Recent Kit Base Performance

[fargus111111111]

Often the kit base gets a bad rap for being inadequate and this year without modification it was, but I want to talk about how they perform in more average FRC conditions aka a flat carpeted floor. In 2014 and 2015 my team used a 4 CIM kit base and did relatively well, in fact if you ask anyone on the team they will probably say that those were some of our best robots in recent history (last 6 years). I would credit much of their success to how quickly they were running for drive practice and testing. With this in mind I want to ask 2 questions.

1. Does your team frequently use the kit base?

2 a. If yes to q1. Do you modify it and if so how?
2 b. If no to q1. How do you reliably develop a drive base to be quickly usable for on field testing, have you formed your own kind of "kit base" that is the origins of each year's design?

I ask because I am interested in encouraging my team to work on developing other types of drive trains in the off season but I'm not sure if it is worth it because of our very good experiences with the kit base.

[GreyingJay]

I think you'll find that a lot of teams use the kit base. I've even heard of teams that certainly have the engineering ability to build their own custom base, just go with the kit base because of the reasons you have listed. Why reinvent the wheel(base)?

We got the AM14U3 and bought an AM14U2 kit at half price during one of AndyMark's sales. These were up and rolling within days of kickoff for prototyping and for software teams to start driving. We quickly realized that we wanted to upgrade the wheels to 8" pneumatics. Other than that and a change of frame configuration, it was pretty much as-is.

We do want to prototype other types of drive systems, but unless we hit a game that really needs it, I expect we'll keep using the kit and modifying it to suit.

[aciarniello]

We have used the kit chassis with great success for many years, but we usually modify it in some way. Here's our history:

2012--(rookie year) completely standard kit chassis, but with larger wheels
2013--Standard kit with super shifters
2014--Standard kit with sonic shifters and omni wheels on one end.
2015--Standard kit with cut out in front for tote entrance and omni wheels on one end.
2016--Fully custom laser cut chassis with actuated suspension

We will continue to evaluate the kit chassis as a viable option every season, since it has some significant advantages:
1) FREE (the first one, anyway)
2) Lots of COTS ways to modify it
3) Uses fewer engineering resources
4) Lots of support in the community

Regardless of the advantages of the kit chassis, I believe that you should encourage your team to develop alternate drive trains. Most importantly, the engineering exercise is great experience for the team members. It also gives another option to evaluate when doing your game analysis. Even if it does not get selected 9/10 seasons, that 10th season, it could be the difference between winning and losing!

The biggest advantage that I found with a custom chassis, though, was the flexibility it gave us in mounting our attachments and gearboxes. It allowed us to really design for maintenance and repair and to package things the way we wanted. This is a benefit that you'll never see from a kit chassis.

[Jon Stratis]

We've used the Kit chassis every year except for this year and Lunacy. With Lunacy, we really don't have an excuse, other than wanting to do our own drive train. This year, we had opted out ahead of time, spent the fall exploring some other drive trains, and were very thankful for our decision and our custom chassis

We have modified the kit chassis before. In 2014 we switched the gearboxes out for toughbox nanos and ran mecanum drive. Other years, we've cut the front and built a super structure to hold it together in order to allow ball intake. But that's really it.

For our custom chassis this year, we did direct drive from toughbox nano's, and welded the nano's to the bottom of the chassis. It was amazingly effective, and set the chassis up high enough that getting stuck was never really going to be an issue. The whole design came out on kickoff day, prior to that we had expected to go with the nano tubes to form the basis of our design.

[D.Allred]

Originally Posted by fargus111111111

1. Does your team frequently use the kit base?

2 a. If yes to q1. Do you modify it and if so how?
2 b. If no to q1. How do you reliably develop a drive base to be quickly usable for on field testing, have you formed your own kind of "kit base" that is the origins of each year's design?

I ask because I am interested in encouraging my team to work on developing other types of drive trains in the off season but I'm not sure if it is worth it because of our very good experiences with the kit base.

You'll have to define "other types of drive trains." There is a big leap between kit bot and 4 wheel independent steering.

Here's our approach.

We have always used skid steer instead of investing resources in omni-directional drive trains. I don't see that changing for 2017. We may switch to the kit bot or VexPro Versaframe to speed up construction.

We used kit base components for our 2013 robot, but had to build our own frame for packaging reasons. Every other year we have built a west coast style skid steer drive train. We do this primarily because we can, not because it is better than a kit base.

In either case, there really is no drive train "development" for us. A west coast style drive train is part of the tube and gusset construction technique we use for the base and scoring section frame. Since we have the milling machines to manufacture our own tubes, it is very easy to customize the wheel spacing and frame dimensions to suit the game. It did take time to develop bearing mounts and gearbox mounting systems. There are now COTS solutions readily available for most of these needs.

David

[tr6scott]

We evaluate it every year, just like the simbotics say to..
We have a old school kit bot on steroids that is a programming mule, test bed.

2014 we ran the new style kit bot, with two speed trans, all the way to 4th rank at Archimedes. Bot is still in use today, with no modifications, we have had to replace the front face twice, due to hits...

https://www.youtube.com/watch?v=hsl-47EOqN4

[GeeTwo]

In our five years, we've modified the KoP drive train three times, and built our own twice - not in that order.

In 2012 (Rebound Rumble, rookie year), we upgraded the KoP C-channel chassis with shifting gearboxes. We were unhappy with the rigidity of the chassis, but this was all on us - we cut open one side to do a ball intake and did not adequately support the breach.

In 2013 (Ultimate Ascent), we had a mentor with aluminum welding capability, so we opted out of the kit chassis did our own. We were happy with the chassis, but the time required to move the chassis back and forth for each change ate up far too much of build season. We also made some unfortunate gearing decisions, gearing too fast with only about 2/3 of the weight of the robot on driven wheels.

In 2014 (Aerial Assault), we again opted out of the kit chassis, and built our own C-channel chassis. We were happy with the results. This was the first time we'd bagged a 90%+ functional robot. That year, we also built a full-up prototype in wood, so we had a chance to do a bit of drive and manipulator practice after bag and tag.

In 2015 (Recycle Rush), we were going to opt out of the kit chassis, but missed the deadline. After we had our basic design (5 omni wheel H-drive), we realized that we could build it faster using the kit chassis than a home build, so we bought a second AM14U2 so we could have a practice robot as well as a competition robot. We had quite a few hours of drive practice with the "fully functional" competition robot before bagging, and pulled only the pickup "rake" to continue practicing and development, as our two rakes were far from identical. We were very happy with the chassis (though we had some issues getting the correct weight load on the center wheel).

In 2016 (Stronghold), we opted out of the AM14U3, but purchased two AM14U2's before the season, realizing that almost any design we were likely to build could be adapted from the KoP chassis faster than we could do it ourselves. We had rough plans for over a dozen re-configurations ready for tweaking into our specific needs. We wound up designing a wedge practically identical to the AM wedge, so rather than build it, we bought it. We had ten wheels in four planes, and our only difficulty with the defenses was when we lost or broke a chain. About all we used of the KoP chassis were the main chassis components, the gearboxes, and a few fasteners. We needed different wheels, axles, and more.

For 2017 (TBD), we took advantage of a recent Tuesday deal at Andy Mark and purchased two "frame only" AM14U2 chassis sets at about half price. They're tucked away on a top shelf in the workshop waiting for January. For those who are interested, this consists of only the six large pieces of aluminum (as of this writing, they're available while stock remains).

The bottom line is that the only problems we've had with the KoP chassis in our three years of using it were related to the modifications we made. It saves significant time over the chassis we've built ourselves, so unless we gain significant fabrication capability, we'll probably continue to use (or modify) the kit chassis for the foreseeable future.

[Sperkowsky]

We have used the kit bot for all the years in our teams history. Next year is the first year we won't be using it. It's not really for the slight competitive advantage I think a custom chassis provides but more for learning. We also had a big issue this year where we were one of the first teams to get the pneumatic upgrade kit for the am14u3 and it had no instructions or exploded drawings. I ended up writing an entire manual myself. It was also really expensive $350 Iirc. In the end going custom this year would have been cheaper and with something like a versaframe we probably would have had it up and running quicker. The AndyMark voucher is worth it for us to abandon the kit bot and go for some of vex offerings.

[fargus111111111]

Originally Posted by D.Allred

You'll have to define "other types of drive trains." There is a big leap between kit bot and 4 wheel independent steering.

Here's our approach.

We have always used skid steer instead of investing resources in omni-directional drive trains. I don't see that changing for 2017. We may switch to the kit bot or VexPro Versaframe to speed up construction.

We used kit base components for our 2013 robot, but had to build our own frame for packaging reasons. Every other year we have built a west coast style skid steer drive train. We do this primarily because we can, not because it is better than a kit base.

In either case, there really is no drive train "development" for us. A west coast style drive train is part of the tube and gusset construction technique we use for the base and scoring section frame. Since we have the milling machines to manufacture our own tubes, it is very easy to customize the wheel spacing and frame dimensions to suit the game. It did take time to develop bearing mounts and gearbox mounting systems. There are now COTS solutions readily available for most of these needs.

David

By "other" I mean some sort of omni-directional drive train as a simple skid steer is something we already know how to do and we can purchase and be running very quickly. We have attempted swerve twice in the past, the first time we were very successful, 2009. The second time, we don't talk about that robot much, 2013, the swerve system was gone by our second regional. I therefore have been exploring that path along with the octo-canum and butterfly paths.

The lack of a drive train development time is an issue I see on my team, as much as this would be an exercise on how to build another drive it is more an exercise in what other options are out there and how to go about choosing the best one. We have a lot of new students who have no idea what the difference is between a mecanum, omni, and traction wheel and they are even more clueless when presented with the question, "what is a (insert your favorite drive style here) drive?" This lack of knowledge often results in us building another skid steer purely for ease and skimming through an important part of the early development of our robot. (not that skids are bad, I certainly like them I just see many competitive advantages to being able to move in any direction)

[D.Allred]

Originally Posted by fargus111111111

By "other" I mean some sort of omni-directional drive train as a simple skid steer is something we already know how to do and we can purchase and be running very quickly. We have attempted swerve twice in the past, the first time we were very successful, 2009. The second time, we don't talk about that robot much, 2013, the swerve system was gone by our second regional. I therefore have been exploring that path along with the octo-canum and butterfly paths.

The lack of a drive train development time is an issue I see on my team, as much as this would be an exercise on how to build another drive it is more an exercise in what other options are out there and how to go about choosing the best one. We have a lot of new students who have no idea what the difference is between a mecanum, omni, and traction wheel and they are even more clueless when presented with the question, "what is a (insert your favorite drive style here) drive?" This lack of knowledge often results in us building another skid steer purely for ease and skimming through an important part of the early development of our robot. (not that skids are bad, I certainly like them I just see many competitive advantages to being able to move in any direction)

I see.

Our students have always wanted to explore swerve. That would definitely take a lot of development to make it a "go-to" solution. Sab-BOT-age has published a lot of information if you want to take the swerve plunge.
http://wiki.team1640.com/index.php?title=Swerve_Central

Simbotics has a great design guide for your students to explore different drive trains and selection criteria.
http://www.simbotics.org/files/pdf/drivetraindesign.pdf

Good luck with your training.

David

[Bkeeneykid]

Many of our very successful robots have been built on a stock drive chassis. It seems as our team gets older, the students and mentors who both have the desire and capabilities to create a custom drive chassis gets fewer and fewer. It always depends on the game, but this year, we knew right away we would have to spend time creating a VERY robust drive train, something none of our few members had any experience with. Both our 2014 and 2016 robots are built upon the newer AM14U chassis, and we've had great experiences with both of those. We modified our AM14U3 to have an indent in the front for the Rock Wall and a ball opening, and we spent considerably less time cutting these on the porta-band than we would have spent on our own. In light of not being able to get AndyMark pneumatic wheels, we adapted their hubs to our own wheels, with moderate success. Frankly, for low resource teams such as mine, I think time saved on the kit chassis is worth more than creating any of your own.

I'm not the only one with this opinion, as I talked to the other teams in my school district about this at Maker Faire KC (2335, 1984). 1984 used a VersaFrame bot, and had troubles with it throughout the entire day. They still said they preferred it, but they also lost a few matches from it. 2335 used a modified kit chassis like ours, but they used a aluminum super frame around it for bumper mounting.

[Billfred]

Originally Posted by fargus111111111

Often the kit base gets a bad rap for being inadequate and this year without modification it was, but I want to talk about how they perform in more average FRC conditions aka a flat carpeted floor. In 2014 and 2015 my team used a 4 CIM kit base and did relatively well, in fact if you ask anyone on the team they will probably say that those were some of our best robots in recent history (last 6 years). I would credit much of their success to how quickly they were running for drive practice and testing. With this in mind I want to ask 2 questions.

1. Does your team frequently use the kit base?

2 a. If yes to q1. Do you modify it and if so how?
2 b. If no to q1. How do you reliably develop a drive base to be quickly usable for on field testing, have you formed your own kind of "kit base" that is the origins of each year's design?

I ask because I am interested in encouraging my team to work on developing other types of drive trains in the off season but I'm not sure if it is worth it because of our very good experiences with the kit base.

Probably best to lay out these answers in order:

2009: C-Base, with F-P motors (or some small-can motor like that) into AM Planetary, because who needed more than that that year?
2010: Full custom frame, with 2-speed. That one worked out pretty nicely for twin 1398 (and partner 343), but oye those robots had teething problems.
2011: C-Base, made into 6WD and wedgetop on the wheels. 2 regional wins playing hard D. No complaints.
2012: C-Base, 6WD, no changes at all. Another regional win, this time on offense.
2013: C-Base, 6WD, belts between wheels and a chain run going to the then-new VEXpro ball shifters. A miserable failure, but given my current affiliation I'll make this clear: entirely our fault. We discovered after the season that the kids neglected to use grease or thread locker during assembly as recommended, and the mentors failed to catch it. One of those years where you'd kill for a mulligan because everything else about that robot was money.
2014: AM14U, stock but with the 3-CIM upgrade. Won Orlando playing literal wear-the-wheels-bald defense.
2015: AM14U2, made into H-drive with an additional AndyMark Toughbox and VEXpro omniwheels. The H-drive component was scrapped after our first event in favor of a ramp. At least the other omniwheels meant it turned well?
2016: Full custom chain-in-tube frame, with dead-spaced #25 chain. Worked great once it was shaken down. Team Cockamamie used the AM14U3 for their Robot in 3 Days build, which also worked pretty well for the purpose but wouldn't have lasted a season. The very small bellypan and open front meant the sides of the drivetrain were beginning to camber inward.

[EricH]

Kit base? What's that?

1197 does custom drive, generally a WCD variant of some stripe, every year. Helps when you have in-house aluminum welding capability. We re-evaluate the specific needs each year, but we can do that sort of drivetrain pretty easily.

[GreyingJay]

Originally Posted by GeeTwo

In 2014 (Aerial Assault)

Tee-hee

Originally Posted by GeeTwo

In 2016 (Stronghold), we opted out of the AM14U3, but purchased two AM14U2's before the season, realizing that almost any design we were likely to build could be adapted from the KoP chassis faster than we could do it ourselves.

This was a clever move, and assuming you got them when they were about half price, represents considerable savings. When I first heard about what you did, I started paying attention to the AndyMark deals and sure enough during the Christmas holidays there was a Tuesday deal on the AM14U2. We hemmed and hawed about whether there would be a risk that the AM14U3 would be different enough from the AM14U2 that designing two "identical" robots from them would be tricky. In the end we went for it (I placed the order 15 minutes before midnight on Tuesday!) Sure enough there were minor differences in the hole patterns, gear ratio, and the wheels, but for the cost and time savings we were happy to work around that.

[GeeTwo]

Originally Posted by GreyingJay

Tee-hee

Oh, did I type that out loud?

Originally Posted by GreyingJay

This was a clever move, and assuming you got them when they were about half price, represents considerable savings.

I don't know if we did for the 2016 season, but we definitely did for the 2017 season!