rotation of body of crab drive robots
 

I'm confused as to how a crab style robot rotates itself entirely (not just wheel modules), particularly coaxial ones in which all the wheel modules face the same angle and are powered by the same gearbox. I see on the wildstang video page that they have a little foot that comes down to pivot on, and I can see how that could work with a crab like the one i'm describing, but it seems like a large turning radius and the necessity of having to rotate the modules to a certain angle and put down a foot would kinda counteract the whole ultimate mobility thing that crab drives have going for them. But then I think wildstang won last year so it's got to work fairly well.
...
...
...
Basically what I'm asking is how crabs turn and how well it works.

Michael,
First off, I don't think I've seen a crab system where every wheel is driven by the same gearbox. That type of setup would only allow for translational steering. The crabs powered by a coaxial system that I've seen have had one gearbox per side of the robot. This allows the robots to reverse the wheels in either gearbox to employ tank steering.

Wildstang's robot, like The Sea Dawgs' this past year, doesn't utilize a coaxially driven crab system. There are four drive motors, one mounted inside each crab module. This setup enables the robot to use tank steering by driving the 2 motors on one side of the robot in one direction, and the other 2 drive motors from the opposite side of the robot the reverse direction.

A robot like Wildstang's and The Sea Dawgs' can also use complimentary steering where, for example, one set of the modules points a certain degree to the right and the other set of modules points a certain degree to the left. This is just another way of turning. The Sea Dawgs used this method in their autonomous mode at the California Robot Games back in September.

If you want to see a crab robot in real life feel free to check out The Sea Dawgs' 2003 robot. You, and anyone else, are welcome to drop by during one of our meetings. If you can't do that then our robot will be on display at the November and December WRRF Motors/Drivetrain classes. Members of 258, including myself, will be there to answer and questions you might have about the drive system.

I'd say that crab steering is great, but it's definitely not a huge advantage. I think 2000 was the last game where having a swerve made a huge difference in the success of a robot. Although, now that I think about it, I can remember 5-7 times (over my team’s 2 event season) where we made it on top of the plateau with 1-2 seconds left in matches where we wouldn’t have been able to without our crab drive system. It all depends on how the game is laid out, and what strategy you want your robot to be designed for.

for body rotation in a coaxial system 2 gearboxes one on the left and one on the right, you make a switch in the program while the wheels are locked in the foreword position and you can rotate and steer as a 4x4 bot, Or with 2 steer linkages one fir the front and one for the rear, you can get some quadrasteer action by throwing the wheels in equal opposite directions Problem with both of thoes you would need some arakeen correction because the wheels would have to slip to the sides.

The Big advantage to the Crab is that you can apply 100% power in any direction and through turns, where something like a tank or 2 wheel drive can only apply 100% power in the direction its pointed and to turn you would either have to reduce power to one side or reverse one side. A kiwi or OMNI can ONLY APPLY up to 50%in a 4 wheel system and maybe 33% in a 3 wheel system of their power in any given direction.

In my design for a crab drive system, There are only 2 drive motors, each supplies forward/reverse power to the left/right side, and the left/right sets have their own pivot motor. This allows for Skid Steering, Crab Drive, Car steering, and "Angle" drive.
I haven't posted my design yet because I'm waiting for some software, and I don't want to give it all away too soon...
Scan the threads for some of my posts though, I've posted more info on my system in various threads.

Not true.

4 wheel omni can apply anywhere from 50% to 100% of the available motor power and a killough platform can apply 66% of the available (all 4 motors) motor power.

The only advantage to crab is not having to use omni-wheels to get omnidrectional-like maneuverability, and it allows one to direct 100% of the total available motor power all the time, permitting that the crab modules are functioning and can point the wheels in that direction.

Um... So it is true

Re: rotation of body of crab drive robots
 

Michael,
A crab drive system does not need to steer to drive. You just point the wheels in the direction you want to go and apply power. Wildstang used the foot this year so as to provide an easy way to reorient the robot to pick up bins or to align with the ramp. There are actually two feet, used one at a time, to achieve the desired effect. The foot raised two wheels off the ground while driving the remaining two in opposite directions. This is not an efficient method but it works when needed. We have used steering in past years that allowed for a different turning radius on one side of the robot than the other. This would give an arc to the robot motion and hence a turn. Max power is still when all the wheels meet the road.