Japan's Life-Sized Gundam, Through the Years
June 10, 2024 4:16 PM   Subscribe


*clicks Buy Now*
posted by HearHere at 5:40 PM on June 10 [2 favorites]

"GunPlayPal operators are standing by now to take your order!"
posted by cupcakeninja at 5:45 PM on June 10 [1 favorite]

I saw a largish Gundam statue in Denden town in Osaka maybe 20 years ago. I was just there looking to buy some electronics and happened upon it. Maybe because of that I just figured that there were large Gundam statues spread across Japan and they'd always be there so it sapped me of any urgency to actually go and seek them out. With a bit more urgency I could have seen the Yokohama one when I visited in 2022.

In my defence I've never watched a Gundam show or movie before but as someone that loved Transformers when I was growing up I'll always have some fondness for large robots.
posted by any portmanteau in a storm at 5:56 PM on June 10 [5 favorites]

Oh man. Every giant robot built moves us closer to a future I want to live in, and every loss is a tragedy. Unfortunately, physics keeps getting in the way of these projects.

The square-cube law imposes a hard cap on how large we can build things with currently available materials. As you scale an object up, its surface area increases with the square of that increase while the volume increases with the cube. However, the strength of your materials stays constant. This is a big problem.

I think this is easiest to illustrate this with examples from biomechanics. This is straight out of the square-cube wiki page:

Haldane illustrates this in his seminal 1928 essay On Being the Right Size in referring to allegorical giants: "...consider a man 60 feet high...Giant Pope and Giant Pagan in the illustrated Pilgrim's Progress: ...These monsters...weighed 1000 times as much as [a normal human]. Every square inch of a giant bone had to support 10 times the weight borne by a square inch of human bone. As the average human thigh-bone breaks under about 10 times the human weight, Pope and Pagan would have broken their thighs every time they took a step.

IRL, one of my best friends is about a foot shorter than me, but clocks in at about half my mass. Conversely, I dated a woman who was only a few inches taller than I am, but easily had 40-50 lbs on me. Furthermore, whereas I have allocated a good portion of my mass to storing all my extra cookies, she had specced into a build that seemed completely comprised of ironlike cords of rippling muscle. Combine those two design decisions and she could induce a force vector of staggeringly greater magnitude at will. What seem like relatively small differences can add up extremely fast.

Kurzgesagt has an excellent, highly accessible series of videos on this subject if you’d like to explore further: Part 1, Part 2, Part 3.

That article from Inverse has some pretty telling details. That 60 foot mech weighs 25 tons. At least one member of the project addresses the square-cube problem:

“When you make something big, everything becomes heavier, and the strength of the material becomes insufficient,” fellow GGC leader Pitoyo Hartono, an Indonesian engineering professor at Nagoya’s Chukyo University says.

And it seems, like most of these projects I’ve seen, they’ve hidden the compromises they’ve had to make when the limits of material science collided with physics:

Though it’s not immediately evident to the naked eye, the robot is attached to a flatcar called the Gundam-Carrier. This keeps it suspended and supported from the lower back, enabling it to move up, down, forward, and backward. During the performances, smoke blows up over the low wall in front of it, drawing the spectator’s attention away from how its feet fail to connect with the ground as it marches in place.

The second big limitation is Newton’s second law, F = ma. This is not quite the same hard cap as the square-cube law. But when you’ve got 25 tons of m, you really do need a tremendous amount of f for even a small amount of a. But even if you could harvest and install an S² organ and instantly summon infinite f, square-cube strikes again. Acceleration induces a force vector and pressure; go too high and your materials will yield.

Robotics is a pretty small world, and I was so. damn. pumped. when I had my first opportunity to pilot a friend’s giant mech. Evangelion was one of the more influential anime I had seen in my youth, perhaps only surpassed by Ghost in the Shell. I was ready for this. This was to be the apotheosis of my boyhood. I climbed into the cockpit, executed the start-up sequence, lights were green across the board, so I finally, finally grabbed the joysticks. And moved like the world’s largest yoga instructor. That company went out of business quickly after their public debut, and every supporter (financial or otherwise) was furious at the founders.

But oh my god, the second time I piloted a friend’s giant mech, it was a completely different beast. It was extremely lightweight, basically just a space frame. Its hydraulic power plant was running off of some of the highest energy density batteries ever created, harvested from electric superbikes. On the whole, his team’s mech was smaller than many mechs I’d seen, but mass was minimized, force was maximized, and you could actually get some decent acceleration. And its control system was unreal. I have known founders that have poured literally hundreds of thousands of dollars into controls to try to get their mechs to walk before having to abandon the effort. It’s a hard problem and the stakes are incredibly high; a small mistake in the development process can destroy things real quick, including the pilot inside.

But this second mech walked. It’s the first one that’s ever done so. My friend had spent no money on controls. Instead, he opted for a human-in-the-loop controller. The mech had four legs in row, two on either side of the cockpit. When I climbed into that cockpit, my friend strapped my arms and legs into an exosuit suspended within. Moving my legs controlled the inner pair of legs, moving my arms controlled the outer pair of legs. I powered up, lights were green across the dashboard, and I took my first step. I fell on my face immediately. Thankfully, that mech was designed with a pair of spring-loaded tusks on the front and rear of the cockpit for exactly that reason, and I bounced back pretty easily.

The training to be able to pilot that mech took time, but less time than would have been spent on software. Writing the controls software to make a robot walk is HARD, but living creatures are incredibly good at this kind of stuff. Our proprioception rocks, and we’re able to solve all the equations necessary for gait intuitively with a bit of training. Watching the speed at which my nibbling progressed from motor babbling to crawling to toddling to walking to running as he established control policies in his tiny developing neural net was breathtaking. My friend had to instruct me every step of the way when I was in the mech’s cockpit. We started with push-ups. Then basic, shaky steps. Eventually, we worked our way up to a gorilla gait (alternating the inner pair of legs with the outer pair of legs) and suddenly that machine became an extension of my body. Which amplified my strength 40x. I stumbled around a huge field. I clambered over a traffic barricade. I flipped a truck. I had fun. I use a photo of myself in that cockpit on my dating profile.

You can see an old video of it in action here. They’ve got a webpage too, but they’re really most active on social media. They’re doing pretty well, and some of the stuff they have in development is absolutely bananas.
posted by 1024 at 7:10 PM on June 10 [23 favorites]

They declined a offer to fight one of the SRL machines.
posted by boilermonster at 7:36 PM on June 10 [2 favorites]

Unfortunately, physics keeps getting in the way of these projects.
*shakes fist at physics*
posted by HearHere at 8:47 PM on June 10 [7 favorites]

Imagine this scenario:

I am a very rich man (IRL, I am laughably poor, but just go with it) and I go to my financial comptroller and I say "Janice? I need a million or so to purchase a giant, walking exoskeleton. Janice, I need this thing. My kids need me to have this thing. My wife needs me to have this thing. Her secret boyfriend that I actually know about and am taking steps to ... ahem ... needs me to have this thing, Well, not really, but go with it. Janice, the world is crying out for me to have this thing! Janice! The world, Janice! The world! W.O.R.L.D. You wouldn't stand in the way ot the world's needs, would you? Think of the starving ..."

Janice holds up her (probably) expertly manicured hand, looks me square in the eye (she's 6'2, played lacrosse in college and still practices with her teammates for fun) and in the driest, most cynical voice that ... Janice has actually been to a very expensive voice clinic with psychologists, voice actors and sound men and women to achieve a voice so cynical, Jeanine Garofalo weeps for joy to hear it (and she never does that) ... driest, most cynical voice that you can imagine and says:

"Why, dawg?"

Help me here. I need to tell Janice something. She's staring at me and it hurts. What is the practical application of this walking exoskeleton thingamajig?

If I say "For sport!", she'll say "No, you pay me to make you spend your money sensibly. Now get out of my office. You're upsetting my gardenia."

If I say "For great justice!:, she'll look at me witheringly until I just leave and yes, she's had training. Her stares once made the Dalai Lama pause, look around and shrug and he wasn't even there.

What are these things good for?!?

Note: yeah, that's cool as hell and I actually do want one to just drive it around. But is there a practical application envisioned?
posted by JustSayNoDawg at 9:17 PM on June 10 [3 favorites]

I should have married Janice, and would have, except her wife is not into thruples.
posted by JustSayNoDawg at 9:22 PM on June 10 [3 favorites]

Be a lot cooler if they could walk around.
posted by neonamber at 1:24 AM on June 11

What is the practical application of this walking exoskeleton thingamajig?

We don’t actually have a Kaiju problem. Exosapien’s PX1 serves very much the same role as the pretty damn awesome life-sized Gundam in TFA, which you didn’t feel compelled to write a story detailing Janice’s dismissal about. I watched at least one Mobile Suit Gundam series as a kid, and I would have absolutely loved to have seen that Gundam. I’m sad that I’ll never get the chance. The first Mobile Suit Gundam anime was released in 1979, so I assume I’m not alone in this feeling.

Although this is near the bottom of TFA, which would require reading it, they do mention that seeing the Gundam is not free. You do have to pay for a general admission ticket to Gundam Factory if you want to see the 59’ life-sized Gundam as it performs every 25-35 minutes, and I doubt there is any way around this cover charge. A premium ticket that grants one access to the Gundam-Dock tower to get as close as they’ll allow you to costs twice as much as the general admission ticket, on top of the general admission ticket. In addition to the regularly scheduled performances, Gundam Factory’s site is still up and has some a schedule of special events that change throughout the year. There used to be a cafe, as well as a museum and an event venue they say was used by “companies, research institutes, and schools,” presumably in exchange for money. All their merch is listed as sold out, but interestingly enough, looking at the site plan it seems that the shop was not located between visitors and the exit as I would expect. Instead, it looks like visitors had to pass through the shop to get to the toilets. It looks like they had food trucks too.

But while Gundam Factory Yokohama closed on March 31 and will never make anything ever again, 17 days later in April Grimes rode a giant mechanical spider Exosapien built onto the stage of Coachella. Bella Poarch took a turn at the afterparty. Everybody who saw that machine wanted a chance at the controls, but there is never, ever enough time to accommodate demand. I can’t talk about anything they haven’t already publicized, but they’ve been doing this for years – Airrack’s got a couple scenes in this video from two years ago where he traipses around in some of Exosapien’s early builds. There’s more content out there going back over 8 years.

I’m not actually familiar with Bella Poarch or Airrack beyond their engagement with Exosapien, though google says @bellapoarch has 2.3B likes and 10m more followers than the entire population of Germany. Grimes, OTOH, is very enthusiastic about deep tech and shows up the oddest places, but that’s a story for another day.

The Exosapien team have studied giant robots intensely. The founder, Jonathan, absolutely eats, sleeps, and breathes this stuff. He’s been working on this for 18 years now. A lot of the work on the PX1 and the other stuff he and his team have built started out of the foundation he launched in Vancouver 16 years ago that’s supported over 2000 artists building large scale, technically sophisticated artwork. Before he could work on Exosapien full-time, he kept the lights on working as senior engineer designing mighty morphing shape memory alloy stents for neurovascular implants. He is a smart cookie. He has studied every giant robot that’s ever been built and talked extensively with many of their creators to get the inside scoop on what worked and where they went wrong. I don’t think there is any man who has ever lived who has as much expertise in the design, fabrication, and business of giant robots.

Where Exosapien is succeeding, there is also a graveyard of business models that have failed and they have learned from. Most recently, that would be Gundam Factory Yokohama, which reminds me I should really text Jonathan some of these links (thank you cupcakeninja!). Their model didn’t work.

Right now, the few players building stuff like successfully are largely following the supercar model. After two years in development, Tsubame Industries came out of stealth last year and got a good amount of press for their transforming Archax mech. Interestingly enough, Tsunami is also based out of Yokohama, not very far from the full-sized Gundam in TFA. Robotics is a small world.

I don’t personally know that team or much about their mech beyond their publicly released specs in that last link. At 3.5t, it seems pretty heavy to me, especially when its exterior paneling is entirely 3D printed ASA and fiber-reinforced plastic. Its top speed seems a bit low too. I do not like the idea of getting into its fully-enclosed cockpit, the hatch entirely enclosing me inside, and not actually being able to see outside directly with own eyes – from inside that chamber, you can only look out through 4 internal screens which switch and combine the feeds of 9 external cameras. I’m not particularly claustrophobic, but I do value binocular vision and enjoying depth perception when I’m operating heavy machinery. They seemed to be very much pursing the Lamborghini / very rich person’s toy model. Last I heard, they were aiming to build 5 mechs, and the sticker price had increased to $3m from $2.7m. Their site shows 5 different colors available, but I bet one could work something out if they really wanted to. If your first thought would be to consult Janice, your “financial comptroller” (???) for permission to buy one of these, I don’t think it’s for you.

My money’s (lol no I won’t ever be afford one) still on Exosapien’s upcoming Exoquad. That thing is fast, it can jump, and its dexterity and articulation it's the closest any ground vehicle I’ve seen to being supermaneuverable.

Exosapien is following much more of the F1 blueprint. You cannot buy their PX1, it’s not for sale. They are absolutely do engage in custom work for individuals and corporations, there are mech and mech-like vehicles they’ve built in private hands, but not all of those private customers publicize them.

I am not a lifelong F1 fan for a bunch of reasons, the first of which starts with “Bernie”. As non-fan, I’m probably speaking out turn here, but I know many professional engineers as well as engineers at heart that can’t get enough of it. My understanding of the appeal is very much informed by my view that engineering, and particularly design, is an act of service. I dive into that quite a bit in that linked comment, but it basically boils down to this: when you’re doing design professionally, you are not building what you want, you have to stay focused on applying your skills to serve your end-user’s needs. Some people chafe under those expectations, but I find a lot of fulfillment and satisfaction in taking this stance.

What I left out of that comment was that you are also making compromises. All the damn time. You always always always have to work with constraints. Sometimes they’re monetary constraints coming from users, sometimes they’re coming from your employer, sometimes there are constraints on how you fabricate things when you making multiple units in batches, but there always limits. Well, maybe not with killbots, I know of one single man-sized robot where the military branch singing the checks shelled out $400k just on custom-machined titanium screws to fulfill some extremely specific requirements, but I won’t build killbots so I can’t speak much beyond that. All other scenarios of constraints of some sort, and a huge part of the whole game is working within those and finding clever ways to mitigate them and delight your customers. That’s very much the job, and it’s a fun puzzle. But when you know there’s a better way to do things but you can’t, one can chafe a bit. Or at least, from time to time, wonder what it would possible to build if one had no external constraints whatsoever, and was free to solve a given problem with absolutely no compromises, using the pinnacle of knowledge that humanity has achieved so far.

I believe that’s the itch F1 scratches. It’s the dream of what one could build and the life one could lead with absolutely no restrictions, lived vicariously through the lucky few on F1 teams. Perhaps this is similar in some ways to how some people relate to celebrities and follow celebrity lifestyles? I shouldn’t speak for fans of F1 or celebrities as I am neither, this is very much my head cannon. But I do know that it is especially rare to find mechanical or mechatronics engineers whose hearts do not sing a least a bit at the thought of creating an absolutely no-compromise machine, achieving the absolute apex of what is possible with our species accumulated knowledge, and then absolutely tearing through the streets of the Monaco Grand Prix in it.

Before spending caps were introduced in 2021, it was estimated that developing and building a top-tier F1 car could cost as much as $400m. That year, F1 capped each team’s spend to a much more reasonable $145m per season, decreasing by $5m per year. For reference, the most expensive lambo of all time sold at auction for $8.3m, although I believe a portion of the sale price may have been accounted for by it’s rarity, as it was one of 9 commemorative units that Lamborghini built for it’s 50th anniversary that originally sold for a paltry $4.5m. But you can’t buy a latest-get F1 car. They’re not for sale. Old ones might be, but then, those aren’t exactly the latest and greatest – they were once F1 cars, but you cannot buy a F1 car, only former F1 cars.

I have had to make many compromises in pretty much everything I’ve ever built. I’m not upset about this, I’ve had extremely good reasons for each and they lead to better solutions, or were otherwise the absolute right move. One time I was in China designing a clutch, and I needed a material for the friction lining. I perused a materials catalog that written in mandarin, which I could not read, but the material properties were all in Arabic numerals so I was able to get by. I found a wonder material in that catalog whose coefficient of friction and ability to withstand heat exceeded my wildest expectations at an unbelievable price point, and I ordered samples immediately. I tore open the package when it arrived the next day and quickly realized what I’d found. It was totally asbestos.

I mean, if I could use that stuff? Wow I’d be using it EVERYWHERE, it’s kinda awesome on paper. But yeah, not an option. If I truly had my druthers, I would be building almost everything I make out of diamond. Seriously look up its material properties, it has 5x the thermal conductivity of copper. But if I used it in a heat exchanger, few customers would be able to be able to buy it, so I will never get to create the highest-performance design I know how to build.

Exosapien is trying to do that with mechs. I can’t talk about stuff that’s coming down the pipeline, but unless there’s been some other team building mechs in stealth for almost two decades, their flagship model is going to be the best that can exist. I will never be able to buy one. I doubt I would ever be able to afford just the fabrication alone. But that’s irrelevant. It’s not for sale.

I’ve worked on many robots making everything from food to medical devices. I take great pride in my work, and I can look at what I’ve built with a great deal of satisfaction. Here and there, I’ve really been able to help people.

But with all the things I’ve built, not once in my life have I felt the sheer fun and raw exhilaration that I have with Exosapien’s robots. When I got into that cockpit for the first time, extended my arms and legs in the exosuit for that first pushup and felt my movements amplified perfectly as this giant machine hoisted me 10’ in the air in perfect synchronization with my body… dude I lost it. I started laughing like a maniac. Total age regression. It was FUN. There is something incredibly immediate and almost tactile about the exosuit control interface. It responds to your every move, and you feel every movement like it’s your own because you’re suspended right at its center of mass. With a bit training, it really does feel like an extension of your body. I have never felt that with any machine in my life. Jonathan’s been working on this for 18 years now, he has clocked far more hours in the cockpit than anybody on his team by a large margin and an extraordinarily talented pilot. But when he gets frustrated, he straps in and goes for a stroll around the mech yard. They’ve worked out a deal with the local junkyard and there are always fresh cars to stomp other huge things to toss around.

It’s pretty good for that.
posted by 1024 at 7:51 AM on June 11 [8 favorites]

I saw the one in Odaiba back in the late aughts. That was fun.
This post reminded me of a group at AnimeExpo many moons ago who were building an actual working giant mech. Not sure if the folks in this video are the same group?
posted by The Ardship of Cambry at 10:22 AM on June 11

« Older Fuzzy wuzzy were five baby birds' head   |   The Time I Built an ROV to Solve Missing Person... Newer »

You are not currently logged in. Log in or create a new account to post comments.