Self-Reconfiguring Modular Robots GO!

was really cute. I haven't seen it either, but Jonathan keeps talking about it, so I think we need to go see it. So before I get into what today's topic is, specifically, the thing I wanted to say about Big Hero six is the movie opens. It's set in me an alternate future really, because you have the main city is San Francikio, so it's a kind of a melding of San Francisco in Tokyo and are just world design.

But the thing that really impressed me early on is that there's all this kind of science fiction ee sort of stuff going on, but a lot of it, at least has its basis in science. Fact, a lot of it is is based on you know, take take sort of the cutting edge stuff that's going on in labs right now, projected out several years, and assume that the research you know plays out and that everything is fruitful, and what do you get? And a lot of the technology you see, at least in the early part of

the movie seems to go along with that. And I was just really impressed. I mean, everything from soft robotics to what we're gonna talk about today, which is self reconfiguring modular robots or as you also noted that they can be called in your script for this video episode, autonomous kinematic machines of variable morphology rolls off the tongue. So what on earth is that talking about? Well, I

think we should start with the idea of the variable morphology. Okay, So something that changes shape, right more fall as he refers to like structure or form or shape as it would in biology. So so what we're talking about is a transformer, that's right, Yeah, it's a shape shifting robot. And why is a transformer? Great, well, a transformer is it's a robot in the SUI, A transformer is great unless you're talking about one of the recent movies, a

cartooning transformer or a real transformer. Either one of those would be great. But it's great because it can do two jobs, whereas most robots can only do one. So a regular robot might be an autonomous car and it can drive down the highway distributing palettes of Captain Crunch. Excuse me, Captain Crunch the hyphens. I would really hope you'd be an admiral by now. But then, of course, after that autonomous truck delivers them, it can't get into

big wrestling matches with other robots. Now, you might make a robot that can get into wrestling matches with other robots and stabbed them through the side of a highway with a giant sword. But that robot can't be a truck unless it's a transformer, right where it can be both you can. But but wait, let me do you one better. What's better than a transformer that can assume two morphs? A robot that can assume even more morphs? What about a robot that can assume an infinite number

of morphs? So you're saying, like a robot that is capable of assuming any shape that you need at any given time to perform any given task. Bingo. Well that is something that is is in Big Hero six, the the hero whose name is Hero actually designs self self self reconfiguring modular robots. He designs a tiny robotic unit that, when it connects to others of its kind, make larger shapes. And you might think, well, that's pure science fiction, but

it's not. And we want to kind of go into more detail about why such a thing would be a desirable object in the first place. And you you really hit on it, Joe, the idea that robots today tend to be really good at specific things and crappy at everything else. Yeah, they have one job. You had one, Joba, I asked you to paint a masterful painting. You had one job, and all you did was clean the floor. Yeah, that's exactly right. You know, we do. We design robots

to perform specific tasks, and so the form follows the function. Yeah. In fact, there almost seems to be a direct proportional correlation where the more diverse a robots skill set gets, the worse it gets at all of those jobs. Instead of can be really good at one or very mediocre at several times. This is like that TV trip about about ninja and that if you have one ninja, it's great, and if you have a hundred ninja, they have about

the same capacity. Is that one ninja would have just spread out amongst all of them, right right, That's why the tick can just breathe right through a horde of ninja. Yeah. Yeah, besides the fact that he's also ni invulnerable. Um yeah, they the this is that's exactly right. Yeah. That you build a robot that you want to be able to do lots of things. Usually it can do a few

things and not so great. And this is you know, anyone who's who's played with any of the robots that are social bots that are meant to do like these things, like mostly in the toy department, you know, they're amusing, but they're not exactly the best at anything that they're trying to do, Okay, And there's solid design reasons for creating robots that are that that only have one job, right, I mean, creating robots is kind of expensive and time consuming,

and so if you're trying to build lots of different functionality into them, that's that's a lot more expensive and a lot more time consuming if you if you're narrowing down what you wanted to do, right, if you if you pinpoint that focus, then you say, here's what the robot has to be able to do. Anything that does not directly contribute to this task, we can eliminate. That means that you can, you know, really pinpoint that focus, really make sure that that design is as efficient as

possible at doing that task. But let's imagine a scenario. Okay, let's say that you live in a house by yourself, and you work long hours, and you need a robot to help you clean the house. Beyond beyond something like a roomba that's just going to sweep the floor. Sure, a roomba might be really good at sucking up dust off the floor, that's its one job. But a roomba, of course, cannot clean your gutters, and it can't well I don't know, I was gonna say, it can't soap

and mop the floor. But they might make rumas that can do that. They do. They do make mopping room bas but they're not usually also sweeping room bas So it's in other words, you have you have two different robots that have similar but different jobs. Okay, and one about another one thing. Yeah, another one for washing windows, and another one for doing your dishes. And there there have been lots of robots that do these things, but again they're all individual like their uni taskers, as Alton

Brown would call them. So are you going to fill your house up with thirty five different kinds of robots? Well maybe if you can afford it and you don't mind occasionally getting sharp metal jabbed into you. Well, also, you just you started this up by saying you live alone. I mean that maybe they'll make you a little less lonely to have like thirty five robot friends doing all your work for you. That would make me more lonely,

would it? Yeah? I just named them. I assume if you live some with somebody else, you can bully them into doing all the houses. Is that how it works in your household? Kidding? My wife is wonderful. Yeah, I do the same thing, Joe. Uh yeah, so this is this is exactly the point you're making, though, is is

very much on target right. The idea that it would be very beneficial to have a mod geiller robot, a robot made up of smaller robots, smaller pieces that could form any given shape to perform any task as best as it possibly could, rather than have a veritable army of robots. I mean, even if you were to say, well, let's imagine that I've gotten enough space for all those robots and I would want them all. Just the power bill alone to keep them all charged would be ridiculous.

So what if you could have this group of tiny robots that joined together to form a much larger moving object that can do pretty much anything you wanted to do? Right, So, it could form into a small disc and hoover along the floor like the room but does. Or maybe it could make itself tall, rearrange itself into a long stick with arms to clean the windows, or to reach into

the sink to clean the dishes. Or it could even climb over upon itself up the wall, up the side of your house to reach the roof and clean out your gutters. Yeah, yeah, all of these are I mean, it sounds kind of science fiction e but these are all real things that, uh that engineers are looking into. Maybe not necessarily those use cases, but but essentially the idea of a robot that can change its shape in order to do different things. Whether that's just to get

across a landscape or to actually perform a specific task. Right, Well, I think we should talk about some of the existing models of reconfiguring or self reconfiguring modular robots that exist out there today, right because obviously we don't have anything like what we just described, this perfect self reconfiguring robot in your house, but we we we've basically got a

few steps in that direction that are very encouraging. Yeah, we've got to that we're specifically going to talk about, one of which is a reconfiguring robot, one which doesn't necessarily reconfigure, like it doesn't form a larger robot, but it can act with a swarm intelligence, which it will be important in modular robotics. So we're gonna talk about those. But I should point out there are lots of different types of modular robots out there. Some of them are

not self reconfiguring. Some of them are simply you get all the different components and you put them together. Yeah you can. You can sit there and make you know. There are a lot of ones that teach children the basics of robotics, and they have different units that do different things, and the way you put them together determines their function. Uh, these are modular, but they don't self reconfigure. They don't they don't manage to put themselves into whatever shape.

You are the one responsible for doing it. So we're not really going to talk about those. Okay, Well let's look at M I T S M blocks. Yes, they are momentum driven magnetic modular robots. That's how the commercial would go. You know, I imagine the guy who did the micro machines ads doing it. Uh no, no, I know. How the commercial would go would be like, let M into your life. The M block, you know, M I

T you can just send the checks are away. So they're cubes that they you would think because they're called blocks, Joe's already regretting absolutely going to stay in. Okay, So the M blocks are these blocks that have magnetic corners, right, so each corner has a beveled magnet in it, which is pretty cool. I'll talk a little bit more about that in a second. They also contain a rotating flywheel that can move up to twenty thousand revolutions per minute.

It's pretty fast, and so, uh, they don't necessarily speed turn at full speed every time, but depending on what you need the cube to do you might have it speeding spinning that quickly, and then what you do is you apply a break. Now, the laws of momentum state that that energy has to go somewhere. It gets transferred to the cube, and this makes the cube move. Okay, so what does this actually look like when we're watching

these cubes move around on each other? Have you ever seen a ghost movie where objects seem to be moving on their own accord with nothing obviously moving them forward. Yeah, that's kind of what it looks like. It looks like the blocks are possessed, kind of like kind of like the toaster jumping and Ghostbusters too. Yeah, yeah, exactly. So I don't know if you already said this or not, but they don't have external moving parts. No, it's all

it's all kept inside. The flywheel is actually contained within the cube, right, So the momentum can move one cube onto another cube and reposition it around the outside of that cube without attaching and reattaching via clamps or hooks or anything like that. But it's it's attachment through those magnets, right. Yeah, And so if you imagine one it's just one cube

moving around another. But if you add a third, you could have two cubes moving around on each other, and then the fourth or fifth, and eventually you can end up with a large mass of moving modules that can assume shapes, especially if they're in communication with each other. And you can also include modules that don't necessarily have the flywheel inside them. They might have some other UH

specific sensor or camera in them. Uh. The m I T Group has talked about the possibility of that as well, in which case the ones that have the flywheel would be responsible for moving the other ones, which they can do because that flywheel spinning so quickly. Um, these things can actually jump, so they don't just like the the

videos I've watched show the cubes. I guess rolling is the best way of putting it in order to move across the floor, but they can even Let's say you've got two cubes stacked already vertically and you have a third cube roll up. It can actually spin and break in such a way to jump up and land on top of those two cubes, which is kind of spooky

looking when you watch it the first time. They also, of course are making pretty loud noises because it's a flywheel spinning, and then breaking, so it's it's not silent, it's it's loud and spooky, so but it's pretty cool. And and those those magnets that are in those beveled points in the corners, they rotate. And the reason by the magnets rotate is so that you can always get

a north south connection. Remember, you know, the opposite poles attract and similar poles repel, and you want to be able to make sure that when one block approaches another block, it's able to actually latch on properly. So the magnets in those individual points will naturally rotate so it will be a north south connection so that any two cubes from any two sides can attach exactly. And if you get two poles that are the same, then one of them will end up rotating. Everything will be fine. So

there's literally no wrong way to put them together. Neat. Yeah. Yeah. These also remind me of another project out of m I T when we've talked a little bit about before, previously, before previously, which is the common way to really over emphasize the point um. They're the m I T S reconfiguring robots from their self assembly lab um while researching what's called forty printing. Um. That's three D printing of materials that can then move and change with with minimal

stimul lists or instruction from the outside. Uh. The kids of the Self Assembly Lab created some macro scale robots to help demonstrate the usefulness and honestly the coolness of this changeability of of structure and so so these are these reconfiguring robots are programmable folding chains constructed of of nodes, each of which is hardwired to one or more neighboring nodes.

So if you give one note an instruction, it will pass information on how to move down through the chain until the programmers desired shape is is complete with with relatively minimal input from from that programmer. Right. If you do not remember us talking about it, or or if you want to refresh yourself, you can learn more about these and other self assembly techniques in our video and podcast, both of which published August. That's like over a year ago. Guys,

we've been doing this a minute. Um. The episodes are called the video is forty Printing is the Future of Design, and the pod cast episode is forty Printing is One D Better. I must have titled that one, actually, I think I came up with that line. Strangely enough, Yeah, very it's a very Jonathan kind of joke. So the

other one we wanted to talk about was Harvard's kill abots. Now, this is the one that is not modular robot but they do demonstrate swarm behavior, which again is going to be really important if you're talking about ultimately, if your if your goal is to create a robot, a large robot made up of smaller robots, then you have to build in some form of intelligence for each of those smaller robots to know quote unquote its place and it's

uh and it's function right. And so there are some similar concepts going on between these two uh, these two ideas swarm robotics and the modular robotics. So with a self reconfiguring modular robout, we're generally imagining something that's connected to itself. It's got pieces that may all be the same kind of peace moving on one another to reshape itself.

With swarm robotics, you've still got lots of pieces working together, but they're sort of moving on mass and they might not be building a shape, right, they might they often aren't physically connected together at all, right, they're all individually free to move. If if somehow one was to break free of its programming, it can make a break for it. Uh,

it's not. It's not chained physically in any way. So whereas the ones from M I T were magnetic and linked together, and we know of a lot of others that have physical connections of some sort, either they have connectors that snap in or they have clamps that hold onto one another, these do not. These are simply to allow UH computer scientists to test various algorithms for robotic intelligence, for swarm intelligence. UH and a lot of them are looking into bio mimicry, the idea of of how animals

and swarms behave in a mass. You know, it's not it's it's really fascinating stud At any rate. The problem with these computer algorithms was that for a very long time,

there wasn't an easy way to test them. Like you might come up with what you think is a brilliant algorithm to to inform a swarm of robots how to move for any given purpose, and you can test that through computer simulation to your heart's content, which is not ideal, right, I mean, because you're depending upon the fact that the computer simulation has to be accurate enough to actually reflect

what a physical object would do. So Harvard scientists came up with this idea and professors as well came up with this idea to create a relatively inexpensive tiny robot on which you could test these algorithms by by assembling enough of them to do whatever it is you need to do. And they look like, well, they're little circles

like think of about the size of a quarter. But they are on long, spiny legs and there's a little motor, a vibrating motor on them that depending on how it's vibrates the assuming that this is on a level, smooth surface, the robots will skitter across the ground. Now, their their legs aren't moving, right, it's just through this vibration, so it's not like they're segmented legs. It's not that much like an insect. But they do jitter so that they

move in a specific direction. Yeah, well, these things certainly do that, And there's some great videos online of these moving around in various ways. Usually it's a direct overhead shot, so you just see a little it looks like a hockey puck, just kind of slowly moving across an area, so you can even make these yourself if you really wanted to. I'll tell you more about that in a second. But the way the way the kilbots form that shape

is all dependent upon the algorithm. Assuming that the algorithm was was properly constructed so that the robots know what to do quote unquote um, that they are able to detect one another and move according to whatever the plan was,

and they communicate through various means like there's uh. The version I saw, they were specifically using infrared emitters and sensors so that they would be able to tell how far apart they were from their neighbors to make sure they maintained whatever the distance was that the algorithm dictated. So one of the examples I saw was a follow the leader um example, where one was designated the leader.

They had actually put a green sticker on top of it so it made it easy to follow on the video, and then the next kill abot had the instruction to try and maintain the same distance with the first one. The first one's instructions were move wherever you want, but you can't get too far away from the second one. The third one's instructions were tried to keep up with the second one, but don't get too far away from

the fourth one, etcetera, etcetera. So the longer the chain is, each one essentially has the same instruction maintain as as you know, as as a stable a distance as you can to the one in front of you without getting too far away from the one behind you, and it would follow the leader this way. And it was a cute little demonstration. They also had a foraging demonstration where they had one kilobot designed to look like uh, well, it just had a red sticker on it. Those designed

to act as the nest this is home base. And they had another kill abot that was pound off maybe to three ft away, that had a green sticker that represented food. So if they wanted to have these robots go out and search for So we're teaching robots to be cannibals. Well, it's much better than teaching them to hate humans. I'm not worried about the killers. Things skitter at me all they want. Yeah, they're only called killer bots.

And how one letter off from kill bots? Yeah, so uh yeah, and and well, at any rate, so the way this would work is that the once they got the instruction, they used their infrared sensors to look for the quote unquote food kill and once they found it, then they would travel back and forth between the food one and the next one. Uh. Some would just end up forming kind of a structure around the nest. Uh. Once they were unable to find the food, they just

kind of stayed still. They wouldn't continuously move around, they would only go so far away. And that was a kind of cool demonstration as well. Now, like I said earlier, you can actually get the plans on how to build one of these yourself, or lots of them, since one would not be terribly useful. You can't really test swarm intelligence on a single robot. The it's difficulty. Yeah, all right, just pretend you're a lot of people. Uh. But the

whole design is published under Creative Commons. It's just a non commercial commercial license, so you can't you know, you obviously can't make them to sell them, but for educational purposes you can make as many as you can afford to build. And I say afford to build, because obviously you have to buy all the hardware to put these things together. Though, you know, if you're building them in bulk.

Obviously the cost will go down. Yeah, although it's still pretty expensive because if you were to build say about a hundred, they price it out as being about fifty dollars per robot. It's pretty expensive. If you were to build a thousand, it's twenty dollars per robot, But that's twenty tho bucks when you take that all into consideration, and I don't have twenty tho dollars to build my robot army yet, so uh but at any rate, again, these kind of test out that idea of swarm intelligence.

More recent videos have shown them do things like assume specific shapes, like they give it a command to for the all the robots and I think there were in this case more than a thousand of them in the demonstration I saw, but to form a shape like a star. And so the robots have to have information about which other robots they're supposed to be next to and how they're supposed to be configured based on that, which is really fascinating. You think, you know, this thing is again

about the size of a quarter um. I mean, obviously it's thicker than a quarter is, but still they're able to pack in enough of the microprocessor technology for the robots to be able to determine, oh when when they we make this shape, I'm number fourty seven. I have to make sure I'm next to four and three, six and four and thirty eight. And they do kind of go one at a time to make this shape you like it was it was like in a line, as opposed to all right and break didn't They didn't do

it like that. But it was really interesting stuff. And again, this is the kind of behavior we're going to have to see for something like the modular robot that's in Big Hero six to become a reality. Well, I think this stuff is awesome, and I want to know what's currently in development. I don't know if you've looked into that, Jonathan. I looked into sort of the most recent literature, because this is one of those things where we still have

people working on implementations. No one, I don't think anyone has come up with the ideal implementation yet. It's it's of course not I mean, we're in the we're in the research phase. Yeah, exactly. This This is a developmental era, which kind of makes it even more exciting really, because you know, there's there's so many people looking at different ways of doing this in different tasks that such robots could do. That it really starts to fire your imagination.

So for example, here's here just some titles of some recent scholarly articles that are on the subject. Uh and and they kind of demonstrate now and like by recent, I mean these were all published in two thousand fourteen. Uh, these kind of demonstrate the the breadth of the research design of modular robot system for maintenance tasks and hazardous

facilities and environments. Well that that kind of you know, you can easily imagine having a modular self reconfiguring modular robot would be incredibly useful in that source of situation. Or how about room bots. A hardware perspective on three D self reconfiguration and locomotion with homogeneous modular robot. So here we have a specific scholarly look into well this is you know, we we know what we want to achieve.

What are the way is to do it? Using this specific approach where you've got robots that are all essentially the same shape that combined with one another, then we have design principles and testing of a latching modular robot connector So again looking at different ways for these different

modules to join up exactly. H then you know the modular robot joint design and experienced verification for small openings and enclosed space, which again really awesome because you're you're looking at the possibility of using these robots that could go into places that you know, if it had a solid form factor that didn't change, it couldn't get there.

So imagine that you have to send uh, this robot into a building that is you know, perhaps it's a really old building and it's got U faulty structure, you know, structural damage, and you want to make sure it can navigate all the way through and actually investigate that structural damage.

That could be really useful. Sure, yeah, if there's a collapsed path or something like that and you need to get under it or round it or whatever that it is, that that is right right, being able to break down into all of its creepy little robot parts and skirt her on through is terrifically useful. You know, it's stasis

on the terror. You know. It strikes me that there are sort of like two main ways that this technology can move forward, and really both of them seem crucial, but one of them is about working on the modules themselves, like making each individual piece, and we're assuming we're working with robots that are made of pieces that are all identical pretty much, Uh, that making each individual piece cost effective but very versatile, able to move around, sturdy, able

to communicate, without being a gazillion dollars apiece. But then also talking about the overall swarm intelligence, how do you program the robot through these modules to have some kind of collective brain that enables it to do its work effectively? Right? Right? Yeah? And this is I mean, these are a lot of unanswered questions right now, right can could there be an emergent artificial intelligence from the swarm intelligence that would allow you to essentially tell it here's what I need you

to do, you figure out how to do it. Or would you always need to have some sort of controlling system that would essentially send instructions to that swarm of robots to form a particular shape to complete that test.

Oh yeah, that's an interesting approach. I hadn't even thought about that because I was just imagining we're assuming this has to be autonomous in one way or another, but you could just I guess, have a laptop that's bluetoothed up to your modular robots swarm, and this thing is reconfiguring itself to do work based on orders that are sent to it from the software on your computer. Yeah. Well, I mean when it comes down to it, the question

is how autonomous does it get? Because there's some autonomy that is necessary, right, because if you're talking about thousands of robots, you don't have the time to tell every

single individual robot where it fits in this grand scheme. Sure. Sure, well, I mean you know, if you think about something like like having a a lead butt with a green sticker on it or what have you, um, that you can bluetooth up with and give a command too, and then have it break that down into commands to give to everyone else in the swarms that you don't have to literally contact every single microbot, right. I think there will

be a bridging system. Yeah, I was assuming the commands would be given by software, not like manually by well well, sure, I mean you know, no EI either way though, Um, there are lots of different factors of trying to communicate with a swarm like that, and how you would go about making it the most efficient, and it's kind of like when we talked about neural networks, how every time you add a node, it complicates the entire structure. I don't want to say exponentially, but quite a bit large fold.

So same sort of thing with modular robots. The more modular robots that are in the system, the more complex it's going to be to command it. But that's what you know that swarm and eligence is all about. So that's really exciting and uh, you know, kind of looking into the future, like let's assume again kind of like Big Hero six did. Let's assume that this this research pans out and that we come up with some implementation. It doesn't you know which which one is immaterial for

this discussion, but some implementation that works. And then in the case of Big Hero six, Uh, it was all magnetics that held everything together, and the individual modules look like a sphere with two little uh protrusions that stuck out that could that were a kind of hinge. They could move independently of each other. So uh, that's what the individual component was, and then together they could form

pretty much any shape you you wanted. Um, and they did so using a brain computer interface, which was really cool. Is essentially a headband you could wear where your thoughts get transmitted into computer commands that tell the robots what it needs, what they need to do. Uh, that's awesome. Anyway, we're quite a ways away from that too, But people are working on brain computer interfaces just as they're working

on this modular robot issue. So you would want, in an ideal implementation to have your microbots as small as you could possibly make them, right, because that that would increase their versatility. But also, you know, you can think of it kind of like an image of digital image and in pixels. You know, if you have smaller pixels, you have better resolution. The picture looks better as opposed to large blocks of color than it looks blocky and

and jagged. Same sort of thing with your modular robots. The smaller the unit, the more smooth the robot's going to be, the more versatile it's gonna be. Um, but

that also comes with some pretty big issue problems. But another one is that this ideal implementation would allow it to move through those small spaces we were talking about, whether individually, you know, so it splits apart and all the individual pieces creep through an area, or it forms some other shape, like a snakelike shape to get through certain areas. I've actually seen some really awesome snake bots

that are incredibly disturbing to watch on video. Uh, we'll probably need to power these suckers with something that we haven't thought up yet. Yeah, this is a real problem. Oh sure, you know, well, any significant pushing or pulling or lifting or other interaction requires really a lot of

power right now. Hydraulics are really the most efficient systems for for that kind of stuff, and they're you know, for example, what what drive the most badass exo suits for for lack of a better term, um, but they're also really clunky and don't really lend themselves to this microbot kind of format because you need like a tank of fluid in order to drive. I can't imagine having like super small lines to each microbot and still having any kind of degree of freedom. Yeah, it would be problematic.

Yeah yeah, so I mean, you know, maybe the answer to this is that we were not going to use this kind of system for any kind of heavy lifting. Um. But even so, just powering small things that you want to keep lightweight is a problem. All the time. We

talk about it always add weight. So one other challenge, and we've talked about this, is the fact that we need to have these these individual units be able to communicate with each other at least on some basic level so that they quote unquote know what they're supposed to do and where they're supposed to go. And that's where we're seeing this this work with, like in the Harvard

kill abuts of of Swarm Intelligence. It's it's one thing to see the work that's being done with things like the the M blocks where you're seeing maybe six or seven of the units working together. It's another thing when you start talking about you know, large collections of these devices. We talked about that a little bit before too, in our episodes about ants and also about biommicry as a

whole um and also in our upset about right emergent behavior. Yeah, these are all things that you know, we don't have answers for because we're still asking the questions. But it's really exciting again to see the work and while knowing about the stuff because you know what, we've researched so

much into robots and artificial intelligence just for this show. Uh, the not not just this episode, but the forward thinking series overall that when I went in and saw this movie, when I was watching Big Hero six, That's why I was like, I had a huge grin on my face because I was thinking, these are all the things we

talk about on our show. Uh that where we are now in a world in this movie where all those things have been realized, where all the problems and challenges we talk about have been overcome, people have figured out the way to engineer a system that meets those challenges, and to look at the potential of that world was phenomenal. I was like, Ah, it's like I'm watching the outcome of a forward thinking podcast where we have that bit where we say, what will the future like? And this

is what the future would be like? And it's it's an amazing vision of that future. So um yeah. The first ten minutes of that movie, I was like, man, I wish they had called me because this is awesome. It feels like it's the outcome of a forward thinking show. Um but you know, and then of course it gets into more fantastical elements as it goes on, and and the science and technology get a little more loosey goosey, but they had already won me over at that point,

it didn't matter. So you have a hard heart for scientific implause. There's I can get pretty nitpicky. I admit it wouldn't happen like that. Yeah, yeah, well, I mean they're there. Definitely comes a point in that movie where you gonna say, all right, I just gotta go along with the ride Bernell or I want to be it's it's it's worth seeing you guys should definitely see it. And of course, you know the Smudgler robot research. We're going to keep an eye on it. It's fascinating stuff.

We're learning so much about robot intelligence and robot mobility through this kind of stuff that it's just a fun thing to keep our eye on. So, guys, if you have any suggestions for future episodes, maybe there's something else about robotics you've always wanted to hear about. Obviously we don't shy away from that topic. Let us know. Or if there's anything else you know, just something you're wondering about, what would that be like in the future, send us

a request. We read all of them, we're happy to get them, and we would love to hear what you think. Send us an email addresses FW thinking at how Stuff Works dot Com, or drop us a line on social media. We are on Twitter, Google Plus, and Facebook. At Twitter and Google Plus, we used to handle f W thinking. Just search fw thinking and Facebook. We'll pop right up. Leave us a message, and we will talk to you

again really soon. For more on this topic in the future of technology, visit forward thinking dot com, brought to you by Toyota Let's Go Places,