The Future of WALKING BATTLEBOTS with Mark Setrakian
Вставка
- Опубліковано 16 жов 2024
- We have a SPECIAL INTERVIEW with Battlebots Legend Mark Setrakian. #battlebots #blog #robotics
Mark Setrakian is the man who designed the infamous walking Battlebots Mechadon, Snake and The Master. He is also the engineer who designed the platform upon which the Giant Nut rotates, as seen at the beginning of every Battlebots episode. Mark shows his new walker designed named "Stalker", and discusses the future of Walking Battlebots.
Mark gives us insight into his history with Battlebots, the start of his career and working in robotics. You don't want to miss this!
Follow our journey through Battlebots!
Tune In every Thursday night at 8 PM ET/5 PM PT on Discovery Channel
battlebots.com
battlebots.com/robot/skorpios-2020
Would you like to fight robots yourself? Bot Bash Party brings the robot fighting directly to you! We do parties and corporate events for ages 8 and up. Located in Lafayette, CA, Bot Bash can provide our services from the Northern Sacramento region down to San Jose. For corporate events outside of state that need multiple arenas, please contact us with further details and we can accommodate.
Ask us about Private Tutoring lessons through Zoom!
Contact: botbashparties@gmail.com
Website: botbashparty.c...
Instagram: / botbashparties
"my first Battlebot was a walking robot"
he says that so casually like it isn't the most amazing thing ever, love him lol
One of the ONLY walking robots ever submitted. I remember maybe two others, one of which was Son of Wyachi and that was more of a pseudo-walker simulating wheels.
@@OneBiasedOpinion Pour one out for the shufflebots.
I think they changed the Arena to have a higher up stage , that could be an advantage for walkers.
@@OneBiasedOpinion ua-cam.com/video/QaoG6MYx9Ks/v-deo.html i mean thats kind of a walking robot :D
From an engineering standpoint, Stalker is amazing. The articulation and the ability to go anywhere it wants is really impressive. From a combative standpoint, it's slow. Any faster wheeled bot can take out one leg and it's done. Don't get it twisted, under the right circumstances it could do really well but, I think, it's just too far ahead of its time. Wouldn't mind seeing a stomp-bot division though.
With AI it could be programmed to be faster at revovery and such, but I suppose that's against the rules.
@@prototypeo1404 hi current rules allow autonomous robots and assisting AI as long as there is a remote kill switch, even in 2016(unsure if date correct) there was a robot called Chomp which used AI to know when to activate weapon and I think it also was able to help the robot self right .
Yeah it definitely needs it’s own class in order to have a chance at being competitive
@@54raceman Walking robots are already allowed to be 2x the mass. What more do you want?
The legend returns!
I wonder if instead of a huge weight bonus, simply having a uneven ground at least in some places in the battle area would make walkers more viable. Uneven ground would defocus the ground scrapping battle as well. Or a even a shallow pit, like a toned down version what they had in RobotWars back in the day.
The problem is walking is just so much worse than driving that it only becomes competitive as a form of locomotion when the terrain is basically impassable for wheeled vehicles. Like let's say there was some tactical advantage to being able to climb up on the shelf and the rules allowed it -- you'd just see more robots like Huge that can get up there with a simple two-motor drive systems and some big plastic wheels.
@@wyrmh0le ya, I feel like people would just opt for bigger wheels over a complicated walker bot
@@thunderd7904 there should just be a separate division for walkers imo. There's no way something like chomp and it's insane articulated legs are going to survive a hit from something like tombstone.
@@scotthardin7189
That's why Chomp had belt armor over its legs. I also want to note that Chomp was never actually running at full speed, the team did a demonstration video showing the bot at a full run and it was a lot faster than what we actually saw on TV. Problem is, that eats into its gas supply, since the whole bot is powered by pneumatics except for the hammer turret.
@@VestedUTuber Walker Chomp is an engineering beauty and the chomp team is always trying innovative stuff. That said, the belt armor couldn't even stop hijinx from hitting it's legs and hijinx, while improving every year, doesn't yet have super heavy hitter kinetic energy. I'd LOVE to see all kinds of crazy robots in the arena but the meta is the meta for a reason.
I think it would be cool to have an entire season where ALL the Battlebots MUST be walkers (just to throw a monkey wrench at the builder's). No wheels at all! Then make it a normal 250lbs. weight limit and see what people come up with. and do the announcment in the middle of this current season to give them some time to ponder their ideas.
one team will say hell no and make a snake that somehow can slither across the arena lol
The obvious way to encourage walkers is to have an arena with bumpy ground, steps etc. I'm sure you'd get some folks trying with large wheel designs as well.
What I would really like to see is an autonomous battlebots competition where the only controls you are allowed are 'start' and 'stop'.
That would be so cool
Autonomous combat robots have already been done in the antweight and beetleweight classes. They weren't anywhere near as interesting as you seem to think they would be, and the auto-ant and auto-beetle classes only lasted a few years.
@@VestedUTuber I don't think they would be as exciting as the current radio-control bots, but I think I would personally find it very interesting.
@@timaidley7801
I think you're underestimating how bad the autonomous classes actually were. Only one person actually bothered to even take them seriously, by entering a modified Kitbots WETA kit, and subsequently lost to a shitpost bot. The problem is that there's no suitable sensor equipment that can take the forces involved even in the lower weight classes. Chomp's semi-autonomous functions were very unusually reliable for this environment, even the accelerometers used in "Melty Brain" bots can't really take the abuse.
That Stalker is an amazing machine!
That final pose at the end, all it needs is a large brain with a monster face on it and it becomes an Arachnotron from Doom Eternal!
Proof that Mark Setrakian really is a Spider Mastermind! 🕷
Minigun on the bottom, finished :p
thats really cool! im also attempting to build a walking bot with no formal training, and i never considered approaching the control systems from a music software approach, but it makes sense because each leg cycle is like a looping song you need to sync with the other legs!
last friday I got to eat lunch with Zach and Dianne. they are awesome people.
*I. Love. Walkers*. Seeing Mark coming back with a walker is something I unexpected, but something I welcome too. Wish the best to Mark in his run in Battlebots!
I never realized just how versatile, maneuverable, and stable a pentapod is
Mark is MY HERO! I can't even imagine where the sport would be without him pushing the envelope of what you can do.
LET'S GOOO! WALKER BOT NATION!
What a treat! And I love how Diana is just holding her face in awe in the beginning of the video.
The two things I think killed walkers were: 1) the arena is designed from the ground up to cater to wheeled vehicles. It's flat. An irregular surface would be much more amenable to locomotion competition. 2) the definition of a walker and the weight advantage - SOW is not a walker. It's a series of cam-driven feet. A dog is also not a walker, though dogs can walk awkwardly a short distance. A walker manages its COG relative to its feet, using a minimum of 3 degrees of freedom (humans use 5-6 DOF) in each individually articulated limb, and can both lift itself, and balance itself - with a single limb. A walker can also hop on one limb, which leads to it being able to run. Speed is an advantage. Today, I think as a walker, there needs to be rules which mandate that advantage, so that a walker is something respected, not just poorly used to access a weight handicap. Designs that need 3 limbs on the ground at all time shouldn't be designated walkers. But regardless, fix the arena so a handicap isn't required. Nobody wants to win a battle and have people in the audience saying they only won because of the weight handicap.
I couldn't help but think about Boston Dynamics and thought what if your BattleBot was able to jump over its opponents. Potentially could be a big advantage in a fight.
Jumping for 3 minutes doesn't win you fights.
@@Narcan885 Jumping on the enemies might
@@avataru_ Yes, please, jump into a 60 lbs tool steel drum spinning at 200 mphs, see who wins that impact with a frail walker robot!
@@Narcan885 The conundrum of hammer-saws fighting Bloodsport only its your drive system you're planning to stick into the ferrous wheel XD
@@Narcan885 What I was thinking is you could jump over your opponent and attack them from behind. Most bots have their weapon and most of their armor at the front so attacking the back would be more effective. Just not sure what weapon would be practical to use in this situation.
Stalker is definitely a well designed bot. It moves so fluidly and I hope to see it win some matches eventually. It is an engineering marvel at this point to coordinate the movement. And there haven’t been many walker bots since Chomp.
🔥🤯 Wow..... This guy is insanely good.🤯🔥
That dynamic walking cycle is amazing. it feels alive.
The primary issue I see with walkers in BattleBots is their reliability. Wheeled and tracked bots already lose often because their motive systems get FUBAR’d by the opponent’s weapons. Having exposed, articulated legs as your primary source of mobility seems like a very easy way to get disabled very quickly, especially against the ubiquitous kinetic spinners in the game.
I'd say the most likely answer to Setrakian's problem of 250 lb Walkers is to reintroduce middleweights.
Either that or a walker multi-bot.
It would be really cool if Battlebots had a special walker bot exhibition fight from time to time.
It would certainly be better than the walker fights we had on Robot Wars UK😄
Forget about being a Battlebot with this, I can think of so many other excellent uses for this design such as rescue, NASA rover missions, and even as a Star Wars droid, etc.
I love the beauty and brilliance of Mark Setrakian's walking designs. They're amazing works of art and science. I cannot, for the life of me, imagine how one could compete in modern battlebots. It seems as if one solid hit to the legs from any sufficiently impactful kinetic weapon would immediately end the match, and their speed and maneuverability is simply too limited to avoid such a hit. I would love to be proven wrong, though!
My thoughts exactly, but perhaps we need a walking bots only tournament, could be fun.
@@klausknudsen106 yeah , i agree with the idea
that or one that is desinged like chomp but make the turret sensor controlled and let it face towards the enemys without inputs. with the right kind of weapons that could realy work. making such a bot 100% ai controlled shouldnt be that hard in a fixed arena . the biggest problem would be top spinning weapons the dmg the weapon of the walker.
This needs to be in the show!!! The interview is spot on. You did Greg a favor by doing this and trying to expand the sport and hobby.
Now I can't help but imagine the claw hand that holds the giant nut tossing it aside, flipping itself over, and taking off running like Thing. Personally, I think arena design with rough or uneven terrain that favors walking over wheels is the best way to make walkers viable in combat robotics. I like to have as few constraints in the actual rules as possible to allow maximum innovation by the builders.
Rough terrain wouldn't make walkers viable, it would just screw over designs reliant on a low ground clearance. Look at the bots that were successful at TLC's robotica if you want to know how rough terrain would change things, they're mostly conventional but they're just built with bigger, chunkier wheels.
Imagine a 8-legged robot jumping around the arena, shooting webs at spinner-bots, grabbing bots and throwing them out
I wonder if going smaller would be better - loads of fast moving much smaller versions all armed with drills and magnetic feet which could climb onto the opponent in a sort of swarm, stick to it in a place it couldn’t reach and go to work. If the opponent is still driving then it’s not a pin or anything so once onboard the mini walkers could just go to town!
That bot is already amazing. But amazing is what I'd expect from Mark Setrakian. Some thoughts on turning this into a viable Battlebot: 1) horizontal spinner, and the bot can tilt itself forward to reach low bots, and 2) use the 2X weight bonus for walking bots, but make it a multibot.
Mark Setrakian makes the most awesome bots I have ever seen!!
See that gives me SO many ideas for a bot based on Stalker. I wish I could talk to Mark directly! BTW Stalker is AMAZING! Make it... JUMP! As an attack! that would be would be awesoume!
very impressive as always from Mark Setrakian
I still have that old video tape from the 90's with Mechadon. How times have changed.
Wow Stalker is awesome! Make two arms with saws, one arm with a flipper, one arm with a spike, and the last arm a flamethrower!
Mark has always made the greatest walkers and is one of the best robotic engineers
He's a legend! I saw Mechadon back in the day when Battlebots was on Payperview.
its incredible just how nimble it is. it walks so stably and fluidly while also being able to adjust its height and tilt in all directions. most legged robots just kind of hobble along, but not this one. it also seems like it doesnt have a set stepping pattern, which is incredibly impressive. I could imagine this structure being viable for walking drones, rovers and other applications.
It is my goal to make a walking combat robot! I'm new to combat robotics (1 year in so far and I'm obsessed) and wanted to make an easier bot for my first bot but one day soon I wanna make a 60lb walking combat robot to bring to the Angry Vehicle Competition. I was design it around a hydraulic system but maybe a servo system would work. I would totally be down to but a walking bot to battle other walking bots!
Imagine if we could get even 8 walkers for a side class! That would be so freakin sick!!!
Then someone enters a walking shell spinner with enclosed legs and dominates the whole class, because cool-looking walkers just don't work.
The software side of things is completely out of my element which makes these things even more impressive, but he's 100% right, Mechadon wasn't the best fighter but it was a thing of beauty and just something that makes you say wow when you see it, sticks in your mind so many years later and then you look at the results "Oh... I guess it didn't do so well... but man was it still awesome"
As for what purpose Stalker could have as a platform, on one hand I'd think it'd be a cool platform to hold some kind of spinner underneath it so it could straddle an opposing bot and just have the disk dig into it, almost like a Hammersaw in a way, at the same time that just seems like such a high risk/low reward idea I don't know if that'd be worth it
so basically in a lower weight class the bonus weight might allow it some potential in a combat scenerio... Id love to see it!
i cant imagine them surviving anything fight wise up in the 250 class
Yeah it would get its teeth kicked in against a 250lb bot
Okay, now let's see him develop a meltybrain walker :)
I don't think limbed bots will be viable until there's an automous bots class. Moving limbs out of the way of spinners and drums quickly enough can probably only be done by an autonomous bot
I think a walker that's a rim or full body spinner would be cool
I wish more Mechadon footage existed on YT. It's tough to explain to people who weren't watching back in the early aughts.
Totally worth the wait!!
Part of me wonders if the solution to all of this is a slight less even or less perfect floor. Take away the fork meta, thus hurting the verts. Walkers get the advantage of maneuverability and control on the rougher floor
Underappreciated reply... I think the fork game is definitely not helping level the playing field.
do a multibot, 2 walking bots so you get the weight bonus, and extra bot
I’ve read the rule book, and this is unfortunately not possible. Battlebots will not allow a walker multi bot due to the fact someone would make a tiny walker and invest the 250 extra pounds into the bot with wheels. Unfortunate but it’s a huge loophole closed
@@GrizzlyDust7222 even if both of them are walkers
Mark continues to be worthy of his crown as "THE MASTER".
Looks fantastic. Love to see this battle another walker even if it was for a one off exhibition fight.
What an absolutely gorgeous machine 😍😍😍😍😍💙🦂💙😍😍😍😍😍😍😍😍
😲😲😲😲😲😲😲😲😲😲😲😲 just in awe of how cool this is.
I'm thinking a standardised walking platform design that can either be purchased in kit form or as a set of CAD files. This platform can then be modified and adapted to each teams specification. I can only imagine designing and building something like this must be very time consuming and expensive even for a sport that's already expensive and time consuming.
The way this is constructed you could have leg kits/CAD files that could be purchased and teams could design and build their own central 'hub' (I see no reason why this couldn't work with 3, 4, 5, 6 or more legs). If the control software that Mark has already been developing is made available it would drastically cut down on the time/skillset investment that would be required for a team to start building a heavyweight walker and make it much more accessible.
Walkers are a daunting prospect even for experienced builders and realistically I think the only way it could work given the contemporary, _uber_ destructive wheeled heavyweight machines would be to have a separate walker division, at least initially.
I am wondering if it is possible to invent an extremely bulky arm to intervine, grab & stop the momentum from spinning weapons. It's so nimble and wide base could support the leverage.
I love the look and design of walking bots but I just don't think they are viable in the sport. They should have their own class like they were talking about, now that would be amazing to see. I just don't see how they beat a drum or bar spinner like Tombstone for instance. Truly a work of art and I hope they get their own class someday.
Beard goals. Man that's one of the best looking goatees I've ever seen. His beard is almost as cool as his robots
That's just amazing!
Walking bots need completely comprehensive automated control systems before any advantages legs offer can be fully realized.
I need more of this in my life!
I fear for stalker, those legs looks so fragile, and it seems to flip over very slowly
Wait a minute I’m able to subscribe to the scorpion? Hell yeah!
I feel like there should be a different class unto itself for the walking battlebots
I think your on to a innovative design. This is sick...you have a a bad ass military platform...this could be shot from arty rounds...
Beautiful creation! I love it!
The only possible way for walker robots to be viable in the current meta of Battlebots is to make it heavy and less top heavy. More armor on the legs definitely so that it can withstand hits and even then you can't ensure that if a single leg gets hit it won't just straight up destroy the walking mechanism and there's the problem of what weapon you would place on it. It's definitely a good robot idea but the complexity would definitely put a strain on repairs as well after matches anyway it's a fun idea to think about if anyone has ideas feel free to chime in.
There's a featherweight walking ring spinner that's pretty much been the most successful true walker in the entire history of the sport, simply by just being a relatively conventional design but with legs instead of wheels. Even then it kinda struggles.
So cool!!
if it ran around very quick and could stick and jab and dodge, outmaneuver other bots, then it could be a real changer.
idk, the legs will have to be built super tough though, unless it had a fast battering ram weapon mounted on top, that could have a powered thrust ram to shoot out and maybe then flip over the opponent.
Imagine an under cutter at the bottom of the bottom of the platform, with horizontal spinners at the knee joints, this could be amazing
I think it would be interesting to have an "off-road" version of Battle Bots with irregular floor.
I think a walking robot could do relatively well in a case like that.
That's sort of what Robotica was. Walkers didn't really take off there, people just went with bigger wheels.
@@VestedUTuber, but maybe walkers would be more viable there.
At least better than in a flat floor.
@@thomasfplm
The problem with true walkers is a severe lack of mobility. Most of the true walkers that have competed in robot combat tend to have very slow walking speeds and very slow turn speeds. You don't really address this by introducing rough terrain, you're just changing what wheeled designs are effective.
Walker viability is ultimately entirely on the design of the walker, not on the arena or the ruleset. One thing I have not seen in any form of robot combat, at least until now, is the use of round walker layouts. A "round" design like Stalker is capable of omnidirectional movement, something that inline walkers like Chomp can't do. And it's the only way to implement omnidirectional mobility without the loss of traction inherent to omniwheels or mechanum wheels, and without the delay of swerve drive. Additionally, they can more easily mount weaponry with a full 360 degree coverage around the bot.
The most effective true walker I've seen is a featherweight ring spinner built around a hexapod walker.
I would put most of the extra 5000 lbs weight in the forward arms of the bot, so they would act like big energy syncs for spinners. (maybe even have a program that could detect when an arm was hit, and release all its grip so the arm would transfer less energy to the body) having such a wide stance and a strong spot to hit could make for a challenge. Add in a central weapon like a diamond coated cone drill and let it envelop bots, or use hooks on its arms to flip/ act like wedges to deflect (with that low crouching stance) really a small robust design of this with the programming right would be interesting.
DARPA is going to come calling, I see a lot of potential there.
Please tell me you built two so when one gets destroyed you still have the other beauty I can't get over how cool this thing is.
Until they can find a way to speed up the walking they won't find their way into realistic competition. That's why the og whyachi changed the game on "walking" because they found a way, albeit suuuuuuper cheesey, to make a fast walker. I don't know the engineering side of things, but is there something mechanically keeping them from moving faster? That would be the edge plus the weight bonus. But it seems we are far off from that. But also it's pretty cool to see. I want to see him and the chomp team square off idea wise!
There's a lot of stress on the joints which requires more torque, which means higher gear reductions to handle the load, which means more weight, which means more inertia decreasing acceleration times
got the chance to fight his robot "the master" when i was on team odin robotics with ankle biter, he was a super nice super smart guy, but ankle biter still impaled him on the spike strip
Now imagine this with a swiveling double jeopardy on top with more than 2 shots. It would be able to aim up and down and potentially land a shot on the top armor on the other bot
Climb onto the deck and take pot-shots from up there XD
this man making a Fallen Walker from Destiny lmao
Amazing engineering. Very impressed. Would be destroyed easily but maybe a novelty.
waou mazing gaes
Magnificent
I feel like the big flat open floor is what makes the boring floor scraping meta happen. The more varied the terrain is the less effective wheels are.
I can't wait to see how BB will evolve. Today's meta has mastered the unobstructed arena. The shelf was the first disturbance to the environment and the bots reflected that. I would LOVE to see walking bots, but they need some sort of evolutionary advantage or be nightmare fuel fast.
I predict BB will start an urban arena.... Reactor online, sensors online, weapons online. All systems nominal.
"I predict BB will start an urban arena.... Reactor online, sensors online, weapons online. All systems nominal."
You mean like Mech Warfare, which pretty much died?
@@VestedUTuberMech Warfare is doing as well as the sportsman class. They nerfed weapons and most of the match is spent trying to move. Legs are complicated.
@@devinholland5892
First I've even heard of it even being active in years.
As much as I love walking robots, the only way they'll ever be competitive is to either allows them even more weight or give them a class of their own.
Needs a giant vert :D
Awesome :))
Very impressive mechanically and control wise, but the thin metal at some of the joints is not going to survive terribly long in an arena.
Hehe .... Nice Stone Island jacket 🦾
When the legs are armored they could perform like a wedge against horizontal spinners, if that front moves in a certain angle.
Is it supposed to premiere at 11 PM?
yeah, im sleepy.
Time zones homie.
I HAVE ALWAYS HOPE BATTLEBOTS WOULD CREATE A WHOLE DIFFERENT CATEGORY FOR LEGGED OR BIPED BOTS!!!👌
I'm no mechanical engineer, but although this thing is very interesting and impressive as a piece of work, I can't see it having any place whatsoever in a Battlebots arena, unless you're going to give it a chance by creating a "ponderously slow plodding" class for it to fight in. Look at the material thickness at each of the leg joints - that is not going to hold up against a single hit from a heavy spinning mass. I will admit that I didn't watch EVERY SINGLE episode, but honestly I never saw Mechadon or Snake really have much success at all.
I would love to hear a conversation between Mark and Shane of Stuff Made Here about inverse kinematics and control systems for a walking combat robot.
I think walking should be its own class
I mean, 2x 250lb walkers operating as a cluster is already a big advantage if they're as nimble as stalker!
Cool
walking robots should be allowed to carry like a .22 pistol
Its a comical idea but unfortunately it would do almost nothing to a modern BattleBot.
Its like a Monkey Lord from Supreme Commander.
Ask him is is smile okay
Oh wait wrong guy.
Stopper looks pretty cool btw
are people happy with the upper deck? This is a dream for walking robot, they could make arena in way that only walker can activate something that gives them huge advantage. The goal is fighting, not play around like ninja warrior
I'd like to see autonomous or semi-autonomous bots become a thing. The old Chomp sort of was going that way with the weapon being able to trigger if the opponent is in a good spot. Mind you with verts becoming "the thing" im not sure how semi-autonomous you could make something.
@@JustinWilliams-ed2ug Quantum was in the show a couple seasons ago with a pincher/crusher arm designed using machine learning, and I think Beta has some parts designed the same way.
The speed and maneuverability that wheels offer, with simplicity both mechanical and logistical as well as compactness and defensibility, seems to make walkers almost entirely unviable in the battlebots ring. I'd love to see some sort of outdoor or otherwise more rugged competition that takes that advantage away from wheeled bots. Unless this can somehow increase the versatility of whatever weapons system they design for it by quite a large margin over a wheeled platform, it just doesn't seem like it's trying to be an optimal engineering solution to robot combat.
Wasn't there a guy in texas who tried to build a similar Battlebot called Boris but technology as well of his understanding how to use it to build that kind of machine hadn't caught up yet? What he said about a walker division is interesting also like the idea of "ultimate" battlebots (giant machines with a massive weight limit) which was talked about when Greg was trying to attract people from the SF robotic art community like Kal Spelletich and Mark Pauline into battlebots as so far only Mark Setrakian as well as Zander rose from Inertia labs (his photos from volunteering at SRL are awesome) are the only guys associated with that scene who came over.
😍😍😍😍😍😍