Low-effort animation looking into the possibility of constructing the Panels seen in Portal and Portal 2, as well as just an overall look at what they can do. I say it's possible, but I like to think big.
If I may ask, where is this sudden surge of activity coming from? Are y'all just seeing this video randomly in the recommended tab or did someone mention it?
Always makes me happy to see folks acknowledging the panels' AI. They have personalities, too. Remember the one that was just spinning around with a cube on top of it until you grabbed the cube, then it collapsed into the floor shyly? The floor is adorable. I rarely get to say that.
@@NightNinja Well... Some things are, some things aren't. For example: Literally anything that has to do with sentient AI, we MIGHT be able to make portal guns, but that's way too far ahead in the future for us to guess. But like... If given the resources, we COULD make turrets, panels and whatnot. Maybe even a lemon-ade.
@@Leekodot15 oh sorry, I meant you could physically build pretty much any of it in real life (ignoring the stuff like the AI's, portals and other non robotic things like that), like you could probably build the cores in real life using the in game models as a template but obviously the AI is way ahead of our time so it would just be at most a remote control robot, I remember hearing somewhere that while designing the models the creators wanted everything to be theoretically makeable in real life. Though I can't actually remember where I heard it though so I could be wrong.
I love how whenever Aperture Science built anything they decided to give it sentience for no reason. I wouldn’t put it past them to make the portal gun sentient.
Oh god, now all I can think about is a portal gun with a sassy personality mocking the player whenever they're stuck in an area or get themselves killed or something XD
They are basically industrial robot arms (that exist IRL) holding a square plate. We already can make robot manipulstors of that size, strength, and speed, the only practical problem is that such robot arms cost as much as a car.
Considering it comes from the company that on the verge of bankruptcy decided to buy a buttload of poisonous moon-rocks, the economic consideration is moot.
@@voodoodummie Yes but dont underestimate costs. Sure, the moonrocks are common and expensive, but you know whats also expensive? these damn things, and the fact is Theres MORE of these than square inches of moonrock. Dont forget that the black panels are also using these things, as well as the fact that theres t h o u s a n d s, maybe even f u c k i n g m i l l i o n s. The fact aperture is fucking existant in portal has to be one of the most expensive thing in the entire fucking storyverse, because the combine citadels are big, aperture is bigger than any individual one, and the combine are likely communists, giving you just enough to feed yourself unless you're joining their ranks.
@Cyberman112 Unknown Yes but consider the extreme cost of building OLD APERTURE, and that ONLY GLaDoS'es awakening was killing the scientists. until then they still had people and were trying to sell these products stupidly. Also IIRC it was said by word of god that the combine TRIED To get into aperture, but failed repeatedly until they just gave up.
I actually got to interact with one of these arms. My neighbor works at a robotics research lab and he took a small mobile one home to run tests and do work on the software
Can you imagine how expensive it would be to build even one of these panels? And Aperture has tens of thousands of them throughout their facility with replacements at any given moment?
I suppose they manufactured all of the sub-components too, and all of their operations seem to be autonomous, so probably not as expensive as it seems. Perhaps raw materials go in, and panels come out, all that's required beyond this now is energy and space.
programming this would definitely not be the hardest part (aside from the personality thing). just slap some inverse kinematics on it and you can decide exactly where and in what direction you want your panel, and the arm things will do the rest for you.
I have to reply from my personal acct because I'm at work but I'm so happy the guy with the Mugi profile pic thinks the same way I do. This is exactly what I had in mind for controlling them
"All the doors in this spaceship have a cheerful and sunny disposition. It is their pleasure to open for you, and their satisfaction to close again with the knowledge of a job well done."
Also Also Aputure: We don't use our portals to send cube into tests, we use a clogged up vactube system that if routed incorrectly by a moron central core can impede science
I mean, that AI is not "impossible" just very impractical. Anyway, you could make these things and you could even make them without the cameras built in. A few gyroscopes and accelerometers is all you really need. With the right sensors in the right spots, it could never move without it triggering the right combo of sensors. With that, you can do some actually fairly simple math to calculate the panel's current position. With the panel's current position, you can calculate how strongly and for how long to trigger each little motor controlling the panel's position, or just constantly update how much you're revving up those motors as the position constantly updates until it's in the desired position. The most complicated part I think would be finalizing that position by locking it in place somehow, but then making it easy for it to be moved again at a moment's notice. All of this added ontop of the robotic arm itself would be quite cheap, even. The real expense would be the robotic arm itself, those kinds of motors do not come cheap and that is a lot of thick heavy-duty steel. It's the same technology, math, and programming used in VR Controllers today, just for a completely different purpose. :D
You don't really need gyros even, if every motor is a servo you can know the exact orientation of the panel, as long as the foundation they are mounted on is not movable.
This is in essence a robotic arm... industrial robots typically have servo motors that track their absolute position as well as an inverse kinematics model that calculates the required motor positions.
@@yshouldicar3 well fair enough but i mainly pointed out gyros and accelerometers because those are much more accessible on a consumer-level i think, like who sells full-on servo motors on a consumer level? but gyros and accelerometers are pretty accessible because they're used in all kinds of hobbyist electronic things
@@Templarfreak servo motors *are* sold "on a consumer level", and they serve completely different purposes then accelerometers and gyroscopes (they're motors for christ's sake). Come on man, this is simple stuff.
I'm a Computer Science student and I'd say something like this is definitely possible, and while I don't know much about mechanics it seems simple enough (if you've seen the versatile arms used in factories for constructing things such as cars, this really isn't much more complex than what's already in use) In fact I would go as far to say it'd be fairly easy to build something like this, just very expensive. Excluding the part of the AI which isn't much to go by (in what sense are they AI? What information do they gather, and what are they even learning in their response? This could be relatively easy or outright impossible to implement realistically) this really shouldn't be much of a challenge to make a reality. While it's obviously for a massively different purpose, the robotic arms used in factories are very similar. Great versatility in terms of movement, smart computers, and can even move heavy items, smart with lots of sensors. We already use all of the technologies needed to make this a reality in the world today. The only reason why we don't have anything like this in the real world is that we have no application for it! Likely the cost is the biggest obstacle I imagine
I love how this video has not only gained an absurd amount of attention recently, but has been seen by multiple engineers and computer/tech science students and employees, mostly agreeing that it's relatively possible but rather expensive and pointless haha! Thanks for the input, I really appreciate it!
Why is no one talking about them cooperating? I'm sure they need AI primarily to cooperate with each other so they take into account the movement of their neighbors and don't produce collisions or lock-ups.
Im a year late but I love panels. Portal/(2) is obviously meant to look very modular, excepting Old Aperture ofc, like most things in Aperture, they’re sentient, and they can build a chamber in minutes. I like Laser Relays/Adrenal Vapor/chamber 16 (I believe) because it shows that panels aren’t just part of the chambers: they *are* the chambers. Cave Johnson couldn’t have described them better. Fully configurable. Infinitely variable. Aperture brand panels will help [(or hurt/block)] your subjects every step of the way. I like to think that every single test element, at least smaller ones (not crushers, at least partially immobile, obv not the block stairs, not goo obv, and probably not vital apparatus cents [droppers {NOT the vacuum tubes, known as Pneumatic Diversity Vents}]), are all on panels. Including but not limited to: Butttons (all 3 physics object buttons + pedestal) Hard Light Thermal Discouragement Beams Funnels Gel droppers maaayyyyybbbbeeeeee not really unless you want them to be delivered with bare diversity vents like cubes could also be Fizzlers, both kinds plus BEE2 (harder since the projectors must be lined up) Antline connectors Checkboxes And literally pretty much everything else!
I adore the "Panels" characters so much. They felt like the most "sciencey" part of the game while giving the "you're looking at background workers just doing their job despite the craziness" vibe. Given the game's type of lore, they're either an AI hive mind of former Aperture workers, or a random unscene core which can control them when Glados isn't around
Yes the programming part could be used Basically use really strong stepper motors(or servos) with a fuck ton of power from a fusion reactor to power every plate in the facility Basically defining basic movement (Turn x degrees(x, y), push hydraulic piston out (y millimetres), take photos, etc) however it is pretty hard tp define X, Y positions for servos as they take degrees pf movement. Then assign those functions to plates. Then plates to sets of plates. Then testing “labs” of plates. Then testing tracks of plates(you get the idea) It would be quite impossible for the “AI” in the plates though
Aye Ye fookin w0t m8 why would the AI be impossible? If not linked to a master computer, the panels could enter AI mode, driven by the sensor data from their environment.
Ayden Shurley I mean it’s really hard to program emotions into a Ai along with the storage space within a panel Mainly it’s just not efficient, useful, or helpful. Really it would be irrelevant and too hard to implement to get anything out of. When the Ai is mentioned I’m thinking it as more of a AGI than a narrow Ai It’s very hard to implement emotions/ways of expressing said emotions If from a master server then it could be possible but quite useless. I was talking about it from a single panel point of view without a server
I love how much effort was put into the details for each level in Portal. Everything feels quasi dynamic and interactive like in real life. These panels especially help.
My favourite one of Aperture’s investment gadgets. I feel like the animators don’t get enough credit for all the panel movements seen through the game. My favourite is during the core transfer segment. The panels in GLaDOS’s chamber are tied to her emotions, and when she gets replaced by Wheatley, they all go limp, reboot and recalibrate to respond to his emotions. Really puts into perspective how much control over the facility you’ve just bestowed to Wheatley.
Why is no one talking about them cooperating? I'm sure they need AI primarily to cooperate with each other so they take into account the movement of their neighbors and don't produce collisions or lock-ups.
I want them to be a reality! Instead of stairs, I'd just stand on one of these and get lifted to the 2nd floor of my home. I get sore muscles a lot because I have a heavy backpack to deal with. Please make my life easier!! ;-;
I've been wondering this since the game came out. I see no reason why it would be impossible. There might be practical problems though like how to get them all enough power and getting that much equipment onto a network.
The Best Strucutral part of Portal 2 is the Facility itself, (excluding the Heavy-AI Based Robots/Cores/Turrets) it completely consists of panels that have personality you could just say the entire facility is Alive in a way i would love to have a game that is all about this concept
Funtime Freddy the gel ain’t possible but portals are by using quantum energetic dark matter causing a mini wormhole causing 2 interdimentonal portals that can be reset by relay technology
Two main reasons: the concept of a fully-autonomous artificial intelligence system with at least a basic personality style still has yet to be perfected, and packaging one into each panel's subconstruction and linking them all to eachother and a central network with today's technology and infrastructure setup simply isn't possible. AI is still quite a long way from being fully-functional, especially for something like this. Thanks for asking!
@@TetherAnimations9999 i know the AI part is impossible, but fitting the technology is not, for example raspberry pi's (a single board computer) would very likely fit into the panel ( at least the pi ZERO W should.
In the future yes now no splitting the nuclear atom was only the first step we have to squeeze a black hole into the gun and that could take a lot of time
It's the people like you on RU-vid that actually bring interesting videos onto RU-vid. Everything nowadays just seems bland and copied. We need more original creators like you on this site.
Great to see some more videos. I have a dream of making a science facility similar to aperture science, and these are one of thes things that I'd love to make.
My favorite part about these little dudes is how they are used in the core transfer scene, in the starting trap room they don't all move at the same pace or in the same way, they look at you, Wheatley and GLaDOS like an audience, and there's so much more. Another reason why this game is amazing.
Remember that one panel that was just super angry that it couldn’t fit into the space and was slamming against the wall in a fit of rage? Yeah that was my favourite one.
Man you can animate really well because i cant do that, i am trying to make animations with source filmmaker but i suck, so greetings for great animations!
I like this moment when crusher (what have construction like a panel) crush the panels and in few seconds later subs said "let's make them real". Ok, i'm really need one crusher, so yes let's make this spiky things real
Something I've wondered about as well, but haven't had the time or energy to build one with it's relatively limited usage in like... My house. If you want leads to track down, I'd look into robotic arms like others have suggested, assembly line robotics and such, *but* if you want a stronger lead to go down, look into Inverse Kinematics for a model on how to make them move. It's a way to sort of mathematically understand what all the linkages are, and then control them all just by determining the position of your tool, which, in this case, is the actual panel itself. I don't have a lot of in depth knowledge to share, just enough to understand the gist and know that it's exactly what you should be looking at here, but if you get IK working, and get it accepting coordinates, then your "AI" would just be as simple as determining what location you want it to be in, and then telling it to go there. If you wanted it to act as an elevator you could have a pressure switch on the platform so that when something is placed on it, it raises itself up. Etc. Have fun, do something cool. 3D printers are cheap these days, and arduinos are cheaper than a nice chocolate bar.
![[SFM] Portal/2 Panels: Are they possible?](/img/logo.svg){kind=logo}
