What is T-boning?

[Kevin Leonard]

T-boning can be a t-shaped pin. It can have the effect of the robot being t-boned moved sideways, or it can be against a wall.

T-Boning is generally creating bumper-to bumper contact with another robot with the front end of your robot pushing the other robot perpendicular to that robot's direction of motion.

The effect it has is often preventing motion in the desired direction, and thereby keeping the affected robot from scoring.

This can be better explained by someone smarter than me.

You can see it in most FRC matches this year that had effective defense.

[Max Boord]

What 1065 does to 233

[zinthorne]

T-bones are very hard to get out of especially when using mecanem wheels. The wheels have such little traction that you can't drive out of a t-bone like 254. Without using different bumper material or chassis, the best way is to have your driver get stuck in a few and then learn how to avoid them or drive spin or drive well enough to get out of them. I was caught in tons in the first district I drove and then was not caught in them hardly at all after. In my opinion the easiest way to avoid them is to be constantly moving. Also if you know that one is coming you can essentially spin out of it, if they do not hit you directly in the middle of your bumpers. Spin the corner the your robot the same direction that the robot is coming at you. Watch at about 1:04 when 3663 gets pinned by 1983.

https://www.youtube.com/watch?v=WbHHUtCGc1c

[tickspe15]

Originally Posted by zinthorne

T-bones are very hard to get out of especially when using mecanem wheels. The wheels have such little traction that you can't drive out of a t-bone like 254. Without using different bumper material or chassis, the best way is to have your driver get stuck in a few and then learn how to avoid them or drive spin or drive well enough to get out of them. I was caught in tons in the first district I drove and then was not caught in them hardly at all after. In my opinion the easiest way to avoid them is to be constantly moving. Also if you know that one is coming you can essentially spin out of it, if they do not hit you directly in the middle of your bumpers. Spin the corner the your robot the same direction that the robot is coming at you. Watch at about 1:04 when 3663 gets pinned by 1983.

https://www.youtube.com/watch?v=WbHHUtCGc1c

If you have your mecanum programmed properly and your drivers know how to take advantage of its holonomic capabilities i dont understand how you can possibly get t-boned. When somebody comes along side you all you need to do is translate away.

[Caleb Sykes]

Originally Posted by tickspe15

If you have your mecanum programmed properly and your drivers know how to take advantage of its holonomic capabilities i dont understand how you can possibly get t-boned. When somebody comes along side you all you need to do is translate away.

This.
Although not directly comparable, 33 this year was a good example of why having a low sideways cof could make it easy to slip out of t-bones. Given, this is all dependent on good driving, I'm sure it would have been easy to take advantage of 33's drivetrain when defending had they had poor driving ability.

[Chris is me]

Originally Posted by tickspe15

If you have your mecanum programmed properly and your drivers know how to take advantage of its holonomic capabilities i dont understand how you can possibly get t-boned. When somebody comes along side you all you need to do is translate away.

Gearing is also important. You need to be able to apply force to the ground while your wheels are being backdriven and your rollers are spinning due to the pushing force. Ideally, you're right - translating away from the pin will allow you to escape. Just be sure to do some testing when implementing your mecanum drive as this kind of stuff can be hard to model accurately. Lots of factors involving play at the rollers, construction of the wheels, rigidity of the frame, etc. are at play.

[Karthik]

Originally Posted by tickspe15

If you have your mecanum programmed properly and your drivers know how to take advantage of its holonomic capabilities i dont understand how you can possibly get t-boned. When somebody comes along side you all you need to do is translate away.

There are a surprising number of mecanum robots that are not actually programmed to go sideways; they simply use mecanum wheels in a tank drive setup...

[BBray_T1296]

Originally Posted by Karthik

There are a surprising number of mecanum robots that are not actually programmed to go sideways; they simply use mecanum wheels in a tank drive setup...

So like a normal omni-tank drive, but with 4 wheel independent gear-trains. Sounds expensive.

[Andy A.]

Originally Posted by BBray_T1296

So like a normal omni-tank drive, but with 4 wheel independent gear-trains. Sounds expensive.

I doubt many such configurations are planned. It's probably more a case of not getting the software (and/or hardware) working correctly, and tank steering being a contingency plan.

[IronicDeadBird]

If you are going by the conventional concept of T-boning as applying a force perpendicular to an objects force then it is actually something that you can prevent from happening with design or driver skill.

In retrospect I sound smarmy as all get out...
Think about how you can change how other robots apply force onto your robot.
Think about making sure your drivers know how to redirect forces applied.

[JVN]

Originally Posted by Karthik

There are a surprising number of mecanum robots that are not actually programmed to go sideways; they simply use mecanum wheels in a tank drive setup...

Just because they don't EVER move sideways, doesn't mean they're not programmed to do so. It is bad scouting to make an assumption like this...

[techtiger1]

Originally Posted by JVN

Just because they don't EVER move sideways, doesn't mean they're not programmed to do so. It is bad scouting to make an assumption like this...

*Or that the driver can use this capability effectively.

[Michael Hill]

Originally Posted by Karthik

There are a surprising number of mecanum robots that are not actually programmed to go sideways; they simply use mecanum wheels in a tank drive setup...

This blows my mind. With code available online that does the work for you, how do teams let this happen to them?

[Mark Sheridan]

Originally Posted by Michael Hill

This blows my mind. With code available online that does the work for you, how do teams let this happen to them?

Having done mecanum and octocanum successfully (or not) for about 4 seasons with 3309. I had to debug a lot of other teams's mecanum drives. The number one reason I saw was not implementing a gyro. With inadequate practice, drivers just revert tank. Some just avoid strafing, others just use the tank code instead. I think most teams I met did not know the gyro was important, the remainder could not get their gyro to work. I speculate the later was due to a lack of time for the programmers to work on their robot.
the second issue I have seen is that teams over tighten the Andy-Mark mecanums, locking the rollers preventing the strafe. this issue can very from all the rollers to just one. Even with one roller jammed, strafing is awful with every rotation of the wheel.

I had lot of fun with mecanum, its a cool off season project but I am having 3309 go back to tank. To explain I will steer this conversation back to the t-bone topic. There are defensive analogues for mecanum. 3309 and 3476 have practiced a fair bit together. 3309 ran a octocanum this year. 3476 had a west coast tank that were slightly faster than 3309.

Two moves that tanks can do against mecanum are:
1. Plowing: Essentially a T-Bone but you carry through with your hit. this was common tactic in the wedge robot era. 3476 would ram the side of 3309. 3309 would attempt to roll dodge. then its a battle of skill , 3309 would try to roll off and continue but if 3476 turns with 3309, 3476 would keep pushing or rather chasing 3309. So its essentially looks like a high speed t-bone. If 3309 stops moving, 3476 will push them or if 3476 lets off, 3309 would escape. If 3309 can't escape, eventually the robot will wither run out of room to roll dodge and get pinned into a wall (with pining rules being applied) or attempt to straighten out to shoot where the pushing force 3476 will start shoving 3309 off course. typically, 3309 would roll to the goal and thus 3476 would shove 3309 into the wall forcing 3309 to low goal score.
2. Wall defense (I know this as crossing the T): i think this is the best against mecanum. 3476 would drive back a forth perpendicular to 3309. We tried having 3309 strafing to get around but the drive was not quick enough. Mecanum could not push through, so 3309 dropping to tank mode to push 3476, it would work but wasted a lot of time. So it forced 3309 to adopt V turns and fake outs to get around. This was when 3309 decided to not do mecanum again because the bulk of the moves that would work were tank drive moves that 3476 uses. We just started running out of uses for the strafe.

I think these are the terms that are colloquial to FRC. It would be interesting if anyone know the history of these driving techniques and how they evolved over time. Personally I feel mecanum is safe from the 20 second long t-bone but they are not immune to defense. More importantly, mecanum can't excecute a t-bone without being a octocanum. finally with ballistic nylon and sail cloth bumper material options, any team can build a very t-bone resistant tank drive.

[Pat Fairbank]

Originally Posted by Michael Hill

This blows my mind. With code available online that does the work for you, how do teams let this happen to them?

I've seen worse -- teams that chain both mecanum wheels on each side to the same gearbox.