Re: Swerve Drive vs Mecanum Wheel drive?
 

Re: Swerve Drive vs Mecanum Wheel drive?
 

I've offered my views on both systems vs tank drive despite having intimate experience with each and was told I was "wrong". So you guys are on your own here. I can't be bothered to argue with people on the internet anymore.

Everything posted on an internet forum should be treated as opinion. Some has more value than others based on the credentials of the individual posting. Typically, what users define as "fact" tends to be what most people agree on but that doesn't mean it's correct. The OP should realize this and when sifting through these discussions, use the statements brought forth to form their own conclusions.

Re: Swerve Drive vs Mecanum Wheel drive?
 

Well what's the point of comparing a good tank drive to a bad mecanum drive? Of course the better drivetrain will always come out on top. That's not even a question.

Nearly half the people in this thread have claimed that. My post is not dismissing mecanum as bad, it's pointing out the truth that mecanum drivetrains are not good for pushing which multiple people in this thread claimed otherwise.

My bad, I should have been more clear in my first post. The point I was trying to get across is that both a good, and a bad mecanum drive will lose pushing matches against a well built tank or swerve drive. Re-reading my post I realize that was probably not the way it was interpreted.

I will agree that there absolutely are differences in the pushing power between the best and the worst mecanum drives. However, I also assumed from the beginning that we were comparing drivetrains that used the same COTS parts because it's pretty obvious that two drivetrains made using different COTS parts will be different. I figured that was a given since it really applies to every drivetrain, not just mecanum.

Re: Swerve Drive vs Mecanum Wheel drive?
 

I was not going to get involved with this thread, but because my team has done both, I figured I should probably share my opinion.

In 2012, at about week 3, my team decided to use the wild swerve module from AndyMark. It was quite a bit of work to implement, and the performance was lackluster. But ultimately, it was part of why teams 1717 and 469 picked us for their alliance on Newton. So in hindsight, it was probably worth it.

In 2013, my team tried to do a custom independent swerve drive, but was not prepared enough so we ended up changing to mecanum before it was too late. I was way more disappointed with the performance. We barely had any ability to strafe because the rollers had too much friction, our top speed and pushing force were diminished by the inefficiency of each gearbox, and the drive train was heavy because we used gearboxes with steel gears that were made to take two cims.

This year we made the jump to custom independent swerves. I could barely be any happier with a drive train. The swerves are light (about 7 lbs each), somewhat affordable ($220 each), reliable (knock on wood because we've had no failures yet), and highly maneuverable. I will admit that they were a lot of work, but I truthfully believe that they did not have any negative effects on the rest of the robot.

In conclusion, I would strongly recommend that any team with CNC machining capabilities spend their off-season research time on swerve drive rather than mecamum. I realize that many people disagree with me on this, but this is my honest opinion.:D

Re: Swerve Drive vs Mecanum Wheel drive?
 

Our team is in its 11th year (I think). We've used swerve for 2 years, mecanum for 3 years and tank drive the other years.

Our mecanum was very well done, almost as well as it can be done w/o adding locking wheels or brakes of some kind. We had no trouble with the average tank drive opponent in qualifying matches. We just spun off them or juked them - often it was embarrassing. But late in eliminations rounds tank drive robots with good drivers assigned to specifically stop us could really slow us down. But we stuck with mecanum for a goofy reason I reckon, the students loved to drive the thing. It is a lot of fun.

Our swerve drive worked pretty well the first year (lunacy) but not so well the next. We had some steering problems but it was a design issue (windows motors, not enough torque etc). We fixed it eventually. We loved swerve also - it is also fun to drive. But swerve involves a lot of relatively complex machining. I disagree that the software is complex - the math is pretty straight forward. We stopped doing swerve because one of our mentors (and the best machinist) got sick the next year.

For the last several years we've tried chain and belt driven single and two-speed tank drives. They work well and are relatively simple to design and build. It takes us a single afternoon to hammer out the design. It is very strong and reliable (except for changing out the occasional chain or belt).

So what advice can I give you? Do NOT try swerve w/o a foolproof off-season design working. You could kill a build season schedule (for most teams). If you have a good swerve you will be more maneuvrable than a tank drive. Is it worth the weight and complexity? ... depends on the game. A properly geared two-speed 6-motor tank will still push you around a little (not like a mecanum though) because you will probably be limited to 4 motors and a single gear.

Finally persons on CD and in competitions are too hard on mecanum drives. They can be better than most people think. And as Ether points out (in dozens of other threads) the mathematics and operation behind mecanum is almost universally misunderstood. For example ours could park on a bridge (but not push someone else up the bridge), fly over bumps and play passable defense against tanks (if we caught them on the corner and spun them around rather than push them). But the prejudice against mecanum is not w/o some basis and there are MANY teams that will NEVER pick a mecanum alliance partner.

HTH

Re: Swerve Drive vs Mecanum Wheel drive?
 

As a Team who has had both a swerve drive and a mecanum drive. I love our swerve drive. In our teams history we have built 2 mecanum drives 2 tank drive's(if lunancy counts) 1 crab drive(failed so bad) and 1 swerve(2014).

In 2010 my team attempted a crab drive(they call it a swerve). At that time my team did not use cad. The drive was huge 80/90 pounds and did not work... ended up trying to keep it driving straight with zip ties....

In 2011 we did the first mecanum drive. This was the best performing robot up to this point we used the standard AM frame. When using it we never had problems with being pushed. It was very slow though much slower then i would have liked and the drive frame was heavy.(i think our team just builds heavy drives) But it was still a good drive.


2013 We went back to our mecanum drive this time using lighter gear box's. this drive could not strafe very well at all. The rollers had way to much friction. We got pushed around a lot. Even though it was geared on the high end it never reached that speed. Could not drive straight at all. We got it to work in the end but not nearly as well as in 2011

2014 We built a 4 wheel independent swerve.

My team has never had very good metal working tools. We had a plazma cam that we used a lot but when our metal shop teacher passed the new one wanted nothing to do with us. Losing us all of our metal working tools. We do have a very small CNC that breaks a lot. (it was on a total of 30 minutes the whole year). but one of our mentors has access to a cnc shop that we can use.(limited use) The day of kick off we lost our best mechanical student and our coach.
But that did not stop us. We had to travel to Portland Community Collage they have mills and laths we can use their (all manual). Me and my sister managed to manual mill/lath the swerve in 4 weeks. The first week was updating the design. We then pulled 3 weeks of going to school/machining as our whole life. We used a mentors CNC shop for some parts that their were to many of to manually make or to tight on tolerances for a total of less then 5 hours of CNC work.
If their is a want their is a way.

The performance of the swerve drive was a dream. It is the fastest drive we have ever built. It can also plow our mecanum drive our of the way in static tests in the shop.(on official carpet) It was very powerful. It was also 4 pounds lighter then last years mecanum drive and 10 lighter then 2011. With each modular weighing in at 7 pounds 4-8 ounces. The swerve drive out performed every drive we have ever built up to date.

Half of the reason i am in favor of the swerve drive over mecanum is that it presents real engineering problems. The felling our team got when it was driving for the first time was truly amazing. We got our first engineering awards this year because of it. It was a lot of work but it was worth it even if we did not win.

Re: Swerve Drive vs Mecanum Wheel drive?
 

After reading through most of this thread, I am posting ONLY in regard to the question of whether the performance level of a poorly designed/build mecanum drive is very much different from an excellently designed and built mecanum drive.

My teams mecanum drive experience is limited to this year's 2014 FTC competition (we do both FRC: 3135 & FTC: 3507). After evaluating VEX 4" mecanum wheel, we made a last minute decision to switch to mecanum midway through our FTC bot tank build (but completed the tank too in case mecanum was a bust).

The PICS below show the basic concept of our design. This was a more robust implementation of a mecanum drive than what your typical FTC mecanum efforts are like. Twice when a slightly errant autonomous route had us collide with the scoring buckets' balance beam, we became wedged under it, going back on our rear wheels only, and totally lifting one entire side of the hanging bridge, while doing a wheelie, as we pushet it across the field, yet no damage to drive train resulted.

We were not pushed around very much by tank drive bots and could play decent defense. Granted that at FTC level, not as much drive train differences can be seen between drive type designs (compared to at FRC level). However, the point I make is that, compared to the other mecanum drive bots we saw, there was a lot of difference between a good design and a weak design.

Our "H" frame had just enough flex to help keep all wheels down most of the time, yet stiff enough to deal with high stress situations. We were able to position accurately on the sloped bridge and were stable once locations were reached. The soft urethane of VEX wheels was very grippy and gave excellent traction, better than Andy Mark 4" mecanum wheels. Our weight was fairly evenly balanced. We won a design innovation award at our 2nd regional qualifier and were first seeded team captaining the winning alliance at 1st regional. We were 2nd seed in state after qualifying rounds.

Bottom line is there was a significant range of performance difference between the weaker mecanum designs/builds of bots (not very many though) compared to ours. We plan to add field oriented gyro control, improve our speed, and do an independent wheel suspension scheme for our next year's FTC mecanum build, unless the game design makes mecanum an inappropriate option. Our first FRC mecanum drive could also be happening later in 2014.





-Dick Ledford

Re: Swerve Drive vs Mecanum Wheel drive?
 

It's been a while, but I'll reply anyway...

While it's true that you can make improvements like these, they are more design solutions and optimizations rather than conceptual changes. A mecanum wheel, no matter how well deisgned, will still be just that: a mecanum wheel. Rollers tilted at 45* angles on a wheel. There can't be any kind of mecanum revolution unless somebody figures out how to lock the rollers without wasting weight.
This is not a bad thing, as optimization is a very good thing and makes good ideas better, but it does show inherent limitations in mecanums. Other kinds of drives can be done in many different ways by comparison.

Re: Swerve Drive vs Mecanum Wheel drive?
 

I haven't used both but I say choose swerve over mecanum. Mecanums don't have as much power and most teams write them automatically and will be hesitant when choosing a team with mecanum drive train for an alliance. However, don't listen to those people who that say mecanums are horrible. We've used them before in the past and their not that bad. They are much easier much to build and program than swerve.

Our team is currently going through a similar debate. We think a swerve drive is great but are unsure if we can get one working reliably over the summer. In the event that we can't, we will be trying an octacanum drive train. For those of you who don't know, with an octacanum drive train you can toggle between high grip wheels and mecanum wheels. Secret city wildbots (4265) built a fabulous octacanum chassis this year. I would possibly look up an octacanum drivetrain because you can get the pushing power of a standard drive with the maneuverability of an omnidirectional drivetrain like a swerve or mecanum. Secret city wildbots (4265) can be found on YouTube under "dragonfly reveal".