Robot wars

Are their movement pre-determined?
Gooks, for how fast they may be, they can't possibly be RCing at those speeds.

Whatever your opinion is, these robots are quite limited in their abilities. Battlebots requires more creativity and cleverness in order to win.
These robot-rats here, all they have to do is push. In fact they can barely be recognized the one fr om the other.

Apparently the robots have to find the opponent and avoid falling by themselves:
en.wikipedia.org/wiki/Robot-sumo
That doesn't really require less cleverness than for Battlebots.

They have a preprogrammed AI on how they should behave. They might have a RC for starting / stopping them though.

>Robot-sumo, or pepe-sumo

i literally prefer paying 10k usd to 10 homeless, and put them in 2 teams and spend another 5k getting some AKs and ammo letting fight between them for the total prize

That explains why some robots have white flags, which I suppose are used to distract the enemy's photo-resistors/camera from well perceive the surrounding.

The white flags are probably trying to imitate the border of the arena.

You're aware that in both robot-sumo and robot combat there are R/C and Autonomous categories, right? Plus the most successful bots in combat tend to be armored wedgebots relying on pushing power alone (pic related, the most successful heavyweight ever in Robogames is a wedgebot). Of course these types of bots won't do in televised shows that cater to the "me no see flying bits. me bored" crowd, so they enforce rules that prohibit bots without active weapons for instance.

I did not know. Many of those chink-robots appear to not actually determine the opponent's position, but instead just sprint around the whole arena, which is kinda a bummer, while the one in vid related is really fucking stunning as it moves behind the enemy.

these robots rely on infrared and or ultrasonic sensors to detect the other robot. ultrasonic can be absorbed by the sheets infrared won't. but you'll notice sometimes the robot only lowers a white flag on 1 side. it's to give an opponent a false signal to the side of the robot that can then turn into them when they charge and push their opponent off. it may also help if an opponent samples multiple positions before charging, lowering a flag on only one side will move the center of the signal to the side of the robot rather than the center of the robot. so hopefully the opponent will charge at the side and not dead center.

If those robots either use infrared or ultrasonic for surrounding detection, then why would the arena have a thick white circle to contrast the black center? I mean, if they aren't affected by color surface, then what's the point of such neat contrast?

I would guess they use both: visible spectrum for the border, ultrasonic/infrared for detecting the opponent.

yea. it looks like most players use the same source code. would be pretty easy to sneak in and win the championship with our own code.

Seems kind of pointless to create a physical change to your robot, when the base functionality isn't even optimal. there could be improvements in the motor speed.
what if it was like an internet of things type battle, like ctf where each other had to hack in real time the others bot. then there would be no need for extra stuff like flags or wedges. art of war by sun tzu my nigga

It's good to see that our next generation is literate enough to spell "yeah" correctly.

This shit is boring as hell and lame. All this gay safety prevents cool weapons from being used. I'm sure butthurt losers crying about their robot getting obliterated into useless scrap also has to do with it.

You're allowed to program in any sort of behavior. It can be fully scripted, it can track enemy movements, it can use cunning strategies, anything goes. It just has to push out the enemy robot out of the ring - classic Sumo rules. Hence it's called Robot Sumo.

Programming a 90's microcontroller to accomplish things like that is no easy feat. And you really don't want to use modern ARM let alone PC parts for this, with the kind of speeds and collisions these robots routinely undergo, they'll get disintegrated half a second into the battle.

Then you don't need robots for that at all. And such games already exist.

That's because it's a sports game, Sumo specifically. The object of the game is to push the enemy out of the ring, not to get it destroyed. You can always start your own robot tournament with weapons allowed. But I'm pretty sure that, unless object of the game is expressly robot destruction, same type of robots would dominate the scene. Also most of the time robots get obliterated anyway.

It's basically the same as complaining that Formula 1 series does not allows car ramming and shit like that. Go be a retard somewhere else.

Protip: "Gook" is a slur used to refer to Koreans. The slur you'd use for the Japanese would be "nip."

they typically have an infrared sensor array near the ground at the front. white reflects infrared more strongly, black will absorb infrared. hence the colors for the ring.

the ultrasonic sensors are usually ping sensors. there are also passive ultrasonic sensors. the opponent detecting infrared sensors are the sharp model sensors for detecting obstacles within 1 - 5 feet, not for detecting the ring.

it's possible the competition mandates using beacons; that would depend on the competition. and no at least for basic microcontrollers i.e. arduino these don't use cameras. with an sbc - possibly.

for reference, here's a kit and some relevant parts:
pololu.com/product/2510
pololu.com/product/1137
pololu.com/product/702

and pics related are typical ping sensors, infrared obstacle detection sensors, and infrared line detection array.

If you want destruction, you start right by your regulations, limiting the allowed types of robots, like the tv shows (like said above, bots without weapons aren't allowed), and like what Battlebots specifically did, a scoring system that is heavily biased towards spinner weapons.

...

The image is you, right? Read again.

I kinda want start building something
I am a third year EE student, so I have some knowledge of circuit theory
Does anyone have some books that I could read?

/pdfs/

The best manual fighting robot on a flat surface has already been invented. The shell protects it from any impacts, and its entire shell spins. The spinning prevents stab or hammer damage by deflecting it perpendicularly. The gyro effect prevents it from being bullied by heavier bots. The blades on the edge are slightly inclined so they have a LIFT effect, and a CUT effect.

It has never been defeated or destroyed in battle, and is so effective that this type of robot is banned in most robot games because when it hits opponents pieces of shrapnel fly all over the crowd. Out of all robots in Robot Combat Sports it is by far the simplest, cheapest, and easiest to build or repair.

If your echolocation can't see where the enemy is, should your robot:
A) Stay in one place and try to locate it.
B) Move to a predetermined location in such an aggressive manner, that if the enemy robot is there by chance, it will be knocked out.

It's very simple, the robots are designed to have no OODA loop.

Go be a faggot somewhere else.

Slow down there Mr. Mythbusters Fanboy. Blendo was just ahead of its time when it came out, that's all. It had pretty much no competition in 1995, but when it competed in Battlebots later on its opponents were much more capable of handling the hits Blendo dished out. In fact I don't remember it winning a single battle in the entire time it competed in BB.

I wasn't really sure, so I went and checked back. Blendo entered Battlebots 4 times and didn't score a single win, lost all 4 in the first round. Also in it's last run it pretty much grenaded itself after getting hit by an hazard. So
Is wrong on both accounts

It would be interesting to build an adversarial evolutionary simulation for the external design of these bots. For the software side using an evolutionary neural network or reinforcement learning model would be useful.

First of all it's called Robo Sumo. Now, the point is that this sport is not about destroying robots. It's about accomplishing a goal defined by the game rules. The goal is to push the opponent robot out of the ring. With the kind of movements these robots do, damage is often inevitable. But destroying the robot isn't the point. Just like in boxing it's not the point to pulverize opponent's face, it's to knock him down.

Yeah that's great until I come along and flip you onto your side.

Granted, Blendo ripped Biohazard's front panel off in the first hit, but what good does it make if you stop working after delivering a single hit?

There were some bugs, and the Blendo platform should have been improved as every other robot gained mass and power... but the basic platform is undefeatable.

Blendo is like Linux, roughly speaking.

Ugly and bloated?

Holla Forums battlebots meetup when?

Shock-mounted thick armor is what it takes to have any high power horizontal spinner destroy itself from its own weapon kickback alone. Also, the weight of the shell makes fullbody spinners have a very long spin-up time, when facing them, play aggressively and hit them before they're up to speed, and there's little they can do to you.

you know you can do this with arduino?

there ya go, at long last, some fucking content from this guy...

You can do that with a plethora of micro-controllers, what's your point?
Arduino is just an Atmega328 with a couple of other components all integrated in a board. Again, what's your point?

you sound knowledgeable. so how would you make the robot win if you were the roboticist?

nobody here is a roboticist, after all, thats why this thread was buried...

so how would you make him winner?

let answer = if (you == roboticist) {something} else {nothing};

i thought you are a microcontroller expert. do you not know how to program?

thats a great picture of yourself there your mommy must be so proud

this is me

this is not

if its not quite clear by now i would not have put anything meaningful into the threads your necroposting out of principal, not to mention that im not a roboticist so i have no idea how to answer your question, and [heres the thing youre repeatidly failing to understand] if i did or was and there was at least 1 more person who cared then this threead wouldnt have died! nobody cares what you want to see on the front page, the world does not revolve around you!

Holla Forums does

if that were actually true the threads you 'like' wouldnt need bumping

i'm not a bot!

this is JUST what a machine would say

Jesus fuck my IQ dropped to negative values just from reading this

Back on topic, Robogames coming in late April should have some robot sumo events if anyone is interested, though I'm not sure they'll be casted, last year's livestream was limited to robot combat from light, middle and heavyweights, and I figure the same will repeat this year.
In any case, this should be a good year at least for robot combat. Battlebots S3 was already announced, and China is coming in big with 3 large events already (King of Bots, ClashBots and FMB). Also genuinely impressed by some chink bots in KoB (already airing), thought they would all be cheap knockoffs of western ones, but there are 3 or 4 really high quality ones even beating well established western teams.

King of Bots is sponsored by superstar martial arts actor Jet Li. Fighting My Bots (FMB) came first.

I know FMB already existed, it's just that only recently they really invested on international advertisement, it previously was mainly a national event. Also from what I gathered KoB offers much higher build/repair funding and prize pools than BB, and for that quite a few teams that were previously on Battlebots did not apply this time, and are going to KoB instead, as Battlebots S3 and King of Bots 1.5 will be filmed around the same time.