaux]

I see 30 seconds as unrealistic as there will be 5 more robots on the field and some of those will be playing defense. I believe 1 minute cycletimes will be closer to the norm.

Fair enough, I still think a lot of teams will be able to score a tube in around 30 seconds, especially come Saturday at most regionals.

I also believe that throwing in the tubes will only entice your opponents to take the tubes your human player has 'provided'

This is where your human player strategy is key. The human player absolutely cannot throw the tube too early. Doing this leaves the tube open to being stolen by your opponents. A truly skilled human player will time his or her throws perfectly so that their alliance's robot is ready to pick up the tube the second it lands. Human players should definitely be factored into scouting this year as an alliance with two skilled human players can help tip the scale in an elimination match.

[thefro526]

Out of curiosity -- did the P60 failure you experienced happen before or after you replaced your link with aluminum?

Our gearbox is, so far, doing okay, but we're also investigating an aluminum replacement and I am wary that the added rigidity might cause or accelerate a failure.

Madison, the gearbox failure happened during initial testing of the Aluminum link. That being said, we were never able to work the PVC link as hard as we were the Aluminum link because we were afraid that the PVC was going to snap. It's a bit hard for me to say that the aluminum caused the failure, but at the same time it's difficult for me to say that the aluminum didn't cause the failure.

Sorry I can't be of more help.

[Madison]

Madison, the gearbox failure happened during initial testing of the Aluminum link. That being said, we were never able to work the PVC link as hard as we were the Aluminum link because we were afraid that the PVC was going to snap. It's a bit hard for me to say that the aluminum caused the failure, but at the same time it's difficult for me to say that the aluminum didn't cause the failure.

Sorry I can't be of more help.

No worries -- I just want to keep an eye on things as we continue testing and our robots are very similar. I think the PVC is absorbing a lot of the shock loading we've imparted into the system by rapidly stopping and changing direction. The aluminum won't do that nearly as well and we'll likely need to be kinder to the gearbox.

I just ordered a 1.25" OD x .083" wall 2024 Aluminum Round Tube. We'll swap it in on our practice robot tomorrow and see how things go. Also, for what it's worth, we've been running our motor at 40% of its full power. It's otherwise too fast.

[thefro526]

No worries -- I just want to keep an eye on things as we continue testing and our robots are very similar. I think the PVC is absorbing a lot of the shock loading we've imparted into the system by rapidly stopping and changing direction. The aluminum won't do that nearly as well and we'll likely need to be kinder to the gearbox.

I just ordered a 1.25" OD x .083" wall 2024 Aluminum Round Tube. We'll swap it in on our practice robot tomorrow and see how things go. Also, for what it's worth, we've been running our motor at 40% of its full power. It's otherwise too fast.

Cool, once we get our robot back together tonight, I'll try to get some more data for you.

What speed does 40% power (I'm assuming 40% signal) equate to? I know that at 100% signal our arm goes up in ~1 - 1.1 seconds (90* per second), but we usually go up in around 2 seconds because it's more controllable.

[ZeroGAdam]

Great Job Guys! I am always amazed at what your team is able to come up with every single year! Its always effective, simple , and a marvel among all teams. I wouldn't be surprised to see you guys go very far into Saturday like every year. Awesome job!

See you at The NJ Regional!!

[Jeffy]

We are running a similar mechanism with an rs550 motor in a 256:1 gearbox with a 2:1 reduction after the gearbox. Are you able to backdrive the gearbox when the robot is off?
It does not seem possible with our robot.

[thefro526]

We are running a similar mechanism with an rs550 motor in a 256:1 gearbox with a 2:1 reduction after the gearbox. Are you able to backdrive the gearbox when the robot is off?
It does not seem possible with our robot.

Yes, we are able to back drive our arm while the robot is un-powered, it takes a light push to get it to start going down and then once it starts going it usually goes all the way down by itself.

Our current setup is as follows: 1 RS775 Motor in a 256:1 P60 Gearbox on each side of our tower (total of two), each with a 15 tooth sprocket that drives a 48 tooth sprocket on our arm for an effective ratio of 3.2:1 from the transmission or an 819.2:1 reduction from the motor. We also have a decent amount of surgical tubing on our arm as an assist which also makes the mechanism harder to back drive.

We're swapping out out the 15 tooth sprocket on our transmission for a 10 tooth sprocket tonight, which makes the ratio 4.8:1 from the transmission or 1228.8:1 from the motor. I'll report back tonight or tomorrow if this changes the ability to back drive the arm.

Also, how much, if any, assist is on your arm? Enough springs or gas struts would make the arm nearly impossible to back drive by hand. You can also look for sources of inefficiency (binding, improper shaft alignment, etc.) and see if that is adding extra resistance which would make the arm harder to back drive. FYI, as the banebot transmissions start to fail, they will become nearly impossible to back drive without power. Check your transmissions too.

[Jeffy]

Yes, we are able to back drive our arm while the robot is un-powered, it takes a light push to get it to start going down and then once it starts going it usually goes all the way down by itself.

Our current setup is as follows: 1 RS775 Motor in a 256:1 P60 Gearbox on each side of our tower (total of two), each with a 15 tooth sprocket that drives a 48 tooth sprocket on our arm for an effective ratio of 3.2:1 from the transmission or an 819.2:1 reduction from the motor. We also have a decent amount of surgical tubing on our arm as an assist which also makes the mechanism harder to back drive.

We're swapping out out the 15 tooth sprocket on our transmission for a 10 tooth sprocket tonight, which makes the ratio 4.8:1 from the transmission or 1228.8:1 from the motor. I'll report back tonight or tomorrow if this changes the ability to back drive the arm.

Also, how much, if any, assist is on your arm? Enough springs or gas struts would make the arm nearly impossible to back drive by hand. You can also look for sources of inefficiency (binding, improper shaft alignment, etc.) and see if that is adding extra resistance which would make the arm harder to back drive. FYI, as the banebot transmissions start to fail, they will become nearly impossible to back drive without power. Check your transmissions too.

Thank you very much. We have fixed it. It was a shot transmission that was likely messed up somewhere in very early prototyping. And now it has a burnt motor too :(
Incase you want a comparison (in no way am I trying to brag):
We are running a 256:1 w/ a 550 motor and a 12:25 after giving ~533:1.
Our arm will hold its position in about 30 degrees of motion only from the surgical tubing. It holds fine in all parts with the motor connected.
We also upgraded to aluminium arms with the same OD (1") and half the wall thickness. They are slightly heavier (~20%) than the pvc setup we had.

BTW, your bot looks great. I hope to see it sometime in person. I am also a fan of anything orange too.