Hey guys, just figured I’d drop you a line on a new project/competition that I am organizing.
The premise is a combat competition using bipeds, tripods, and quadrapods in an urban-style arena environment (think 2-3’ buildings used for cover) and airsoft based weapons (in the future, more advanced potent weapons will be implemented- microrockets and co2 pellet systems).
Hits will be scored by simple touch sensors on the ‘mechs’. Pilots will be only allowed to view the fight from a first person POV wireless camera mounted in their bot.
This is an ambitious idea, but I like it. You must have considered that the entry barrier is quite high in terms of both cost and technical expertise required to participate. How many entrants are you expecting the first year?
Since you are allowing 4 legs, bipeds will be at a distinct disadvantage, especially if you expect them to mount an airsoft gun.
Tracking and counting hits is a rather difficult problem. For this reason, I see this being more of a free-for-all than a true competition. It would be fun though (at least until the ammo runs out).
How do you expect to handle the “I’ve fallen, and I can’t get up” scenario? This problem will be exacerbated by the number of airsoft pellets that will carpet the ground.
All that said this is giving me a serious reason to consider a quad!
We’re discussing some things in terms of balance for bipeds vs quadrapods. For example, quadrapods may be subject to additional surface area for touch sensors on the legs, making them easier to hit.
We are shooting for 4 entrants for the first year, and we already have that roster pretty much filled, so we’ll probably end up with more. I’m well aware of the high barrier, but I’m hoping to attract people with pre-existing bipeds who are willing to retrofit their bots for this competition as well.
As far as tracking hits… wont so much be a free for all as we plan on focusing on 1v1 at first, and then exploring 2v2 matches. That is currently my biggest obstacle. I’m working on something fairly straight forward though, and it would be standardized and modular to the point where we can simply build 4 units that contestants can plug and play with when they are up for their match.
Most likely we’ll have a small micro that takes the switch contact (from the touch sensors) inputs and relays that via wireless to another micro that will essentially have LEDs that light up when each unit registers a hit. With these standardized units, contestants will simply have to plug in their touch sensors to the unit and be good to go. I’ll also provide the schematics to the system openly so people have the option of making their own.
In order to address the danger of pellet debris causing walking problems, we’re suggesting that foot designs have a downward curve at the edges, so they can essentially walk on top of the pellets. We also encourage you to build your bot so that it is capable of righting itself if it falls over.
A lot of this is still being discussed and planned, we have about a year to get everything squared away so I think its a reasonable timeframe.
Ok, so it sounds like there will be target area where hits are recorded, and the rest of the bot is essentially invulnerable. Are you planning to have the target be easily identifiable by targeting systems, or are shields or other defenses allowed?
Can the bots be autonomous, or do the have to be RO?
Sounds like an awesome idea! They make 6mm paintballs for airsoft guns which would be a much easier way of scoring hits than any type of electric impact system. The 1st person camera idea is cool however I wouldn’t go so far as to require it. If you don’t make that a requirement then any bot that could mount an airsoft gun would be able to compete! Also agree with the biped disadvantage comment, especially since you’ll be loaded down with compressed air and ammo. If there’s enough interest then maybe you could have classes of similar types and sizes of bots.
Target plates will likely have to be a specific color scheme. We’ll probably give a choice of colors to choose from, but they will be very basic (Red, Yellow, Green, Blue, etc). Switching up the color to be tracked is fairly simple with the vision systems I have seen.
As far as target plate placement, since we are leaving construction guidelines pretty loose, we’re going to have to be pretty open to placement. The two things every robot will have is some sort of torso, and legs (some may not have arms, or just have weapons for arms) and so that is where target plates will be placed.
Shields are not allowed.
Bots can be autonomous, just keep in mind they will probably be at quite a disadvantage vs RO bots. Theres talk of eventually creating an autonomous league though, but for now everybody will be clumped into one big league.
The camera isn’t too much of a limitation imo- complete wireless camera systems can be purchased for $100 and are very light weight. The point of this competition is to make a real-life mech simulation, and to essentially play the mech as if you were piloting it from inside, hence the first person POV only rule.
Besides- aiming from an outside perspective could prove quite difficult.
Paintballs have been discussed, and while it makes initial scoring easier as we don’t need an electronic system, it would require each mech to be actively monitored by a judge to keep score… and thus be prone to a bit of inaccuracy. The electronic system will be much more accurate in terms of hits registered.
As to paintball, there’s no way I’d want paint splattered on something I spent a small fortune and countless hours developing…
I like the idea of color-coded plates. It would be cool if each hit lit up one LED in a row of 10, and when all 10 are lit they start blinking. You could attach the LED row to the plate to avoid having to transmit and receive the hits, greatly simplifying the implementation. The LEDs would be highly visible, so I don’t think it would be an issue recognizing that a bot has been eliminated.
Killer idea, as long as they were easily visible it shouldn’t be an issue at all. I could design a simple device using a basic stamp with 4-6 touch sensors on it. Thanks for the input! Much simpler and cost effective.
Jon Hylands and I discussed the targets/hit registering a lot today. This is what we ended up coming up with:
As simple and straightforward as A-Bot’s suggestion is… we’re wanting to implement some sort of scoreboard for the event so that the crowd can see how the scoring is going.
Your target formula seems to solve the balance issue. Having only one target gives a lot of incentive to build a biped.
I understand your reasons for wanting a scoreboard for a big RoboGames-type event. But given that there will probably only be one wireless master device on the planet, I still like the idea of a standalone LED device for casual competitors.
Casual competitors can put the LED on their robot. The slave will either pulse each time a target is hit, or send a packet over that line at 9600 baud, or something like that.
You will be able to buy a master as well - its going to be very similar to a slave, but running different software, and with a USB interface so you can plug it into a PC.
That’s cool. USB is good. You guys are not foolin around. 8) Are you going to provide the scoreboard software for PC as well?
On an unrelated note, why not allow hexapods that fit within certain dimension limits? Size-wise, I’m thinking stock Phoenix would be ok, CH3R not. There are a lot of Lynxmotion hexapods running around, which means more competitors.
It was a tough call for me, knowing that there would be far more hexapods out there than tripods or quads.
Heres the deal though- hexapods, especially Jim’s hexapods, are incredibly stable and can move in all directions. You put a turret on top of that and you essentially have a tank. They also have a lot higher payload than the lesser-legged bots and can generally move faster.
So while I’d love to see more people in it, hexapods would just have an unfair advantage against the other bots even with say 5-6 targets.
I am not however against having a different class of competition that would pit hexapods vs hexapods. I would be willing to host it at Mech Wars if we had enough interest.
I understand what you mean, but I think it could be balanced with 6 targets, two on each side of the hex and one each front and back. That’s a big target.
Also, quads are going to improve over what’s out there now. But it remains to be seen how more advanced quads will compare to the 3DOF hexapods.
One final comment, I think the formula should be one target per leg for all bots. I don’t think it makes sense for a biped’s back to be invulnerable, or for any one side of a quad to be invulnerable.
^ On that point we already touched on and updated the site last night.
We thought the same thing, its not fair for a biped to only have one target as it leaves the back invulnerable, as it would penalize people for being sneaky and getting behind an enemy biped.
Because of that we decided to do 1 target per leg. On humanoids that means front/back torso, on tripods/quads that means you have to have a target visible from pretty much any angle.
Man, you snooze, you loose. I was thinking about creating an event much like this.
The main difference in the event that I was envisioning was that I would allow mechs to damage each other, and keep it to only bipeds. Scoring would be very simple, and no extra equipment needed. Disable your opponents, or at least a certain time limit to allow them to get up onto their feet.
I know, there are a lot of people that don’t want to take a chance at having their robots damaged. Well, in my opinion, most of the people that are affraid that their robots would get damaged, wouldn’t enter into these events in the first place. i.e. they wouldn’t even enter into a Robo-One wrestling event.
I got three bipeds that spend more time collecting dust in my office, looking for opportunities to smack on some other robot. I even entered one of them into a 3 kg sumo event. Outweighed by more than 2:1, and it did just fine. A few new scratches, but no damage. Did just fine.
