View Full Version : To Strafe or not to Strafe
I've been reading up a lot about mecanum and omni-wheels here and other forums online. Most of what I'm seeing is "X wheels are great and easy" or "rule 1- never use X-wheels again", both with sometimes one line for why.
Watching youtube videos of both, I'm seeing that they certainly get the job done, but requires a lot of skill from the driver.
I guess what I'm asking is if people could give a little more detail as to why or why not they are using mecanum or omni-wheels.
To help me search (and possible consolidate the list later with answers and credit) please list your response as:
Mech +
good reason
Mech -
bad reason
Omni +
Omin -
Mech/Omni +
Mech/Omni -
This is a link to a fantastic power point made by 1675 that answers the questions you are asking.
https://www.google.com/url?sa=t&rct=j&q=&esrc=s&source=web&cd=7&cad=rja&uact=8&ved=0CEQQFjAG&url=http%3A%2F%2Fwww.chiefdelphi.com%2Fmedia%2Fpap ers%2Fdownload%2F2878&ei=X4muVPTbA4mjyQT09YLIDw&usg=AFQjCNEXVJgXzSxZrEKUG3-fitLaeCXIcw&bvm=bv.83134100,d.aWw
theCADguy
08-01-2015, 08:58
Keep in mind that a lot of CD posts warning against mecanum and omni are based on previous games, where the lower pushing power of these drives allows the to be susceptible to defense. However, this year, change has come, and omni and mecanum will probably be much more common.
nighterfighter
08-01-2015, 09:03
The best answer to the poll, would be to select ALL the answers.
They can all be valid answers, but it depends on how you implement it, if you have previous experience with it, and how much time you want to put into designing, programming, and practicing, with your drive system.
Would that time be better spent on a game piece manipulator? Maybe. That's your decision. Just remember that for every hour you spend tweaking one subsystem, that's an hour taken away from something else.
Chief Hedgehog
08-01-2015, 09:12
@Fields:
You have one of the best teams (in terms of Mecanum) right in your backyard - FRC 2052 Knightkrawler (Irondale). You may want to reach out the them and Ask what to expect.
FRC 4607 CIS will not attempt Mecanum this season - but from what we have seen from slide (omni's in a H-config) we most likely will be working with that.
Why a need to strafe I feel if your team is comfortable and know that your crab, omni, mech wheel combo will work but the bump is so low there is no reason to dodge them you can get across the field much faster if you traverse the bumps well you really have nothing to out maneuver
If you chose to use the "H" or slide drive wont that middle wheel get high centered on the bump?
Lewis Nerone
08-01-2015, 09:55
4269 plans to use a mecanum drive again this year. We developed a system that utilizes a gyroscope to track the robot's orientation on the field. When the drivers push the joystick, the robot will move in that direction, regardless of it's orientation.
If you do go mecanum, you need to start working on it now. I'm not a programmer, but the code is very complex. You also need to make sure that your chassis is designed to run mecanum (wheels must be in a perfect square, neutral camber, etc.). These were challenges last year.
Good luck!
MrJohnston
08-01-2015, 10:18
We are sticking with what we already know we can do very well: a tank-drive system. We believe that a mecanum system is probably better for this year's game, but having not done one for a long time, we would introduced a lot of unknkowns into our build season and potentially create a very real time crunch while trying to perfect our manipulators.
We can make up for many of the advantages of a mecanum drive through some create creative programming for a tanks system (we know we can do this) and by making sure that our tote-acquisition-manipulator has a way to make lateral positioning adjustments....
daniel12997
08-01-2015, 12:58
the programming libraries have built in functions for mechanum and because of this my teams programmers believe it will not be very hard.
MrJohnston
08-01-2015, 13:07
It all depends on the complexity and necessary precision of the task you are attempting to program. :)
We miscalculated on the amount of defense last year, and went mecanum. Our programmers got mecanum working from the libraries in about two weeks from a cold start. We're going omni H-drive this year - we're doing a "bottom stacker", and feel that a bit of strafing is going to be essential to quickly and precisely line up on the totes. We decided to pass on mecanum this year because of the issues with a moving center of gravity.
We're going omni H-drive this year ... We decided to pass on mecanum this year because of the issues with a moving center of gravity.
When you experimented with H drive in the off-season, did you find that it was less affected by variations in center of gravity than mecanum is?
alopex_rex
08-01-2015, 19:00
We can make up for many of the advantages of a mecanum drive through some create creative programming for a tanks system (we know we can do this)
What kind of creative programming are you going to use? In my experience with tank drive I haven't done anything more fancy than limiting the rate of acceleration, I'm curious as to what can be done.
To Strafe, w/ an Omni Crawler Drive though. This is the year to do it. The year without real defense and no bumpers required or necessary. If you use the triangular 1 it can tip over and still keep going....and going....
https://www.youtube.com/watch?v=BTp2UAaihaI
Hee, hee...Then there's the Water Game! (Ok, that water is slightly frozen...But the 2016 Game could be held in outdoor colder locations and the field just a wee bit larger and colder is all). Posted to try and give the GDC ideas to give the fans what they have always prayed for...A Water Game! Using the 1929 Fordson Snow Machine concept of course. (There has to be a way we could tweak that so it wouldn't turn the carpets into huge balls of yarn). Any ideas?
https://www.youtube.com/watch?v=zBjlSJf4274
So what happens when the field melts in those warmer regions? Still a water game. (Only wetter and muddier).
https://www.youtube.com/watch?v=H5NsmZcLAdM
"Guardrails can't hold us!"....I know how I'm getting to the ski slopes from now on....Anyone still have a 1929 Ford around that's still running or in decent shape? Lol. I'm going to make me 1 of those Fordson Snow Machines.
Caleb Sykes
08-01-2015, 22:27
You also need to make sure that your chassis is designed to run mecanum (wheels must be in a perfect square, neutral camber, etc.). These were challenges last year.
Why do you think that a square chassis is required to run mecanum?
JorgeReyes
08-01-2015, 22:39
Has anyone had any luck with having two traction wheels in the front and omni wheels in the back? This would allow you to pivot at the front rather than the center which makes getting totes easier
jijiglobe
08-01-2015, 22:49
Has anyone had any luck with having two traction wheels in the front and omni wheels in the back? This would allow you to pivot at the front rather than the center which makes getting totes easier
We did it in 2008 and it worked fine but you need to make sure that turning around the ends of your robot is what you really want.
I found the kitbot for that crane we have been looking for, and it Strafes nice at 2:14 of the video.:D
https://www.youtube.com/watch?v=0nWFA8wv0-I
MrJohnston
09-01-2015, 02:44
What kind of creative programming are you going to use? In my experience with tank drive I haven't done anything more fancy than limiting the rate of acceleration, I'm curious as to what can be done.
Tank allows for fairly precise movements with pre-programmed routines.... It's one of the reasons I'm certain we'll have a strong autonomous....
Abhishek R
09-01-2015, 08:39
Tank allows for fairly precise movements with pre-programmed routines.... It's one of the reasons I'm certain we'll have a strong autonomous....
What is it about tank that lets it accomplish tasks more precisely than other drive systems?
MrJohnston
09-01-2015, 16:10
With good traction tires on a tanks system, there is very little slippage, allowing you to program precise numbers of rotations, etc.... I am not a programmer, so I really can't give much detail. I do know that as we were discussing the pros and cons of different drive systems, that our programmers were very clear that they could have more precise autonomous routines with the tank system. In the past, I've seen them consistently get our robots to autonomous go to an exact location (within an inch, or two) repeatedly.
Perhaps we simply aren't experienced enough in mecanum - Okay, I know we aren't. However, I think very highly of our programming team - both students and mentors - and tend to trust them when they say, "We can program much more precise movement with this drive than with that."
Physicguy
09-01-2015, 16:25
Perhaps we simply aren't experienced enough in mecanum - Okay, I know we aren't. However, I think very highly of our programming team - both students and mentors - and tend to trust them when they say, "We can program much more precise movement with this drive than with that."
This is true in past events tank drive system that have been used has been really successful in competition. We are at the same crossroads on whether or not to go with an omni-directional drive or to just stick with tank. We had our fair share of trouble with mecanums before.
With good traction tires on a tanks system, there is very little slippage
When turning there is a lot of slippage. That's why it's called skidsteer.
Abhishek R
09-01-2015, 16:48
With good traction tires on a tanks system, there is very little slippage, allowing you to program precise numbers of rotations, etc.... I am not a programmer, so I really can't give much detail. I do know that as we were discussing the pros and cons of different drive systems, that our programmers were very clear that they could have more precise autonomous routines with the tank system. In the past, I've seen them consistently get our robots to autonomous go to an exact location (within an inch, or two) repeatedly.
Perhaps we simply aren't experienced enough in mecanum - Okay, I know we aren't. However, I think very highly of our programming team - both students and mentors - and tend to trust them when they say, "We can program much more precise movement with this drive than with that."
I would encourage you to try omni's in the offseason if you have already ruled them out for this year. I don't have experience with mecanum, but we used omni's as our main drive system on our 2014 robot. Our autonomous worked pretty well consistently; we didn't have many accuracy issues.
We, Team 4276, tested pushing totes on carpet and up ramps using our 2014 drive: 6" AM Mecanums with 10.7:1 TBMini's.
We pushed 4 totes with bin and we're happy with the results. We drove the robot on top of poly sheets (not HDPE, sticker missing so not sure what it was exactly) and there was little difference in ability to push totes.
That coupled with the ease of strafing and rotation has led us to keep Mecanum for this year. We'll be upping the gear ratios though.
MrJohnston
09-01-2015, 18:52
I would encourage you to try omni's in the offseason if you have already ruled them out for this year. I don't have experience with mecanum, but we used omni's as our main drive system on our 2014 robot. Our autonomous worked pretty well consistently; we didn't have many accuracy issues.
We are looking at some possibilities with omni wheels... The increased ability to turn (over a tank system) is appealing.
We miscalculated on the amount of defense last year, and went mecanum. Our programmers got mecanum working from the libraries in about two weeks from a cold start. We're going omni H-drive this year - we're doing a "bottom stacker", and feel that a bit of strafing is going to be essential to quickly and precisely line up on the totes. We decided to pass on mecanum this year because of the issues with a moving center of gravity.
When you experimented with H drive in the off-season, did you find that it was less affected by variations in center of gravity than mecanum is?
Did I say we experimented with H drive in the off season? Actually, we built a tank drive for a prototype chassis because we recognized that it was the best chassis for 90% of the games for teams without massive funding or the ability to crank out a swerve drive in the shop (like us).
Anyway, we do not expect a major variation due to (front-to-back) variation in the COG on an H-drive. We're going to put the COG at the transverse wheel when we have about a 75% load. The short moment arm to this wheel should keep the torque about the COG small when we have 3-6 totes. We'll have to learn to adjust at lighter loads, when it's less critical. I also can't think of a reason why, in a tank-mode spin, an H-drive wouldn't rotate around its COG - lateral forces balance about the new COG, leaving only torque.
Actually, when I later thought about the forces on a mecanum with a COG that moves forward and in reverse, I became a little less concerned, but still glad that we switched. Moving forward and back would be largely unaffected by moving the COG forward or back. Strafing or translation with a strafing component introduces a torque so that the robot will rotate towards the strafe direction. On an attempt to simply rotate, the robot will tend to strafe in the direction of rotation on the side to which the COG moved. Also, the axis of rotation seems to move to in the opposite direction of the COG, which sounds likely to cause the robot to fall on its face eventually. I did not think about combination maneuvers involving both translation and rotation. Disclosure: this is all based on pictures in my head, not on paper or even a whiteboard, and certainly not experience.
Did I say we experimented with H drive in the off season?
Reading between the lines. It sounded like you had some supporting test data you might be willing to share.
We are looking at some possibilities with omni wheels... The increased ability to turn (over a tank system) is appealing.
Last year we used an 8 wheel tank drive with 4 traction wheels in the center and 4 omni wheels on the outside. It worked very well and was one of the smoothest drive trains i have ever driven, it was just incredible:D .
JorgeReyes
11-01-2015, 01:07
Last year we used an 8 wheel tank drive with 4 traction wheels in the center and 4 omni wheels on the outside. It worked very well and was one of the smoothest drive trains i have ever driven, it was just incredible:D .
Are you going to use this again this year and if so, what size wheels? We might do something like this
Are you going to use this again this year and if so, what size wheels? We might do something like this
We are doing a 6 inch mucnum drive this year. But monday night i can post some pics of last years chassis if i get a chance.
We used 4 inch wheels.
JorgeReyes
11-01-2015, 14:07
I don't think that 4 in wheels will clear the scoring platform. I'm guessing your planning on driving around them then?
I don't think that 4 in wheels will clear the scoring platform. I'm guessing your planning on driving around them then?
That was last years chassis. This year we are using 6 inch mecnum.
Lewis Nerone
11-01-2015, 18:40
I don't think that 4 in wheels will clear the scoring platform. I'm guessing your planning on driving around them then?
Correct, no 4" wheel will be able to clear the scoring platform. A 6" wheel is going to be your best option.
Thad House
11-01-2015, 18:53
Correct, no 4" wheel will be able to clear the scoring platform. A 6" wheel is going to be your best option.
This is not right. 4" wheels will absolutely be able to clear the platform. But it will require a little bit of creativity to do so.
smclean1969
11-01-2015, 20:09
We used mecanum wheels in Ultimate Ascent. They were great on the mobility aspect but had some significant drawbacks. Here are the drawbacks: they slip when they initially start up. This makes autonomous very challenging because it's imprecise during the initial movement. It has very poor pushing power so it's not good for defense or beating a defensive robot.
So, for this year, there's no defense so that part of the game wouldn't be applicable.
We used omni wheels on Logomotion, our first year. They made the robot highly mobile which was great for our ability to rank 7 / 65 teams that year--that's our beast seed ever so I can attest they can work very well. The downside is limited friction so again, they're not very effective for defense or beating a defensive robot. So, that's not applicable in this year's game.
Here are the drawbacks: they slip when they initially start up. This makes autonomous very challenging because it's imprecise during the initial movement.
Is your team familiar with command rate-limiting?
Correct, no 4" wheel will be able to clear the scoring platform. A 6" wheel is going to be your best option.
I guess we didn't read this...we are going over just fine with them ;)
I guess we didn't read this...we are going over just fine with them ;)
Were is the hub of the wheel in relation to the bottom of your chassis?
Correct, no 4" wheel will be able to clear the scoring platform.
Did you mean no 4" mecanum? (https://www.youtube.com/watch?v=2ZbCseo-_ks)
Were is the hub of the wheel in relation to the bottom of your chassis?
I guess would could say that one set of our wheels is centered on our chassis, there is no drop. That is as much as I will say until Reveal night :rolleyes:
Are you going to use this again this year and if so, what size wheels? We might do something like this
Here are the promised pics of our old drive train. It uses for inch wheels 4sim drive and chain for the drive transfer. Overall it is a great driving chassis, one of the best i have ever driven.
vBulletin® v3.6.4, Copyright ©2000-2017, Jelsoft Enterprises Ltd.