Robot Combat Best Practices

Introduction

There are a lot of guides out there that will talk about specific parts that are good to use, materials you should build from, and where to get parts. Those are all important things to know and are a critical part of gaining the understanding you need to build a combat robot. This guide isn’t about that. This guide is on ways to approach whatever bot you’re building that can set you up for the best experience possible.

What are “Best Practices”

Best practices, in this context, are the tools and techniques you use in the design, build, and battling of your robot that set you up for success. Success doesn’t necessarily mean wins, as that may not be the goal of the bot. Success means delivering on the idea you had for the bot.

Core Best Practices

Design for Your Manufacturing

The best design in the world isn’t worth anything if you can’t actually make it. What fabrication options do you have available? Do you have the time, tools and skills to do it yourself? Do you have the budget to outsource? Is this even something that can be made? These are all things you need to think about when designing.

Design for Availability

The perfect part may exist, but you need to actually be able to get it for it to be useful. Everything from screws to speed controllers can run into the problem of it being the “perfect” solution that you can’t order. You also don’t want to find yourself in the spot where you can’t buy spares of your preferred weapon motor because it was discontinued 4 years ago and the ebay seller you’ve been buying from finally ran out of stock.

Design for Safety

No matter the size of your bot, thinking about safety features early can be a massive help. How will your weapon lock actually work? Where are the sharp corners/edges? How easy is it to access your power switch? Not spending the time on this early can result in having to modify your design after the bot’s been built to accommodate a mandatory safety feature. Beyond that, having an easy to install/remove weapon lock and easy to operate power switch will make your load in and load out go smoother which means you’ll be ready to fight or repair faster.

Design for Assembly

When designing your bot take time to think about how each part goes into the bot. It’s very easy to design yourself into a corner where all of the parts are physically capable of occupying the space as designed, but due to how the bot is constructed or the order in which things go together you can find yourself unable to install a part or without access to a critical fastener.

Design for Repair

This is robot combat. Stuff will break. Unless you’ve got a spare robot for every fight, which while technically an option is a massive expense and likely a significant waste of resources you’re going to have to make repairs. When you look at a design you need to consider the parts that are most likely to take damage and look at the process to repair or replace them. If you’ve got 20 minutes to repair between fights and it takes you 15 minutes to get a busted gearbox out you’re likely not making your next match with a working robot.

Design for Failure

Classes in robot combat are defined by a weight limit. This means that for any given class you don’t have infinite weight to make parts stronger. If a bot gets hit hard enough something will fail. Ideally, this something is not catastrophic to your ability to finish the match, is easy to replace, and isn’t the most expensive part of the robot. Designing mechanical fuses so you can influence the failure points can create a situation where you’ve got inexpensive, easy to address repairs between fights instead of having to toss out more expensive parts because none of the cheap stuff broke.

Design for Maintenance

Similar to repairs, inevitably there’s stuff you’ll need to do to your robot between fights. Maybe it’s replacing worn out wheels, sharpening the leading edge of a wedge, or adjusting belt tension. It almost certainly will also be charging or changing your batteries. Think about how these tasks will be performed and try to minimize the time this will take. You’ll thank yourself later when you have repairs that need to be done on a tight timeline and all of the regular maintenance isn’t occupying most of that time.

Advanced Best Practices

This section covers topics that are much more involved and time intensive than the core best practices section. Treat these topics as “worthwhile if time allows” as you can get by without them when necessary, but integrating them into your process will be well worth the effort.

Driving Practice

The more time you can spend driving your bot the more comfortable you’ll be running it at an event. To the degree practical, matching the type of floor and floor coating will only help. With non-spinners and bots that aren’t likely to throw parts/debris you can get a lot of practice in by using a floor and a small protective barrier to keep the bot contained. On bots with spinning weapons or other systems that can send chunks flying you’ll need to think more on containment which may limit your practice surface options. Ideally you’ll be at the point where you’re comfortable enough with your bot that the bot is doing what you want without you actively thinking about how to manipulate the controller to cause that action.

Design for Aesthetics

There’s nothing wrong with keeping things simple, but taking the time to add some color, shape, or style to your bot can result in a bot that is memorable in and out of the box.

Minimize Tools and Hardware Variation

To the degree you can, you want to minimize the number of different wrenches, nuts, bolts, drivers, and other equipment needed. This will save both your back and time in the long run. If you’re working on a bot and you’ve got some 10-32 bolts in there in 7/16”, ½” and 9/16” lengths you should probably take a step back and see if there’s a good option to reduce that to one length, similarly if you’re designing the bot and have an entire catalog of thread sizes it’s worth the effort to look at areas where you can change the bolt size used and cut an entire bolt size out of the spares pile for an event.

2D Profile Gearmaking

Gears can be expensive. Custom gears with weight relief can be even more expensive. 2D fabrication processes can be used to significantly cut these costs if you’re willing to put in the time and deal with a bit of extra prep/cleanup.

There are plenty of vendors out there that’ll sell you a gear, but more often than not the gear you can buy isn’t exactly what you’re after. Similarly, places like https://www.rushgears.com/ will make custom gears to order, but at a steep price, particularly at the low quantities most projects will need. Maybe if the loads are light you could 3D print the gears. Given the choice between “not right”, “too expensive”, and “wrong material” there’s got to be another option, right?

This is the spot I was in when looking to do a custom gearbox for a 30lb combat robot. I was able to source a few gears that were “right” but to get the full reduction at the strength, weight, and footprint required I would need to go custom.

Using 2D fabrication (waterjet and laser cutting) I was able to get gears cut to spec at a much more affordable price via https://www.bigbluesaw.com/.

For my gears I opted for laser cutting from AR500 plate. I also opted to go with a stacked gear approach to minimize the impacts of taper and HAZ due to the cutting process.

Fresh from Big Blue Saw

Two stages in this gearbox would be done via stacked, laser cut gears. The smaller gears were cut from ⅛” plate and the larger gears were cut from 3/16” plate. For each stage of gear you stack three small gears and align them with two large gears. The reason for doing this is to ensure that there isn’t a situation where a single stacked plate is only engaged with another single stacked plate. This minimizes the risk of one of the stacked gears taking on the entire load passing through that stage of the gearbox and reduces the chances of the binding that could result from one gear wearing faster than the others. I also opted for hex shafting as it would better spread the forces out than a single keyway, reduces sharp corners (the cutouts in the gears have rounded features at each corner) and avoids creating a thin section near the keyway that would be a likely failure point.

The first step in gearbox assembly was match marking the gears so I could easily assemble each stage without having to determine which of the 12 possible orientations was the correct one for installation. For this I set each plate on a common shaft and used a green paint pen to mark the teeth.

Stacked for Marking

As you can see looking at the above image, some of the teeth look a bit rough. This is illustrative of one of the downsides to this fabrication method – As features get more detailed and plate thickness goes up the surface quality of the cut will tend to go down. Luckily, this is a case of things looking worse than they are.

With the gears marked it was time to make the shafts and spacers for the gearbox. (Along with a few other shafts used for the build)

Shafts and Shims

With the shafts made it’s test fit time.

Test Fit

Here you can see how the differing plate thicknesses force all gears to be engaged instead of allowing for the possibility of only one gear being engaged during use. The gaps you see between the plates can be closed up with additional shims if necessary.

At this point the gearbox was ready for run-in using valve grinding compound to smooth out the rough surfaces on the gear teeth.

Grinding the Gears

Once the gears were running smoothly the gearbox was disassembled and cleaned to remove the valve grinding compound.

Ready for Cleaning

After a full cleaning it was time to reassemble and lubricate the gearbox.

Fully Assembled

With that all done a quick hand check showed that things were running smoothly and were ready for use.

Hand Test

Notes and Lessons Learned

  • At 16p, 3/16” thick plate resulted in aesthetically poor teeth, but a bit of finishing work got them running well
  • Taper could quickly become an issue at higher plate thicknesses
  • Using low taper waterjet cutting may allow for full thickness gears with no HAZ and should be considered if the project budget allows it
  • Using hex shafting required some additional work on the front end but should simplify maintenance
  • The ability to design weight relief into the gear profiles is a massive benefit when dealing with tight weight limits
  • Good gear tooth profiles are important, as is modifying the profiles to accommodate the fabrication method. Kerfs and beam/jet radius need to be accounted for in your profiles and should be designed into the part so you’ve got better control of it.
  • There are plenty of sources out there for gear tooth profiles that you can use as a baseline, https://geargenerator.com/, https://evolventdesign.com/pages/spur-gear-generator, and https://www.engineersedge.com/calculators/spur_gear_calculator_and_generator_15506.htm are just some of them

2 x 72 Grinder File Release

With testing complete, drawings updated, and the launch video uploaded it’s now time to release the full design.

The link above contains a STEP file of the full assembly along with the PDF assembly drawing and the DXF you’d need to have the flat cut parts fabricated by a waterjet or laser shop. (SendCutSend) With the DXF I’ve left a border on the drawing with the overall sheet dimensions needed. For fabrication shops with automatic quoting you may need to delete this to not get an inverted version of the parts.

This project was a collaboration with Goat n Hammer. The files above are being provided at no cost for anyone that would like to build their own version.

If you’ve got any specific questions feel free to reach out via any of the social media accounts linked on this page.

Build Progress: Custom 2×72 Belt Grinder

A while back I began work on a project with the fantastic folks at Goat n Hammer and it hit a major milestone today, the laser cut (thanks Big Blue Saw) frame was built up enough that the motor could be mounted and the tilt functionality tested.

Once the grinder build is done the drawings will be updated to reflect any changes I want to make after building the prototype and the drawings along with some build documentation will posted and made freely available.

Nyx and Algos CAD Files

I’ve seen a few people request cad files of bots in varying weight classes recently and had already published variants of these elsewhere, however between site updates and host migrations the old links aren’t likely active. Beyond that, I wanted to republish these in a neutral format with both bots in their “final” state.

The above link contains .step versions of the forked lifter and hammer configurations of Nyx along with the final design for Algos and a slightly updated weapon design that should make manufacturing easier. In the case of Algos, there was some hand finishing to get it within weight (wedge sharpening, shaving off small non-critical areas, etc) but it is the chassis design as fabricated.

If you don’t already have a cad program that can open step files one great option is getting a maker license for Solidworks. The SDK-IDs that are working at the time of this post are 9SDK2020 and 9MAKER with the former likely updating each year some time in the fall.

March 2021 Update

It’s been quite an unusual year. With everything going on there’s a lot to talk about as I’ve had some real time to dig in to projects that have needed some attention.

Getting back into robot combat

After a much needed break I’ve started designing and building fighting robots again. This effort started with the un-retirement of Algos, and continued with development of M-80, a modular 3lb bot, Herbot West – ReRaminator, a 12lb rammer with a “hypodermic spike” and Eccentricity, a tubular robot with hand forged attachments. More coming soon including a collaborative project with Wedge Industries. A small preview of that can be seen at the end of this post.

Shop Equipment Design

I’ve been lucky enough to be able to work with the wonderful people at Goat n Hammer on turning a piece of equipment they made into a design that made freely available to anyone that would like to build it. The first prototype has been laser cut by Big Blue Saw and is being built right now. Following the prototype build the design will be updated to reflect any changes that will aid with assembly. Once that’s done the build and all of the drawings necessary to fabricate the equipment will be published and made available at no cost to anyone that wants them.

Merch

With the increased quality of print-on-demand services in recent years it seemed like the time to get some real merch into the world. At the moment the two places you can get Near Chaos gear are our Spring store and our Etsy shop.

Other Recent Work

Recent Projects and Site Rework

As is pretty clear, NearChaos.net has a new look. This is to go along with a new focus. Previously the main focus was robot combat, however my interests have expanded to include a wider range of fabrication projects. Future posts will document new work. For now, here’s a few things I’ve worked on recently.

Mailbox Post
Custom welded mailbox post
J Hooks and Knife In Progress
Forged hooks and seax style knife
Headphone Holder
Hand forged headphone holder
Forge
Home made dual burner forge
Mounted Anvils
Two anvils mounted to custom stands