Are we doing bumpers right?

FIRST General Forum
1

I’m pretty sure we can all agree that the answer to that question is probably not, or at least we can do better.

As I see all the threads popping up here about bumper questions, and I have worked with my own team on constructing our bumpers, I’m left wondering, once again, if our current bumper rules are the best way to accomplish the bumper objectives. The plywood, pool noodle, and fabric construction has served us well for quite a while, but as FRC has grown and evolved, the bumper rules seem to have grown more and more complex along with it. Are plywood, pool noodles, and fabric even the best way to build bumpers in the modern FRC era? If we were to wipe the slate clean and design a system to protect the robots from impact, what would modern bumpers look like given our now long experience with plywood/pool noodle/fabric bumpers?

The “Bumper” Criteria:
• Must reduce damage to the playing fields and other robots as much or more than the current solution (plywood/pool noodles/fabric)
• Must be easy for teams to source materials and construct (including international teams - currently plywood and pool noodles can be hard to source internationally)
• Must either communicate the alliance color (red/blue) and team number or contain a proposal for a separate way of doing so

How would you make the “bumper rules” clear and concise without all the nitpicky tolerances (1/4” gaps, 2-1/2” (2-1/8”-2-3/4”) pool noodles, 5” (+/- 0.5”) plywood, etc.), what’s “supported” or “structure”, what’s red/blue, what “unambiguous” numbers (stroke width/height), etc.) and still accomplish the criteria above (i.e. easy to source and build protection for all teams that identify alliance and team)?

Don’t limit your thought process to solely editing the existing bumper rules (which is great and welcomed) but also think outside the box on potentially better alternatives to the existing bumper design.

Some Ideas:
• A custom extrusion of some sort produced by a vendor and included in the kit (rubber? skinned foam?)
• A DIY solution but that looks substantially different than today (what other padding materials are commonly available, how else could the bumpers, be covered, etc.)
• A solution that looks a lot like current bumpers (plywood/pool noodles/fabric) but start from scratch on the rules

53 Likes
2

Chuck, this is a great thread idea. I was just chatting with some FRC folks last night about how it might be about time to start from scratch with the bumper rules. The current rules are based on the original rule set from the mid '00s, when robots were moving at about 10 ft/s, FRC specific COTS parts were just becoming a thing, and the majority of teams were still using wheelchair wheels. Not to mention that these current rules have been stacked upon for nearly two decades, as such they’ve gotten unwieldy, despite some very smart attempts to clarify and simplify things.

I very much like the idea of a custom extrusion included in the kit, and would be curious what the current and potential new suppliers might come up with if this a RFP was put out. I’ll be paying close attention to this thread to see what ideas people come up with.

48 Likes
3

Some assorted thoughts:

  • The current bumper construction rules pre-date most COTS parts. Teams can slap together a swerve chassis or get vision tracking going on their robot faster than they can make a good set of bumpers. The new bumper solution needs to be much more “plug and play” as opposed to current version which involves a lot of discrete steps. If it were as simple as “apply numbers, cut to length, add mounting” we’d be in the right ballpark
  • Obviously “add mounting” is doing a lot of leg work. But we’re rapidly approaching a world where a large majority of teams are either doing swerve or a standard kitbot. Making sure any new bumper solution can be easily mounted to these two setups would probably be enough, leaving teams using custom solutions to solve their custom mounting
  • A world where you don’t have to source fabric and all teams are using the same bumper surface material would be fantastic. Too much brainpower and money has been by teams in this ongoing bizarre race to have the most slippery bumpers
  • Putting the cat back in the bag to slow down robots is a difficult challenge. A new bumper solution that can regularly absorb the types of collisions (both robot on robot and robot on field) would really help benefit teams and field volunteers
  • I’d be open to moving the robot colour identification out of the bumpers. A lot of problems are solved if teams could make their bumpers a permanent part of their robot structure. Just slap on the bumpers and forget them. With modern LEDs (we’re well beyond the giant signal lights of 2002), there has to be something we can do with 8x10 light box to make alliance colour abundantly clear
37 Likes
4

It is unfortunate that much of the visual identity of what makes a FRC robot identifiable as a FRC robot is the complexity in these rules.

The combination of the red/blue color, size/proportion relative to the robots, and white numbers really is something that visually distinguishes an FRC robot from BEST, OCCRA, Battlebots, etc.

43 Likes
5

I am very much in favor of revamping bumpers to bring them up to the rigors of the brushless age, but I remember the days where your alliance was dictated by little flags in pcv pipes, and it was confusing to say the least. I get the point about them not needing to be removable, but boy does it help immensely with clarify on the field.

19 Likes
6

I’m personally in the camp of a custom extrusion. If it comes pre-mounted on something structural (thinking something similar in size/general shape as the kitbot side plates), then you can do away with the support requirements, it would only need to be cut to length and rigidly attached. Add to that a required set of “license plates” that can be included in the KOP to display the team number - that way we don’t have to worry about height, stroke width, etc.

If we can move away from using the bumpers to display the alliance (with a new control system, can we work something out with LED’s controlled automatically like the RSL is? Maybe even combine the RSL and the LEDs? 2-4 RSL-like devices visible from all sides of the robot?), then we could write the rules to have them permanently attached, removing the need for them to be removable and increasing the odds they’re used structurally on the robot - forcing them to be removable adds a fair amount of complexity.

Also, increase the overlap between max-height and min-height bumpers, perhaps by making the bumpers taller so there’s still plenty of wiggle room. Even with the 1-pool noodle overlap, we still see robots ramping on top of each other somewhat regularly.

11 Likes
7

It seems to me like the biggest issue getting is having sufficient protection given the increased speeds of modern FRC bots, followed by sourcing the noodles. Those would suggest a larger pool noodle or custom extrusion and including that extrusion with the kit. Allowing a plastic or metal backing to the bumpers might be nice for teams struggling with broken bumpers etc. Potentially bumper backing could be made from: 11/16-1" wood or plywood, 3/8-1" solid plastic, 3/16"-1" aluminum, or 3/4"-1" metal tubing .09" minimum wall thickness. Measurements should be actual not nominal, allowable materials should be in a table like areas of the manual that cover specific parts.

I strongly support the use of tolerances in these specifications, “sold as” specifications are uninspectable and leave ambiguity in legality unless teams are going to be asked to show a 2 1/2" label on a noodle.

Bumper support rules could use some simplification: “Bumper must be firmly mounted and may not have more than a 1/4” gap from backing to FRAME PERIMETER" or something of that nature. No rules about what’s behind the bumpers, if a team wants to have a bumper span a 18" gap that’s up to them, but I don’t see it happening, the current trends are toward swerve or the am14u both of which have structure around the entire FRAME PERIMETER unless teams go out of their way to remove it, so a team would have to make a conscious decision to not support their bumper.

5 Likes
8

How about a little flag that goes on the robot somewhere? We could even just have a half pink half green cylinder that is flipped depending on the team for easy vision tracking!

Joking aside, the bumper shape and color has been an identifying feature of FRC robots for years. I wonder if they can create a new identifier that is equally iconic.

18 Likes
9

To avoid having to change bumpers, we could throw the LEDs in a translucent bumper to change the color.

half /s

Light Up Pool Noodle Float

16 Likes
10

We smashing 70g’s now baby!

6 Likes
11

This is a problem I’ve been thinking about for several years now. Without a doubt bumpers are our (1923’s) hardest-to-make robot component year after year. We can resolve almost any engineering problem competitively these days with COTS parts except for bumpers. Therefore I think the solution lies in some sort of COTS bumper construction system that does not require woodworking equipment or any special tools.

My initial proposal would be as follows:

  • COTS lengths of 3/4" thick by 5" tall engineered plastic - something durable and resilient but still easy to cut to length accurately with simple tools

  • COTS 90 degree sheet metal corner brackets of a standard length that represents a rules mandated minimal distance backed by metal on each corner (inspired by the bumper backers from 900)

  • COTS Foam Extrusions in Red and Blue with industrial Velcro or something similar on the back to facilitate changing bumpers without ever removing the backing frame from the robot if desired.

Teams can also opt to build two sets of bumpers and permanently attach the Foam Extrusions. Additionally teams can fabricate their own corner brackets so long as they meet the minimum design requirements to facilitate non 90 degree robot corners. However teams must mandatorily buy and not modify the height and width of the COTS plastic backers and the COTS Foam Extrusions. In the case of the sheet metal corner brackets, I believe it is fair to enforce a MUCH stricter tolerance on the order of +/- 0.05" or something similar if the team chooses to fabricate their own as there is a economical commercially available solution in the market and approved by the rules.

I believe this still allows teams to be creative with robot shape and other advanced design options that require non standard corners, while offering a COTS, solution where the critical rule inspected dimensions are baked into the product, thereby not only raising the floor, but generally simplifying the build process for all teams.

15 Likes
12

Aside from robot to human safety, bumpers are a crumple zone to reduce the g forces sent through the rest of the robot, and you can make the bumper stronger, but that just means that more forces get transferred into your bearings, electronics, and joints. You want a bumper that has the ability to be squished without bottoming out and crushing into the plywood/frame.

As a stop-gap, do we just need thicker bumpers? Instead of 2.5 inches, 4-5 inches?

3 Likes
13

To state the design problem, bumpers are an interesting challenge, as they effectively need 3 components:

  • A hard part: A rigid backing to support the soft part, spread force, and provide a place for rigidly mounting the bumper to robot structural elements
  • A soft part: The “infill” that absorbs most of the energy in a collision
  • A tough part: The outer layer that resists cutting forces (tears) and abrasion against metal/polycarb field edges, as well as other bumpers

All of these components need to be inexpensive, easy to procure, and easy for teams to design and assemble into an integrated system. In addition, they have to vary in length because of different robot drivetrain sizes. Bumpers also tend to be left until last, or given to less experienced members of the team, exacerbating the problem.

Looking at these three pieces individually:

  • For the hard part: Wood is cheap and easy to procure, but teams are generally unfamiliar with it (yeah, there are wooden field elements, but those are of different scale and come with plans, versus having to design in it). The current sizing appears due to compatibility with stacked pool noodles, rather than an engineering reason, and isn’t a standard wood size. It seems like a replacement that is either hard plastic or aluminum extrusion in a standard size that can simply be cut to length would be helpful here.
  • For the soft part: A vendor selling foam blocks in the correct dimension, just cut to length (or maybe in gauge block style “increments” like 2", 4", 8", 12" that can just be stacked to hit an overall dimension), seems ideal here. There’s really nothing quite like foam for absorbing energy, and if everyone uses the same stuff, that should create sufficient demand for a vendor to be interested in procuring and selling it. Shipping is the main downside of anything custom–bulky light things aren’t cost-effective to ship–but I’m not seeing a good alternative.
  • For the tough part: Similar to wood, teams generally are unfamiliar with fabric work. It does have great properties though. I don’t have a great suggestion here unless it’s part of a larger system.

An all-in-one vendor solution that combines all three parts into a single product would be awesome (e.g. a hard plastic backing, foam infill, and pliable plastic outer face), but I don’t know how economical that would be. The complexity might make it hard to make it cut to length, and selling just a fixed selection of sizes would substantially limit robot size options. Maybe it could be a system of cut-to-length “sides” and fixed-size “corners”?

More broadly on bumper rules, I know teams might complain, but simply requiring full coverage bumpers (like this year) and no reversible bumpers could simplify rules and inspection a lot, and help avoid teams shooting themselves in the foot.

14 Likes
14

Perhaps a new RSL that indicates red/blue allegiance would be an easy first step in this direction. This would remove the need to make a second set of bumpers, AND allow for better mounting as swap-ability would not be a feature that’s needed. If teams want to spend the extra time making them swappable, that’s their business.

Allen Bradley stack lights would do a great job of filling this role. You can unscrew the red LED stack, screw on the Blue LED stack, and there’s no code difference between this and the original RSL we’ve been using since I was a kid.

4 Likes
15

False. #LongLiveRotators

6 Likes
16

Bumpers do 3 things.

  1. Dissipate MOST impacts to protect both their own robot and other robots around.
  2. Indicate alliance (admittedly, not well during practice matches…)
  3. Indicate team number.

The latter two were added some little time after bumpers were introduced.

They also have a few other interesting tendencies…

  1. Come off mid-match, disabling the robot in question. We’ll see how that goes this season.
  2. Take damage from the field…
  3. Make inspection interesting for anybody that forgets theirs or misapplies the rules. Or has their numbers too small!

How would I fix bumpers?

  1. Remove both alliance and team number from the bumpers. (And make those indicators “free” weight/up to 5lb total)
  2. Grab a bunch of rookie teams, hand them the bumper rules and a robot frame, and see what questions THEY ask. Over the years I’ve seen more questions about frame perimeter to bumper interfaces than just about anything else. (Back shortly after The Year of No Bumpers, even the inspectors were messing up on that. It’s gotten better, I think.)

Beyond that, a lot of people have good ideas with adjusting the foam and backing to be a bit more durable and easy to assemble. I’d actually suggest finding a foam specialty shop and talking to the folks there, they may have some good ideas and practical suggestions. (Or a rubber specialty shop, same reasoning.)

2 Likes
17

At some point it becomes very difficult to load and unload robots due to the pickup points being further into the robot. Its already difficult with the bumpers hitting my knees, but with enough width they would be pushing my back into places that aren’t healthy.

1 Like
18

Handles. Everyone should add handles (or connection points for detachable hooks/straps) to the frame.

4 Likes
19

Doesn’t fix the distance issue, because the handles fulcrum(?) is still on the robot that is 4-5 inches from my knees. They mitigate the issue, but not entirely.

20

I think that bumpers under the current rules fulfill two roles which should be maintained*:

  1. They protect other robots and the field from damage.

  2. They unambiguously and obviously denote which alliance a team is on.

I think the desired for uniformity has caused bumper rules to be overly specific and constrain. Here’s how I would fix it.

Eliminate the idea of specified “bumpers” from the rules. Include the entire robot’s starting configuration in the maximum dimension rules. Specify that the outermost x inches of you robot must only contain “soft” parts. Instead of specifying specific materials for the protective layer, specify a minimum and maximum amount material in the protective layer must compress when a given pressure is applied to it. I haven’t fully worked out the details, but I’d much rather it be “it must meet this specification” than “it must be built in this manner”.

Most robots already have a hard part called the robot frame. I don’t think the hard part has to be specified in the bumper rules, or the bumpers have to be detachable.

Another part of the current bumper rules that are annoying is that they have a maximum height. There’s a minimum height so the bumpers mesh together, but the only reason there’s a maximum height is because the bumpers are arbitrarily limited to 5" tall. We would have built a lot taller bumpers this year if allowed.

I like the idea of colored wraps with numbers on them for alliance/team identification. Maybe the requirements could be changed so the wrap could be applied elsewhere the the bumpers. Maybe allow shields behind the bumpers that are colored?

On the logistics side, include 4 lengths of 5" wide x 3’ long rectangular extrusions of polyethelene foam in the KOP. Pool noodles are suboptimal because they allow one robot to ride up on another so that the pool noodles mesh.

*There are also some others things bumpers accomplish which I don't think future rules should be designed to maintain.

  1. They force teams to prevent damage to their own robots. Teams have an incentive to prevent damage to their robots already, and it doesn’t make sense to arbitrarily choose damage to your frame perimeter (rather than damage to your mechanisms, bad wiring, the OM5P-AC being terrible, the drivers switching the controllers, etc.) as the one disabling event that the rules force you to avoid.

  2. They an iconic piece of imagery that distinguishes FIRST robots. This is sort of nice, but it motivates the rules to be overly specific and constrain the design of bumpers too much.

  1. They have an aesthetic of being something that anyone can assemble. The image of a group of high schoolers using parts they could find around town to win a competition (ala Spare Parts) is a powerful one in society, and it seems to me that’s part of the motivation for not switching from pool noodles for so long (even it’s not explicit). It’s feels more accessible to a layperson to say “use pool noodles that you can find in your local store” than “buy the specially designed polyethylene extrusion” even it’s in fact not more accessible.

  2. Provide a rules-enforced blueprint for how to build bumpers, in an attempt to make bumpers easier to construct. I think almost everyone in this thread agrees that the number of rules around bumpers.