Episode 108

The History of Robotics

Watch on YouTube

About this episode

In this episode of the Untangling Web3 podcast, we dive into the fascinating and ever-evolving world of robots, robotics, and their intersection with artificial intelligence (AI).

From ancient mythology to cutting-edge tech, this episode explores how our visions of mechanical beings have transformed from mythical automatons to industrial arms, humanoid assistants, and autonomous drones.

We trace this journey from Asimov’s science fiction to real-world innovation and ask: what happens when automation meets ambition?

Key highlights

  • The Origins of Robotics and Sci-Fi Inspirations. The episode kicks off by going way back—literally—to ancient Greek myths like Talos, imagined metal giants powered by magic, and traces that lineage through history to literary visions like Karel Čapek’s play “R.U.R.” (which coined the term “robot”) and the iconic writings of Isaac Asimov. Asimov’s Three Laws of Robotics are highlighted as early governance frameworks that still influence today’s discussions about robotic ethics, AI control, and coexistence with machines.
  • Industrial Automation and the Rise of Robotic Arms. From Henry Ford’s assembly lines to the first robotic arm patents in the 1950s, the episode chronicles the practical evolution of robotics in manufacturing. We cover the leap from basic machine tools to robotic arms capable of precision and autonomy, used in everything from automotive assembly to pharmaceutical packing. Japan’s early dominance in this field, driven by labor shortages and technological innovation, illustrates how robots revolutionized modern production.
  • Modern Robotics, Humanoids, and the Role of AI. Fast-forward to today’s AI-enhanced robotics—this includes everything from Boston Dynamics’ quadrupeds to Tesla’s Optimus and autonomous drones. The conversation highlights how perception, feedback loops, and decision-making make modern robots fundamentally different from their predecessors. We also touch on real-world applications like healthcare robotics, military drones, and the promise (and risks) of placing humanoid robots into homes and public environments.

The world of robotics is no longer confined to sci-fi or factory floors. As AI supercharges automation, robots are becoming more intelligent, mobile, and integrated into daily life—from hospitals and homes to warehouses and war zones.

But as machines become more autonomous, questions of control, ethics, and purpose become increasingly urgent. Whether you dream of robotic butlers or fear the rise of machine overlords, this episode offers a comprehensive look at how far we’ve come—and where we might be headed next.

This episode is sponsored by the VeChain foundation. Learn more about VeBetterDAO here: https://vebetterdao.org/

Transcript

Auto-generated from the episode audio.

Welcome to the Untangling Web 3 podcast. In this episode today, we'll be diving into the world of robotics. Finally, not as sci-fi characters, but as real machines already shaping how we work, live, and interact. We'll trace their journey from Isaac Asimov's imagination to today's global rollout as we see governments and tech giants pouring billions into robotic development. So, what happens when robots become co-workers, caregivers, or even companions? Let's find out. Jack, are you excited for this one? Yeah, I'm excited. I'm a little bit daunted by it as well.

I full disclosure, this is going to be the episode where I'm very much the layman in in the in the conversation because Alec, it's it's your baby, right? You love robotics. You've got a background in it and I don't know a whole lot beyond the sci-fi uh you know what you see in the movies and books and things. So, it's going to be interesting. I think I'm going to learn a lot hopefully and you're going to you're going to teach me, right? Yeah. I mean, I think I I love this work.

I love this topic. I love this area. And I think that so many layman's also do because it's very accessible. Like when you talk to even a software developer, it's very abstract. When you talk to someone who's not in software about software, it's even more abstract. But robotics, it's so physical. Everyone can kind of understand the value proposition immediately and is either horrified or glorified or, you know, whatever kind of thinks of it in a positive light or a dystopian light. So I think yeah, this is going to be something that resonates with a lot of people anyway, the topic whether you hate it or love it.

But the fact that you said that, you know, your only knowledge maybe is the sci-fi novels that you've read is super important because I think that is kind of the defining the art of possible. And I think so many roboticists like me included read the early novels like Isaac Asimov's work and that inspired them. And you see so much of like what people are pushing towards today being based from those novels. And it's kind of like a self-fulfilling prophecy, which I think is really interesting. And that's why I think it's one of those things where we have to start at the beginning to kind of, you know, talk about the origins of the word, talk about some of the early literature, and then we'll go into what's actually happening today.

Yeah, we're going to start this way back in the beginning, right? Like ancient history type stuff, go back to the to the real theoretical history. So yeah, I'm I'm excited for you to take me through this. This is like a proper um kid dressed himself for school where it's like this is like my two favorite topics is robotics combined with a bit of ancient Greek history. Like I am like a kid in a playground right now. So if we go back set your mind back, you know, 2,500 years or something like that, like long before robotic robots were built, we imagined them.

And I think even if you go back to like ancient Greek literature, the idea of Hisus, this god of blacksmiths building this giant effectively automter or robot called Talos. And you know, I think it's quite weird that we kind of describe this as a robot because so many ancient cultures would talk about, you know, kind of organisms or creatures that were made from stone or made from greenery and made from trees and all this kind of stuff. But there's something very distinctive about making things from metal.

And I think it's the fact that it's kind of inorganic and it's kind of like seems like a human creation and there's kind of like an immediate affinity to there being something different about creating a creature or a a a thing out of metal because we have this kind of affinity of metal being this man-made substance that isn't naturally occurring and maybe has like unique and magical properties because of that. Yeah, it's interesting that what you kind of saying here is the one of the origins of the the kind of concept um being kind of related to the Greek gods, right?

Because I think a lot of people see robotics as being humans playing God, right? It's almost like we're we're trying to create things ourselves now that are as much in our own image as possible. Um so it's interesting that it does tie back to to the Greek gods. I mean what was the what do you know anything about like the kind of reception of the of that concept like was it was it in like a kind of religious uh sphere or was it is it intended like what what's the I guess what's the impression of this uh this this early idea that was coming out my understanding is that it comes from like a lot of early cultures derive it individually it's not like there's one source and that kind of spans out to everything that happens in a lot of places and I think like the link is kind of to what I said earlier where you know imagine being a prim primitive human like 3,000 years ago or something like this so much

imagine so much of the world you know just go to rural Wales and you'll get it so much of the world is like organic and you just it's there but then metal is this thing that seemingly we created for ourselves and is you know is associated with so many kind of distinct properties and characteristics because of that a lot of people associated metal with like magic like properties or kind of like you know things beyond like the normal normal realm. And I think that like the association of also being like godlike and like hubris and all these things about you know the more technologically advanced you are, the more powerful you are, the closer to God you are and all this kind of stuff.

So I think there's kind of two distinct camps early on that metal is this special thing and the idea of creating something in one's own image in the same way a lot of religions believe God did. Creating something in one's own image but maybe superior or better like made out of metal because we can craft metal. we can't necessarily like freely craft organic material in the same way um is quite it's quite unique. But then the other side is like the idea of hubus and like you know one's own creatures kind of destroying the person and things like that.

And I think that often comes up in literature as well. Yeah, it's it's interesting. It's like it's tied to the idea of technology and its superiority because obviously we're always looking to to mechanical means almost or or physical means to help us achieve better results, right? Like you know as as kind of you shown me in the research like ancient Egypt you have things like hydraulics and pneumatics being initially designed, right? So ways we can harness the physical world to do things we couldn't do before.

Um, so yeah, and and that the marriage of technology and deism and like gods and things is I think there there's a lot of echoes of it today as well. Like a lot of people treat the big tech giants as and and the the founders and CEOs as kind of the gods of capitalism now. So it's it's interesting that it goes back there and we're definitely going to get to that. And I think like so the other there's that one idea of like this kind of humanoid type thing that's been manufactured is not human.

And then the other things that you mentioned there is the idea of like tools and automation like there's an obvious desire early on across almost every culture to create and advance tools and on the one hand to like be an extension of a human like you know spears bows all this kind of stuff but on the other hand auto to automate and maybe replace certain functionality and that happens very very early on you mentioned there in like ancient Egypt they use hydraulics and pneumatics to open doors they did in the temple of thieves this happened in ear in early Indian culture and early Chinese culture to the idea of automating processes and enhancing processes using hydraulics and all of that you're kind of saying is like part of the origin story of like robotics machines like the development of this kind of like field early on.

One of the things that we have to really mention is Al Jazeri who was an individual from Mesopotamia in I think it's like 1000 around the 1,200 1,300 something like this and that was one of the earliest published books that really related to the idea of like automation robotics and in that he created like basically a hundred different mechanical inventions like a lot of them just theorize and you know kind of like Da Vinci style where just thought about a lot of different things and wrote them on paper but the idea of like robot musicians drink serving automation powered clocks.

Some of them he did actually create and like using like smart cam shafts and valves to create these automated machines which at the time were like magic to a lot of people but now are kind of quite widely understood. I think that the main point that I want to capture in all this is that the ancient world gave us the idea of robots like kind of not as enemies but as tools and like magical servants that could kind of go beyond what we could do early on.

Yeah. And so basically ways of replacing things that I well either replacing things that were being done by humans before and you know often through I guess things like slavery probably right for building lots of early things uh and the other field being how you how you can achieve things that you couldn't do before right I think that's kind of the two main the main strands of of how that links to robotics in my head then I'm glad you you use that word and maybe that's like a nice inroad to the next bit so moving away from like ancient history and literature to like more modern literature that again inspires a lot of the work the robot system.

talked about the you mentioned slavery there and the word the term robot was actually first used it's a Czech word from the word robotter which actually means forced labor so in a way like slavery kind of mentioned there and that comes from Carol Cap um play it's a play by Carol Cap and in the play this was in like 1920 basically a play about robots that are these synthetic biological beings built to serve humans but eventually they you know rebel and wipe out like And this kind of like was very popular, very widespread, sparked early debates about the themes of automation, anxiety around automation like the loss of jobs, control humanity.

So the very things that you know are actually happening right now, a hundred years ago they were very much having the conversation and this portrayed the topic of robotics and automation in a really negative light to the general public. Okay. Yeah, that's interesting. I didn't realize that it actually came from that. So yeah, I'm glad I kind of brought that up as a topic. I think um yeah it sounds to me like you because we just jumped from 2 and a half thousand years ago to like the the 1900s right and it seems like there's been a pathway for what you know eventually we'll come to define like what we think robotics is now but this idea of automating labor or automating tasks that we couldn't do before and it for a long time that's a fairly purely beneficial thing from what I can what I can gather and it's kind of this is just taking us on a better path.

We can, you know, have the technological enlightenment. We can do things we couldn't do before, etc. And then you hit kind of the 1900s and on, and all of a sudden it's almost getting too good and people start getting worried about, I guess, you know, them being replaced, how we coexist in society when technology goes so far and starts mimicking kind of, you know, human behaviors that weren't just kind of simple mechanical tasks, but other kind of, you know, we're alluding to AI already as well, right?

um that kind of thing that kind of fear we have around AI. It sounds like there's a similar one here. Yeah, definitely. I think like it also correlates to like the kind of geopolitical shifts like when you have like widespread large cities with, you know, relative poverty and all this kind of stuff and maybe job competition that didn't really exist 4,000 years ago where there was always a place to do something. But like I say, bit of a negative turn, one like the first negative turns on this stuff.

I imagine there was previously about, you know, witchcraft and you know, kind of crazy people talking about, oh, what is this stuff? Burn the witch, the one building like that valve over there. And then one of the most iconic book series ever created that I am an absolute fanboy over. Have you read this? Isaac Asimov's the robotics trilogy, the foundation trilogy. Have you read the whole series? I haven't read the series, Alec. I know Isaac Aszimov. So this is a point where I do I do, you know, there's something I know about this, but yeah, I know I know the name.

I know the the laws of robotics, the three kind of famous ones, you know, shamefully probably more through something like iroot than anything else. I think that's the one where I first came across the term, but obviously like yeah, he he's, you know, one of the big three science fiction writers of all time, has inspired I mean, you could you could argue his work has inspired virtually everything we now see to do with robotics from like the Terminator to, you know, uh even like people like Philip K.

you know, and Electric Dreams, that kind of stuff that we see more recently. Um, I think Azimov is also foundation, right, which I think is has recently been turned into into a popular series. So, yeah, he's his uh his thoughts are like still very very pervasive in culture now. To think that he came up with this like a lot of these concepts in the 1940s is mindblowing. Like the idea of like early kind of robotic governance principles like that hadn't really been thought of before. there was no such thing as robotics, not even robotic arms or anything like this.

And he was thinking about like humanoid robots, but far like beyond that, he was thinking about the kind of like the political cultural shifts that would happen in societies as humanoid robots started to become widespread. And honestly, if you I would recommend absolutely everyone read this this entire book series. It goes from like the iRoot series to the Empire series to the Foundation series. And all of them are just sort of giant continuum that all spawn with the start of like the introduction of robotics and like how they're designed, how they're programmed, how these three very distinct laws which I think are really powerful and should be mentioned.

He basically came up with these three laws of robotics that were like the ideal ways of governing robots. And the first law, the most important and paramount law is a robot may not harm a human. Okay? And then the second law is it must obey unless that conflicts the first law. So it must obey human orders unless that conflicts with harming a human. And then third and final law is it must protect itself unless that conflicts with the first two laws. And like you're kind of saying that sounds obvious, really easy to live by.

But like there's so many things like it goes into like real depth on what is harm to a human? Like should I lie to a human or you know say no to it if it can cause emotional damage to that human? And the whole book series has a really great kind of consideration interplay of what it is to kind of define these robotic these laws of robotics and how there's interplay between them for any given situation. It is really fascinating. Yeah, this is really interesting, right?

Because I think I mean you and I are quite big into the whole free will debate and and morality and how we derive that and I think this is this is somewhere where you know you people can people are worried in some sense that a a robot could take these laws and interpret the state of the universe in such a way that decides maybe kill all humans. Right. It's like if if if um a paperclip book or whatever it was. Yeah. And if if it perceives, for example, that the death of all humanity would actually be the minimization of suffering, right?

That's kind of one way of defining morality, then maybe it could interpret those laws as a reason to kill everyone and go uh Terminator, right? That that that's interesting. I also wanted to ask you as someone, you know, who's read the books then. Um what how these Yeah, we're finding out now, I think. Um, but no, I I think I wanted to ask that you call them laws of robotics. Are are they like a died act on how like what's the law applying to is it a law for that you should abide by when you're building a robot?

Is that the idea or is it something more abstract? So it's like basically they have the when they start to create more humanoid robots that kind of go beyond typical processes, they kind of they're called posatronic brains. Basically the laws are defined such that any manufactured posatronic bane has to abide by these laws for it to be considered safe. There is basically like early robotics as we're kind of seeing now were starting to be widespread and there was a general kind of resistance to it by humanity by earth at the time and these were the kind of requirements that were entrenched into these posatronic brains these robotic brains basically so that they could kind of coexist with humanity in a way or could be developed safely.

But I think the really interesting things is like it's not hard laws in such a way that's like okay well you know what other laws are there obviously other dynamics like these three laws won't answer any question that a robot has and then you know define an action it kind of they're the overarching governance mechanisms mechanisms and it's like every decision made should try and feed into those three laws by order of preference but like we say a really you know you can have a waiting between each of them right like a really solid order that and we're not going to go too much into that I find it fascinating but we're not going to go too much too much into that.

But the main thing I want to say about this is read the series. It's incredibly insightful from something written a hundred years ago and I think we're now coming into the realm where we're seeing humanoid robots being built being deployed and when they're starting to interact in like the general public there will be questions around like what is the governance mechanisms behind this and do we need you know widely adopted standards and you are right like so much of this feeds into like pop culture like you talked about Terminator like obviously Bladeunner was another one like there's so many references to robotics advanced humanoid robotics in modern sci-fi because of Isaac Asimov.

Yeah. Well, this and I was going to say the other thing that really interests me about this is actually the timing of it. And you know, so we're talking like 1940s for for a lot of this. And it sounds like it's at this point in time, like bear in mind, we don't really have computers yet, right? At this point in time, we Alan Cheuring is still in the midst of cracking the Enigma code, right? And we don't have computers yet. People are writing about how to program logic in humanoid robots.

So you know that whole path from automating technologies to help you know simple tasks as we talked about thousands of years ago then by this point we've already jumped to humanoid robots which is clearly way way down the line like we still haven't quite achieved that yet but this was clearly the worry in 1940s while computing was being invented in parallel. was just fascinating. It it I don't know what he was on or how he managed to like I imagine it's like one of like um survivor bias where there was a million people that were wring absolutely nonsensical things and I gasimov just happened to be the one in 1 million that was on target.

But yeah, like I think that's all I want to say about the literature. I think the the the the kind of the tone that I want to set or the image I want to set for this is there's been so much written like from a kind of creative perspective up until the 1940s that has impacted so much of the work after that. so much of the actual tech technology being developed has been influenced by these early creatives. Thinking about the art of possible purely from a creative perspective and it's kind of like a self-fulfilling prophecy where now we obviously humanoid robots are now being developed by multiple companies and maybe a lot of these people have probably been like inspired by the work of Isaac Asimov.

Today's episode is sponsored by the Vchain Foundation. Now, we often talk about how emerging technologies like blockchain have the potential to revolutionize sustainability and Vchain is at the forefront of this green movement with initiatives like Vet Dao. Let's take a closer look. Imagine tracking small actions like recycling at low cost making them valuable assets. These verified records build a massive collective impact benefiting people through their own actions. Sound a bit complex, but Vchain made it fun. Explore our apps at vb better.com. Vchain's goal is to power a new era of sustainable solutions.

So, if you've got an app idea or just want to earn rewards for small sustainable actions every day, go check out vetero.com to find out more. And now back to the show. So, let's move away from the literature, you know, tech podcast. Let's get into the tech. Maybe there's a bit of an inroad before we really do get into tech. We kind of need to set the foundation of like what foundation series references. too many references here of like what was actually happening in workplaces before then.

So in around 1913 we had this magnificent individual Henry Ford who kind of like set the gold standard of the assembly line and though Henry Ford wasn't a robot like the point was that he was moving practices to assembly line style culture which enabled like mass production and like repeatable kind of tasks consistently. that was kind of like the gold standard for producing on like a a hugely productive um scale. And you know these kind of like very routine repetitive jobs were perfect in the early days for early robots like the assembly lines were perfect for early simple robots to operate in a control setting.

It was because of Henry Ford that we had these principle that were w principles that were widely adopted that enabled robots to make such an immediate impact when they did. Yeah. And I think um so obviously this is this is still okay we've gone from theory to practical now and at this point in time like the early 1900s the practical limits were you know mechanizing simple tasks. So we've had the industrial revolution, you know, we've had uh essentially electrified or gas powered systems in place to automate human work already.

And then yeah, Henry Ford of of Ford, just in case anyone is thinking about that name, uh it is of the Ford Motor Company. But yeah, he was he basically decided to he was looking at efficiency, right? And how what's the most efficient way to produce things on mass? He was producing cars on mass. Um yeah. So I my question on that to you would be so would that come under the definition of robotics, right? If if we're just even if we're not we're not we're not talking about these humanoid robots right now.

We're talking about simply putting things that replace human labor in the production line. Does that qualify in kind of the the larger definition? Um so like a robotic a robot is typically or robotic is typically defined as like a machine capable of sensing, processing and acting. So it's like quite a simplistic view of the world. I guess the early robots with in like say the assembly lines like the very early robots were typically just capable of actuating and then eventually like it became time based actuating and then it eventually became like sensor based actuating.

So I think you probably would describe these more as like machines like automatic machines or something like this but again like these definitions are so like it so gray and vague like it doesn't really it's not really that useful to define things in this sense. But you are right like it perfectly set this the kind of the the scene for these early robots. And there was some early like academic work around the kind of before the like actual mechanical systems were starting to be developed around like logic gates and the idea of how to kind of use logic gates to simulate feedback loops so that you know if input I can do output.

And these kind of really laid the groundwork early on in like uh the cybernetics paper around the idea of control and how you'd start to control these machines when they're in the open world. But you are right, the biggest advent was in 1954 which was the first robotic arm patented by a guy called George Devil in America. And it kind of devil spelled de V o l just for clarity. Right. Exactly. It kind of like it was really simplistic. You can see images of it online and it just looks like kind of one to two axis robotic arm, super old school.

It used um magnetic drum to store task instructions and it kind of was just a repetitive machine, not much intelligence, not much input and it could just do things. But already people started to see there was a huge impact from that. And they started a robotics company two years later. It was called the Unimation Foundation or something like that. And there was this other person called Joseph Elgabberger. I assume that's how you say it. He was describ Yeah. Engelberger. Let's go with Engelberger. Who was described as the father of robotics and they commercialized that early robotic arm into the Unimate robot.

And this started to be widely adopted on those kind of factory lines we were talking about initially at General Motors. And you, like I say, all the benefits that you were reaping from, you know, those factory lines, those automated factory lines that Ford had originally kind of conceived of and pushed were now having robots on them to really take that to the next level. Yeah. And I think um this is it's interesting that, you know, this is still a pretty recent development, but it's now the standard, right?

Like if you go into I mean, I went in on a school trip to the Jaguar Land Rover factory in every every young kid in the UK did. Exactly. And that was actually that's actually probably the only time in my life I've been exposed to robotics because um at that point, you know, you walk around the factory, everything is being done pretty much by these big robotic arms, you know, specialized for for jobs and and and you know, some of them are big, some of them are small because they're doing different things.

Um but we actually got, you know, we spent time sitting down trying to program a very small version of one of these. That's part of one of the exercises. And it's fascinating cuz yeah, it's like they it's something that is quite simplistic in its original form like in the kind of 1950s, but now they're incredibly um articulate, right? They can they can be articulated in many different ways. They can do lots of different things and they're now a bedrock in manu in the manufacturing industry. I I'm pretty sure I read as well that Amazon was produced they have like a Amazon healthcare line and they were they were moving pill packing in to a robotic arms thing right so they basically were having pills put in bottles by robotic arms right because it was just infinitely faster than how they were doing it before which is you know it's kind of mind-blowing but also makes complete sense given what you just said.

Yeah, I mean even in the early days like they were seeing a lot of those benefits like the one they were typically used in quite dangerous settings where there was like hot metals or dyes and things like we didn't want a human in the loop. So it was typically to do you know repetitive simple tasks but dangerous tasks with precision and obviously like the the upside was you know worker safety. No one was involved in that. They could do 24/7 typically. But they were extremely simple machines like they were all electronically controlled so there wasn't any programming where you could change things around.

There was no perception like I said before, no autonomy, no sensors. So like it's really important to highlight that these didn't have any environmental awareness in like if we were describing you're like a blind person who's deaf and can't you know you're going top left to bottom right constantly. That's all you're doing and that's all you know. You don't really understand anything around it. And they were powered mostly by like hydraulic actuators and all this kind of stuff. But if you move like basically 20 to 30 years into the future because I see Jack's itching to push this forward.

Um, you know, we get to 1985 and these kind of principles are just being pushed further and further and like the actuators are becoming more efficient, they're becoming more powerful, they're becoming bigger. We're doing these on a larger scale. Things are starting to be really productionized. Um, and you know, by 1985, you have 20,000 industrial robots in use in Japan alone, 4,000 in the US. And Japan really took hold of this like kind of manufacturing market in like the 60s7s 80s and became like basically the holy grail for robotic development and especially on like manufacturing lines.

Interesting. I I wonder I don't know if you know but like this sounds like maybe it was tied with the the big Japanese economic bubble at the time, right? Because it seems like they were way ahead of things at that point. Is that do you know anything about that? Is that related? Well, Japan really pushed it because they had massive labor shortages. So like they needed to kind of keep up in a manufacturing sense but had huge labor shortages. There was like they wanted to reduce reliance on foreign technology as well and foreign immigrants and all this kind of stuff.

So they really tried to crack down the idea of like again expanding what the current workforce could do with automation and like it wasn't just the kind of the manufacturing the electronics they also well across the globe that like a lot of other things are being pushed. the idea of like vision sensing like actually being able to start to like say take an input where I won't try and grab a vial here until I know there's a vial and then it becomes a question is like how can I know there's a vial is it like infrared I can have like a vision system vision systems were quite like limited early on so it would probably typically be like infrared where it be like a a line from one point to another and if that line breaks I assume a vials crossed it but we were starting to see the idea of like moving from a machine more to a robot because you had this ability ability to sense and make a decision and do something

based on that without a human in the loop. Got it. So, so just to kind of recap, we've gone from, you know, ignoring the kind of the very earliest theory, like the theoretical conceptual stuff in the practical terms, we've gone from the early 1900s, Henry Ford, and we're talking more about machines at that point that can actuate, that can do something. Then as we move forward, we're then talking about um being able to do more complex things, but also being built, you know, pretty specific single purpose.

And then now as we reach the kind of 1980s, 1990s and onwards, we're getting to not only is there an output from these machines, they can also take much more complex inputs. They can sense. They have a they kind of they're simulating perception by this point, right? Not just action, I guess, is is in human. I would Yeah, that's exactly it. I wouldn't say simulating perception. They are perceiving. They are like getting stimulus from the real world and making a decision on it. Like a very simple example is like you like you say you have a healthcare robot that is moving vials on a conveyor belt into certain packages.

As the vial comes in, it sees if it's green. If it's green, it puts it into box green. If it's red, it puts it into box red. That to a degree is robotic. But all of a sudden, we're starting to say that this robot isn't just doing the same thing every single time. It has a degree of external stimulus, a degree of understanding of the environment, and it makes a decision on what action it should do based on that. It's still super limited, right? It's still like basically three three data points.

If red, go here. If green, go here. So, it's not very complicated. So, the the kind of the difference between it being a machine and a robot still is quite like it's quite limited, right? Yeah. Yeah. And I guess at this at this point we're still talking about like predominantly nonhumanoid robots, right? So things that don't look like humans um and they're not exactly general purpose like I guess a human brain is that can be applied to virtually anything, right? Yeah. Like the I think so when we get into the 1990s actually then it it kind of we robotic arms are definitely the golden standard in manufacturing lines.

are starting to see whole manufacturing lines basically dominated by giant robot arms. They can obviously work 24/7, handle heavy weights, no danger to humans, all this kind of stuff. But in the 1990s, we're really starting to see the diversification of robots and the increase of intelligence. Like you're starting to see things like autonomous vehicles, so early kind of locomotion vehicles like little rovers and things like this where they had sensors that can move around maps based on that. Um, you start to have the Da Vinci robots, which is super interesting.

I know a lot of people probably heard of this, which like early surgical robots. The idea of like say basically reducing tremors in an operator's kind of hand, so they can like, you know, you have like a joystick style thing. It wasn't like that early on, but it could reduce the amount of tremor and movement in the surgeon's hand. Um, so that kind of feedback was pretty epic. And you started to see in the late 1990s um early kind of bipedal walking and stair climbing human robots.

It was like really in the late 1990s that we started to see the idea of like robotics starting to come into the humanoid sense and bipedal locomotion is like is not a simple thing. So having like robots that can do it in 19 even now you look at like in 2024 like they don't really look entirely comfortable. I know we just saw like that humano robot do a marathon, but it didn't look like particularly stable or like efficient doing that. Yeah. And I actually just as you were speaking then I I quickly Googled I was like well I I think the example of robotics that most people will have seen in the real world or you know know of in the real world would be the Boston Dynamics ones, right?

Like the the little walking dogs, right? That's I guess that's quadripedal. But yeah, that was founded in 1992 which is really surprising to me. I presume that had been a much more recently founded company. So yeah, I guess these things were already starting by then. And again, like it's not too complicated to have like a couple of motors that look like legs. Like they're not particularly functional, but they kind of look like legs. They don't work very effectively, but like it's fairly not simple, but you can do that.

And you can imagine after like 30, 40 years of development of motors and basic robotics and basic control systems, that actually wasn't too big a leap. It's kind of like the widespread adoption we're seeing now and being able to handle things in like I guess non-controlled environments. That's the really impressive things having like dynamic control systems that can actually allow these to operate in you know the open world. But again, let's fast forward. So kind of we got to 2000 and like I said 2000 plus we're really starting to expand from static robots in place doing the same thing over and over again.

They have their little context box and they know exactly what's in it and they just repeat tasks over and over again to the idea of like moving around locomotion systems. So like robots moving in real world environments, you know, dynamic environments with obstacles moving, people moving, terrain changes. And this is kind of like where it gets really tricky because now all of a sudden it's not just about the robotic system and you know you have a defined environment. You start to go from point A to point B.

It does that a million times. it is. I don't know what the environment is going to look like, but you've got to work out what it is and do something based on that. And this is like where it gets really quite interesting. Yeah, this is uh I this is just the part that kind of blows my mind, right? Because I I imagine the technical challenge it seems to the layman to be huge, right? And I guess uh you know other examples that people would be aware of, right?

So the Mars rovers, that's a robot, right? Because it's it has the locomotion element that you're talking about. It has to move, has to be, you know, I don't know. I I guess I don't know how much of the Mars rovers was being effectively like remote control versus autonomously moving and autonomously doing stuff. I guess there's been a lots of different ones on the way. But then if you think about autonomous movement, then the Roombas would be a really good example, right? Because the little the little discs that run around your your house cleaning and vacuuming, they are obviously making decisions because they're sensing the environment.

They're kind of pinging between when it gets near a wall, it has to turn around. I I don't I don't know if it's doing anything more impressive under the hood. Like is it do they try and like optimized to make sure they clean the whole floor? Like is that a thing? They do now. So they're now like obviously AI everything's AI enhanced, but Rumble was one of the early AI enhanced machines. Like I think I'm glad you brought up the Mars rover. That's something that is probably most people are aware of and there's been multiple versions of that, right?

And the reason there is kind of obvious like it's so far away that to remote control that the latency between you know what I'm seeing on the screen and what it's actually doing is 22 minutes or 10 minutes at closest approach and things like this. So there's just there's no way for a human to be in the loop in that. So you kind of have to have a self-contained system where you say you just go explore. Don't go into a hole. Try and explore as much as possible.

Don't do the same paths over and over again. Go tell me what you see and send the footage back. And obviously one of the first really kind of mobile commercialized systems by far were the Rumbers. Like I freaking love them. I remember seeing some of the early Rumbers. We had like a little project in my old place where we would um basically program a Rumba and kind of like optimize the path planning for the Rumba because it's such like an intuitive way to think about things.

And now obviously they have like little sensors where they can know you know how far is the distance there and they understand they get a rough understanding of what their map is, what their room is and they can optimize cleaning patterns where it's like you know okay you know I had someone walk through part three of my map the other day. I should probably focus on that first and then you know I did phase two really recently so I'll just do phase one next. So they kind of it's not super complex but like they are AI enhanced now whatever that means where it's like they'll optimize the coverage to ensure they're cleaning as effectively and efficiently as possible.

Yeah, that's really interesting. And I guess like the only thing I'd add here is that I think the Roomba is quite a nice showcase of when you have you know of of where we are in terms of um the the space of problems you can solve with a robot and how advanced they are. So rumbers have been around for like what 20 years plus now. Um, and as you said, it's it's quite a simple, very well- definfined problem with clear constraints, right? You're cleaning a room, right?

The room's not changing. Uh, maybe the cat's running around. That's that's thing to deal with. But like the task is fairly simple. And again, robotics as a field has managed to pretty much solve that problem very very well for a long time. I guess it's when you get to more complex problems uh more more you know when you when you don't have such a narrow field that you're applying it to um that's when things get a lot more complicated right I think another point is like if something goes wrong and the rumor bumps into you know a kid it's fine like no one cares so you're very content to push these out and mass manufacturers as long as they're not going too fast exactly if like an autonomous vehicle like a Tesla goes wrong it kills someone so it's like very different degrees of risk there as well so like I say this is one area for locomotion which is quite important like will track robotics kind of like basically taking cars which we think

of as machines anyway giving them a degree of intelligence and like letting them do things. The next one, the one I want to speak about then is kind of like you kind of have this contrast and this kind of like again the idea of animal inspired robotics. Like we look at evolution, we see how efficient evolution is and how efficient so many animals are and there's so much inspiration to be taken there when we design things and we often do that. One of the next big areas of locomotion is the idea of quadripets.

The idea of like Boston robotics, you know, big dog. I think everyone's seen that like it looks kind of like a robot kind of like a robot dog without a head and early it was kind of designed for like military you know terrain navigation like reactive gate all this kind of stuff but more and more we're seeing like spot which is made in 2015 um being used in kind again like dangerous terrain like things like construction settings mining police work disaster inspection is quite a big one and again like it's just about information gathering you kind of want these systems to be how to go investigate an unknown environment and see what's going on to give information to the humans to then do something with with it.

But like you know it's quite tricky to go up a staircase as you know a robot control system or work in hazardous environments where there's fires you know you can't really see maybe there's smoke and things like that. So you have to have a lot of different sensors kind of inputting to make sure that they can do you know the best with whatever environment they're in. Yeah. It's interesting that you mentioned that it's um it was potentially like military applications originally. I hadn't I hadn't actually crossed my mind, but when you think about I mean that's obviously also been the inspiration for like a very very well-known Black Mirror episode, right, for these robot dogs and what can go wrong when you arm them and do other things.

Um and I that kind of reminded me that a really obvious category of robotics that is, you know, a huge topic at the minute is drones, right? So, not just the drones for taking a nice picture when you go on holiday, but military drones like warfare has completely changed now uh by thanks thanks to robotics, right? And you know, for better or worse, it's very hard to tell, but it's like you were saying, you're sending in robots now to go and do the dirty work of things that previously you'd have to you'd have to use human labor for.

Yeah, drones is a big one. Like a lot of my work was previous on drones. Like even the DJIs that you see now that are widely adopted like you think that you're controlling that drone 100% but there is like they've got their own control system, their own sensors that are making your job as easy as possible because there's wind stability, they're very difficult to control. So even when you think you're going forward and just pressing forward, like there's a hundred different inputs going into that drone to make sure it does go forward and doesn't go left when the wind kicks in.

And I think a lot of people don't realize this stuff is so many so much of robotics and control systems that are kind of directly kind of have existed because of the robotic field are in every you know kind of every different aspect of life we don't realize. It just makes our life slightly simpler. Like drone control for example. Yeah. That that scares me a little bit. Right. You're telling me now that my uh my DJI is it makes me feel like I'm in control but actually it's in control.

That's a bit and one day it's going to turn around and get you. That's so drones obviously a big area for locomotion bipedal we've talked about it there's like I think that the kind of early it was Boston Dynamics for those crazy videos about like the Boston Dynamics Atlas doing all the parkour the flips all the crazy stuff but the idea of like say Tesla Optimus coming out and that actually being in people's homes and doing basic chores. Figure AI is another one. We're actually starting to seeing like real competition heating up in the humanoid robot space that's being commercialized.

It's not just in a lab. Like I say in a lab is fine. It's still very freaking hard. I don't want to underell the work that's gone in there, but it's fine. It's controlled. Things go wrong and you don't see it. But when you feel comfortable enough to commercialize and put into someone's home, that is a very different that means, you know, there is less less faults than one in a million, which is crazy to think of that these things are now capable to work on these very dynamic environments and the company feels comfortable that nothing's going to go wrong.

Yeah, it's interesting you mentioned versus in the lab. I saw a video going around. I'm not sure how speurious it is or or whatever, but it was of one of these humanoid robots that was kind of in a lab. They were if you saw this video and they were testing it and it there was some malfunction and it started spinning its arms in like an attack motion almost. I mean, I think it's it's people people tried to spin it like the the robot's become self-aware and it's trying to kill you, but actually it's probably just a mechanical malfunction.

But yeah, that that's the kind of thing we don't see often and it was leaked from one of these labs. But yeah, I mean, were you telling me about, you know, in terms of how far this stuff has gone, was there not a a Tesla event last year that was fully robotic? And then there I think there was some there might have been some uh hoodwinking going on in reality, right? I think it was like people were saying there might have been humans controlling it, but the point being we're at that stage now where companies are trying to showcase that they've got humanoid robots that can do stuff at least.

Yeah, 100%. And I think like I say there are many different types of robots being pushed out there. Autonomous drones. We talk about humanoid all these kind of bip bipedal quadripedal robots. Um and maybe even things that people don't often consider robots like exoskeletons is another big thing both for like and military applications as there always is but al also for healthcare and you can imagine that being used in construction. Okay Jack we're about 40 minutes in so I think we should maybe cut this down and pause this first parter.

This is obviously, you know, we've putting this off for a while, so I think this might be our three-part series and maybe we can break it down here. Um, so in this episode so far, we've covered how robotics has evolved from ancient myths and mechanical inventions to sco influences like Asimov. And then we even went into the realworld rise of factory arms and the early push into more dynamic environments. So I think we cut this first part of this three-part series and in the next part we'll talk about you know what actually makes a robot what are some of the core components like sensors the brains motors and software and then eventually we'll explore how these robots are being used today and maybe in the third part hopefully we'll touch on what's going to happen in the future and give some predictions on this but yeah for those still listening thank you for listening wherever you may be and join us next time in part two of this robotics series to untangle a little more

of web Okay.

Topics: AI & Robotics · Blockchain & Crypto

Related episodes

All episodes