I know this has been suggested before here, but I am looking to start an “open source” chassis project based on designing simple, inexpensive drivebases that can be built with basic tools. The idea would be to develop basic designs (not necessarily detailed cad files, more like simple overall drawings) for chassis that teams without experience or access to more powerful tools can build confidently. I’ve seen too many times that a team can’t compete, or gets on the field and has their robot rendered useless by simple problems with their drive system, and I feel this not only discourages them from continuing, but is also easily remedied.
Our team would like to see this idea grow, and are already planning on releasing the basic design of our drivebase. Does anyone have an interest in such a project, or any ideas on how this could work?
Not to sound too dismissive, but isn’t this the idea behind the kit frame? Now there is a separate issue that we will not know fully until kickoff, that the kit frame might not be included, but if it is, then it is a simple means for less experienced teams to build a drivetrain
The goal of the Kit Bot is to get teams with little to no experience rolling (almost) out of the box. Many teams love this frame, and modify it to meet their needs. Others disregard it entirely. I fall into the second camp, and here’s why.
With a competition as high impact as FIRST, I don’t feel comfortable using a material that I wouldn’t trust my own weight on. In addition to this, I have unintentionally put one of my custom fabricated chassis’ corners entirely through the side panel of a properly assembled kit frame. I simply don’t trust the kit frame without modifications. Modifications are something that I love for not specialized applications. Mod your toaster? Props to you. FIRST is a highly specialized competition, that requires complete control over as many variables as possible in order to succeed in competition. I would personally rather have control over every single detail of my frame.
What would make me very happy would be FIRST giving teams the option after seeing the game to either order a kit frame, or have box extrusion shipped in for the teams that don’t want the kit frame.
As far as an open source chassis, go for it. I did some design work on one in the past, it’s in Inventor 11 (I believe). Assuming I can find the CAD, would you like it?
As far as the kit bot goes, it can be a great base. However, I’ve seen many teams struggle with it for various reasons. The kitbot isn’t really a single chassis design so much as a useable frame. It is designed to be a general erector set like setup, which can cause confusion for new teams and can also cause problems with mounting mechanisms and drive components. Our thinking here was that if teams have a variety of general designs available and explanations of the designs/features/concepts, they would be much more likely to build a reliable, competitive chassis as well as understand a little more about the engineering process in general and drivetrains specifically.
Craig, anything you could offer would be useful and greatly appreciated.
Craig, I have to disagree on the durability of the kit frame. Having used the kit frame this year I can say that the machine took quite a pounding. It survived plowing at full speed into both the center divider’s support pole and the back wall as well as 2 regionals, Championships, Kettering, and WMRI. That frame worked wonderfully out of the box for us.
That being said, the frame has limitations for some teams and I feel that having not one but multiple open source chassis available would be an interesting way of helping out teams.
Sure thing. I don’t actually have class tomorrow, so let me boot into windows and I’ll get to work.
Bumpers. I’ve had bad experiences with kitbot frames without bumpers. If bumpers are mandatory again in 2k9, then the kitbot is a viable options for rookies. I have a feeling they will be, so more power to the rookie teams. I also believe that teams can learn more about material selection and strength of design if they must design their own chassis. If you can correctly design a FIRST chassis that doesn’t bend once, chances are you’ve learned enough to do build your own car chassis (not nearly advanced as automakers, duh).
I strongly disagree about the car frame; making sure something doesn’t bend is only one of many considerations they make. How about dealing with corrosion from road salt? Strong temperature differences (as hot as +120F or as cold as -40F?) in a chassis connected to dissimilar metals? Designing it for maximum possible recycleability of components? Designing it to purposely bend certain ways under certain types of accidents? Designing it to maximize efficiency of the manufacturing process? Or have the least wasted space in shipping containers or trains when making the frames in Mexico and shipping them to America to mate with the engine built in Europe?
FIRST is but a foundation, a way of getting your feet wet. But by no means does graduating the program make you an engineer. That’s what college and a lot of hard work is for.
I’m going to have to say that for rookie teams or with teams that don;t have a lot of machining resources, the kitbot frame, especially with bumpers, is more than robust enough and easy to put together for FIRST. Especially in the standard 6 wheel drive configuration, there is nothing wrong with the kit frame. I would think a good starting point for an open source frame would be the slightly upgraded version of the kitbot frame that team 121 has used for the past couple years.
I do agree that IF a team has the machining resources, and the people to design a better drivetrain, then sure a custom drivetrain will be lighter and stronger than a kit frame, and potentially easy to work on, but that is a big IF that a lot of teams in FIRST should not be attempting. Many teams should just build the kit frame and devote there resources to a great manipulator
Simple sheet metal. drill holes to mount gearboxes to. Cut slits for wheels to come out of bottom. Use box aluminum to outline the perimeter of the bot and to keep it from being too wobbly. This method is the best way (i believe) to build an omni bot.
I will try to put up some of my sketches. I’ve got loads of great chassis designs and dont have time to build any. So if your really interested in making a new chassis…I mean I’ve literally got pages and pages of cheap and easily made innovative base/chassis designs.
What kind of machining resources do you think are actually needed for a good custom chassis? I’m under the impression that any team can build a very competitive robot with just a hacksaw and a drill, the main problem they face is lack of experience or confidence in doing such things.
The experience problem is one that we can try to remedy by building up a bank of knowledge for teams to use. Open source designs would help tremendously with this. As an example, I got to be on the drivetrain design team during our rookie year. Having never built anything like this before, we were all lost as to how everything should go together. We spent at least a week designing an overly complex wheel attachment method before we discovered the simple way (see pic). If we had been able to browse past designs from other teams before hand, we would have had good knowledge about how such things work and wouldn’t have wasted a week on our crazy contraption.
I also mentioned an issue with a team’s confidence in itself, which I think is a huge problem for rookie teams. (It probably doesn’t help that they always read comments like “…well, you’re a rookie team, so I would suggest sticking to the basics”). The assumption that rookie teams aren’t as capable led to what I think was our teams worst strategic decision: In our rookie year (2006) we decided that shooting into the high goal was something that only the hardcore veteran teams could accomplish. We immediately shot it down and decided to be a low-goal scorer. Looking back I realize that it was a massive missed opportunity, and we could have pulled off a decent high goal shooter.
I guess the whole point I’m trying to get across is: many teams are only held back because of a lack of experience or knowledge in FIRST robots. Posting past designs and open-source designs would be a huge benefit for them.
I really like the concept of an Open Source / Community anything. If nothing else, the collaboration effort is an interesting learning experience that can bring teams together.
This sounds like a great chance to (continue to) bring open source to the mechanical world.
A Gallery of past successful designs in great detail (CAD files, images, calculations, written explanations…) would probably become one of the most useful tools for FRC design out there. I know how helpful the FIRST Canada galleries (http://www.firstroboticscanada.org/site/node/96) can be when looking for ideas, and this sounds like an expansion on that idea, so I say go for it.
Otherwise, it might be helpful to incorporate Kit frame based designs onto this page as well. It is a great oppourtunity to play with 6WD and Mecanum designs without going into heavy manufacturing. Giving rookie teams access to things like incorporated supershifter designs, or chain tensioning systems could enhance their on field performance without too much work for them.
I now have a 6 wheel West Coast drive in CAD that has been boiled down to being fairly simple to make. Tomorrow I’ll be adding the Toughbox into it to allow for even less custom work. Once I’m done with that, the CAD will be up for grabs.
Back when I started in 2002, we didn’t get any of this handed to us. We had to figure it out ourselves. It was hard, and we learned a lot. What good is this program if robots are just entirely put together from a kit and instructions? We’re losing the design and analysis aspect of the competition here.
Agreed. The goal of only releasing the CAD, and requiring any teams that actually build it or modify it to also share the CAD is to promote advancement and collaboration of a design. We’re not giving them a box of parts; we’re giving them a vague design (that probably has errors somewhere, I haven’t spent much time on this design…), and letting them decide on the implementation.
Counterargument: How inspiring is this program if a big chunk of your regional is spent stuck in your starting square, or sending out a human player while the robot remains stuck in the pits? Give teams a solid baseline, then let them optimize as their resources (time, money, manpower, machining abilities) allow.
To that end, I might suggest getting in touch with the entities behind the 2009 kitbot. If you can come up with methods of improving future versions, either in durability, versatility, cost, or ease of use, your work may get more use by teams than a separate design posted to CD. The kitbot might go against some of the design sensibilities you or your team relies upon, but what better way to help teams than to help improve the aforementioned baseline?
Well put, this is what I had in mind as well. A large collection of “what works” would benefit inexperienced teams immensely.
This is one of those things that can be ferociously argued either way (similar to the mentor debate). My opinion is that time is better spent focusing on the strategy of the robot or some complex manipulator rather than trying to figure out small annoyances in the drivetrain.
This is also a good idea, since the kit chassis works so well for so many teams. As ,4lex S. mentioned, putting up different configurations for the kit frame would be helpful. I think it would also be good to provide ideas for teams that want to go past the kit frame too, for the small number of teams who prefer a design that won’t work well with what’s provided in the kit.
I believe the open source chassis would be much more effective if it utilized the kit frame.
The kit frame doesn’t out of the box offer a decent 6 wheel, if you guys just make the simplest method of making a 6wd with the KOP frame, people will probably actually use it. Pretty much, use all COTS parts, make some assembly and exploded drawings with complete BOMs, and release that. Dawgma did that a few years back for a non-KOP base (I believe it was the “sub $1k drivetrain”).
Sorry to rain on the parade, but a team capable of making the kinds of bases you describe probably don’t need or don’t want to download a design. Now, the team that meets in a physics room and has no engineering mentors would probably actually use (and be able to use) a 6wd with the KOP frame.
Pretty much, if they have the capability, they won’t want the design, and if they want the design, they probably don’t have the capability.
I disagree. I am talking about a design that would require very few tools (i.e. a saw and a drill), yet would be simple and competitive. The kitbot is nice, but rather than trying to make it work better for a maneuverable, dependable chassis, it makes much more sense to start with a clean slate. It is surprisingly easy and inexpensive to build a frame that is much more suited to your needs than the kitbot, and I think that doing so will make many teams’ lives easier.
