534: I Think We'll Add a Head
Transcript from 534: I Think We'll Add a Head with Aaed Musa, Christopher White, and Elecia White.
EW (00:00:06):
Welcome to Embedded. I am Elecia White, alongside Christopher White. Our guest this week is Spider-Man and his ropebot dog.
AM (00:00:18):
Hey guys, how is it going?
EW (00:00:20):
I mean, Aaed Musa is here to talk about his YouTube video channel.
CW (00:00:24):
Hi, Aaed. Welcome.
AM (00:00:27):
Hey. It is great to talk with you guys.
EW (00:00:29):
Could you tell us about yourself, as if we met at lunch at Open Sauce?
AM (00:00:35):
Yeah. Sure. Yeah, my name is Aaed Musa. Most people know me from my YouTube channel. I like to make robots and a bunch of different gadgets.
(00:00:45):
I have always been interested in engineering. As a kid, I was really into 3D printing. In high school I started a YouTube channel. I would just post a bunch of the stuff that I was building. Started to get a bit of a following, and just kept on making projects and posting them.
(00:01:02):
I am a recent mechanical engineering grad from Purdue University. Now I am co-founding a startup where we are making robot dogs.
CW (00:01:12):
Awesome. And I am disturbed to learn there are people who are adults now, who enjoyed 3D printing as kids.
EW (00:01:18):
<laugh>
AM (00:01:21):
Yeah. I think there are a lot of us out there.
CW (00:01:23):
Yes.
EW (00:01:25):
We are going to do lightning round, where we ask you fast questions and we want short answers. If we are behaving ourselves, we will not say, "Are you sure you do not want to consider that?"
CW (00:01:35):
Or, "Tell us more."
EW (00:01:36):
Or, "Tell us more." Are you ready?
AM (00:01:38):
Yeah. Let us do it.
CW (00:01:39):
Peter Parker or Tony Stark?
AM (00:01:42):
Peter Parker.
EW (00:01:43):
Peter Parker or Miles Morales?
AM (00:01:46):
Peter Parker.
CW (00:01:48):
Miles Morales or Hiro Hamada?
AM (00:01:51):
I have no idea who that is, that second person, so I will go Miles.
CW (00:01:53):
Oh!
EW (00:01:54):
Big Hero 6.
AM (00:01:57):
Oh! Oh. Hiro. Okay. I will still go Miles.
EW (00:01:59):
Hiro made some great robots.
AM (00:02:02):
Yeah, Hiro was pretty cool, but I think I like Miles more.
EW (00:02:05):
Rey, Finn or Poe?
AM (00:02:08):
Rey, Finn or Poe? Finn.
CW (00:02:14):
BB-8, R2-D2 or C-3PO?
AM (00:02:16):
R2-D2. He is too iconic.
EW (00:02:19):
J.A.R.V.I.S. or R2-D2?
AM (00:02:21):
J.A.R.V.I.S. J.A.R.V.I.S. is really cool.
CW (00:02:25):
What was the original planned name of Rom the Space Knight?
EW (00:02:30):
That is just pure trivia.
AM (00:02:31):
I have no idea. What is this from?
CW (00:02:34):
Marvel's "Rom the Space Knight," an ancient comic. It was their first robot comic. The original name was COBOL.
EW (00:02:42):
I should have limited it to Marvel movies.
CW (00:02:42):
They wanted to call it COBOL.
EW (00:02:45):
Not Marvel all trivia.
CW (00:02:46):
Fine. Fine.
AM (00:02:48):
Yeah, I am really an MCU fan. I do not know much about the comics. I just watched the movies and the TV shows.
CW (00:02:53):
As they said, I am very old.
EW (00:02:58):
Stepper motor, servo, solenoid or hydraulics?
AM (00:03:02):
Stepper.
CW (00:03:03):
Complete one project, or start a dozen?
AM (00:03:06):
Complete one, for sure.
EW (00:03:08):
Do you have a tip everyone should know?
AM (00:03:13):
Always seek knowledge. Always learn. Every day is an opportunity to learn something new.
CW (00:03:19):
What is your favorite fictional robot?
AM (00:03:22):
TARS from "Interstellar."
CW (00:03:24):
Oh yeah. Those were cool. Well, he was cool. It was cool.
EW (00:03:29):
Okay. So not in lightning round, what is your favorite non-fictional robot?
AM (00:03:35):
Non-fictional robot? I would have to go...
CW (00:03:39):
You can choose your own. It is okay.
AM (00:03:40):
Spot by Boston Dynamics. Yeah, it would be easy to choose my own. But for me, Spot was my gateway drug into robotics. Seeing Spot in high school really got me excited about robotics, particularly quadrupedal robots. It is just such a cool robot. It is so cool.
EW (00:04:02):
Okay. So you got into robotics in high school, and you built your own robot dog.
AM (00:04:08):
Yeah. Yeah, built several.
EW (00:04:10):
Tell us about the first one.
AM (00:04:14):
The first one was pretty bad, looking back on it, but it was made completely from servos. It was called "Ares," like the god of war. Yeah, it was just made from servos. It was something that I whipped together in my parents' garage.
(00:04:32):
It was able to walk, just not very well. It was fully 3D printed. I just made it from anything I could just find on Amazon and I could find around the house. I think if I did not build Ares, I probably would not be building these cooler robots today.
EW (00:04:54):
If you compare Ares to CARA 2.0, what is the biggest improvement?
AM (00:05:02):
Pretty much everything.
EW (00:05:02):
You cannot choose everything. <laugh>
AM (00:05:09):
<laugh> Yeah, that would be my answer, "Everything." But more specifically, the actuators. The actuators are the heart of pretty much any robot. If you have bad actuators, then you just have a bad robot.
(00:05:22):
With Ares, I was using servo motors, which are cool for certain applications, but they have super high reductions and are not very compliant, not very back drivable.
(00:05:31):
With CARA and pretty much all the other quadrupedal robots I have built since, they have used brushless motors with very low gear reduction gearboxes, which allow the actuators to have high compliance, high back drivability, high torque transparency. It feels more like a actual biological limb, than a servo motor does.
CW (00:05:58):
We are firmware engineers. What does "compliance" mean?
AM (00:06:02):
Compliance is basically-
CW (00:06:03):
You know about motors. I do not know about motors. She is looking at me like, "You are dumb."
EW (00:06:06):
I am using my hands to show him. It is not doing great.
AM (00:06:09):
Yeah. So it is basically, I would say, the springiness, the softness of an actuator. Your limbs are very compliant. If you are standing and someone pushes you, you are able to soften your pose to be able to rebalance yourself. But you can also stiffen your joints if you are, I do not know, punching something or lifting something heavy.
(00:06:35):
Compliance is that ability to be able to soften the actuator's joint. It is very important in robotics, because you want your actuators to be able to soften and to be able to stiffen, depending on the application.
CW (00:06:53):
Okay.
EW (00:06:56):
I am glad you answered that, because I actually did have it wrong. I thought it was mostly about- I had a really, really bad arm robot, which I loved. I was trying to use software to deal with some of its difficulties.
CW (00:07:10):
And it was servo driven.
EW (00:07:11):
And it was servo driven. But if you went home and you sent it to a place, okay, fine. You went home, you sent it to the same place again, you did not end up- You were a couple of centimeters out.
(00:07:25):
You were not going to get to the same place each time, which is why I was trying to figure out the software with the camera and helping it and stuff. There was just a lot of play in the whole thing.
AM (00:07:37):
Yeah. That is backlash. If the joints have slop, if you command it to the same position 20 different times, and it ends up in 20 different places, then probably have a backlash issue. That is just something that is inherent with gears. It can be removed a lot, but if you have gears, you technically will always have a small amount of backlash. But some gears are more sloppy than others.
EW (00:08:03):
But you have been really focusing on capstan system gear thingies, that decrease the backlash quite a lot.
AM (00:08:12):
Yeah. Capstan drives basically have zero backlash. They are rope driven speed reducers. They are basically gear alternatives. They sort of work the same way gear does, except instead of teeth, they use rope.
(00:08:28):
Imagine you just take away the teeth from two gears. The way they interlock is they have rope wrapped around both. As one is turning, it is essentially using tension to pull the other in the other direction. They operate the same way as gears. It is just that they do not have teeth. They have basically zero backlash, and they are super back drivable, super compliant.
EW (00:08:54):
But they have only recently become more available, in part because of the rope technology improvements?
AM (00:09:04):
Yeah. The key thing with capstan drives is, if you want to be able to have that zero backlash, your rope cannot stretch.
CW (00:09:13):
Oh, right. Okay.
AM (00:09:14):
Rope loves to stretch when it is put under tension. Pretty much all ropes will stretch, except for this one rope called Dyneema DM20, which basically has a 0% creep rate, which is amazing. That is basically the only kind of rope that you can buy commercially that will not stretch.
(00:09:38):
It is a newer rope. I forget exactly which year, but I think it is younger than 20 years old. I forget all the details of how it was made, but yeah, it is a relatively new rope. It is really cool, because it is the only way you can make a longstanding capstan drive.
CW (00:10:02):
What about chain? Well, I guess chains have the same problem. They are not under- Chains are metal cables.
EW (00:10:10):
Like tracks? Or like bike chains?
CW (00:10:12):
I guess. Yeah. Bike chains.
AM (00:10:14):
Yeah. Steel cable does not stretch. So I guess I should qualify my statement and say it is the only synthetic rope that you can buy that will not stretch. So yeah, steel cable will not stretch. The thing with steel cable though, and I have tried this, is it snaps if it is wrapped around a small drum and a big drum.
(00:10:39):
There is something called a "D to d ratio." With steel cable, it requires a certain D to d ratio, meaning that you cannot wrap it around a really small drum and then wrap it over a really big drum. That change in drum diameter causes it to snap over time.
(00:10:57):
So if you are trying to make a gear reduction that is like eight to one, meaning that your smaller drum is eight times smaller than your bigger drum, it just will not work.
EW (00:11:08):
So everything got better, as you made more robots.
AM (00:11:12):
Yeah.
EW (00:11:12):
Did the software improve a lot?
AM (00:11:17):
Yeah. A bit. I am not a programmer, so all of my robots thus far, up until now that I am doing the startup, have basically just been hard coded step sequences, trotting sequences where the diagonal legs move in pairs. I have improved that.
(00:11:35):
But now that I am doing the startup, I am taking bigger leaps, because now I actually have someone who does software. Now we are using reinforcement learning. But with my own personal robots, the software has improved a bit. But now I am taking an even bigger leap by having AI do the walking, and not manually programming gait sequences.
EW (00:11:59):
But with dogs- AIBO was long before you were born, which kills me. But trotting is not the only thing dogs do.
AM (00:12:11):
Yes. Yeah.
EW (00:12:12):
I have seen you make it wave, which does not seem that truly difficult, other than the balance problem. But you are not having it do play bows and asking for pets, and rolling over on its back to get tummy rubs. You do not have a dog agent in your robot.
AM (00:12:37):
Yeah, that is correct. Yeah. All of my personal quadrupedal robots have not had any sort of intelligence. It is all hard-coded commands.
(00:12:51):
But yeah, what we are trying to do with the startup right now is the exact opposite, is we want it to feel like it is alive. We want it to be able to emote. It is funny you mentioned a bow and a wave. We actually just last week got our robot to do a bow and a wave.
(00:13:07):
That was all learned through a reinforcement learning policy. Yeah, that is where the technology is going today. Five years ago that was not possible, but because of AI, now that stuff is.
(00:13:21):
The goal is to make a robot dog that feels like a real robot that can go and fetch something if you tell it to, that can roll over and do cool tricks, that can hop onto the couch with you and be expressive.
(00:13:38):
Maybe your robot walks different than another person's robot, and it can show expressiveness through even just its gait. It does not walk the same. So the world is its oyster. It can essentially just learn to do practically anything that an actual dog can do, and more.
EW (00:14:04):
We took our dog to the beach. There are dogs that prance, and dogs that stomp. Dogs walk differently, and it is not just about how tall they are.
AM (00:14:13):
Yeah.
EW (00:14:14):
But-
CW (00:14:15):
Well, wait a second.
EW (00:14:16):
I have a bunch of questions about the reinforcement learning, but I started to ask about the mechanical stuff.
CW (00:14:22):
Okay. But-
EW (00:14:24):
But before either one of those.
CW (00:14:25):
You were talking about the software and how you did not think it was great or whatever, because you do fixed sequence. But you do have a control system in there. I have seen in your videos that you can push the dogs and they right themselves. There are stability feedback control systems in there.
(00:14:41):
I think you are selling the software short a little bit, because it is doing some very complicated things just to be stable, right?
AM (00:14:50):
Yeah. Yeah, yeah. So there is a bit of stability embedded into the software. It has an IMU, an inertial measurement unit. It can sense its own roll, pitch and yaw, and how fast it is spinning. It can correct for that if you give it a push. So yeah, that is another improvement that was made over the course of my robot making years.
(00:15:16):
Something else I was able to program, I call it "pose control," where the robot is able to tilt in roll, pitch and yaw, and demonstrate its kinematics. That was another thing that has improved over the years.
EW (00:15:33):
Yes. And if you ever put a head on your dog, you might want it to tilt its head.
AM (00:15:39):
Yeah. Yeah, exactly.
EW (00:15:41):
Okay. So the reinforcement learning, what are you reinforcing against? Are you showing it videos of dogs?
AM (00:15:50):
You can. But for just getting it to walk, we are essentially giving it a bunch of targets to hit. In simulation, it is running many, many, many simulations, hundreds of thousands of simulations, and it is trying to hit those targets.
(00:16:08):
So for example, we have a bunch of- They are called "rewards." Maybe one reward is, "I want you to walk at two meters a second. I want you to keep that velocity, keep that pace." And if the robot in simulation is walking at 1.5 meters per second, well there is clearly a gap there and it is not doing something right.
(00:16:33):
So then it will do another simulation and try to get closer to that. Maybe in that next simulation, it walks at 2.5 meters per second. "Oh, too high." Okay. It is going to adjust its gait to try to achieve that.
(00:16:43):
So we have a bunch of these rewards, of things that the robot is trying to achieve. Over the course of every simulation, is essentially trying to hit all of those rewards. When it hits all of those rewards, then we have a functioning policy.
(00:17:03):
The whole work with reinforcement learning is determining what those rewards should be, and what the gains of those rewards should be. So, how much should it focus on hitting that two meters a second gait, versus another reward? Maybe we want it to, I do not know-
EW (00:17:21):
To be able to stop in a half a second.
AM (00:17:24):
Yeah, or something like that.
EW (00:17:25):
Because sometimes your rewards are like that, where you have one where you want it to go fast, and one where you want it to stop, and you really have to do both.
AM (00:17:32):
Yeah. We have a bunch of rewards. There are around 20 for what we are doing right now. Determining what those rewards should be, and the gains for each of those rewards, is the whole work. That is what a lot of reinforcement learning research is going into. It is like, "How do we determine how a robot should walk? What are the parameters? What are the rewards that we need?"
EW (00:17:59):
This is on top of your control system. You are not just giving it raw ADC values, and raw DAC values, or drive values. You are giving it, "Here is how to do a step"? Or are you- Where is your control system, versus your reinforcement learning? I know it can be anywhere in there, but where is yours?
AM (00:18:23):
If you are talking about the classical controls, like a PID controller, that all just lives on board on the actuator. The reinforcement learning policy is determining how the robot should move, and then we just send those position commands to the actuators.
(00:18:43):
If we have tuned the actuators correctly, then we send the actuator to move to a certain position and then it is able to move to that position. We do not really even touch the low level control on the actuators, after we have tuned them.
EW (00:18:58):
Okay. Cool. I have seen people try to do reinforcement learning from the lower levels. I never quite understand why, because there is a point at which you just want it to move to where it is supposed to go. And yet I had that terrible arm, where-
CW (00:19:18):
Yes, but the goal with the arm was to do stupid things. To be clear, it was not a badly built arm. It was because of incompetence.
EW (00:19:24):
It was a very cheap arm.
CW (00:19:26):
It was, you bought the cheapest arm you could find, and hoped to fix it using tons of software.
EW (00:19:32):
Yes.
CW (00:19:34):
And instead you just made comedy, which has its own value.
EW (00:19:37):
Yes.
AM (00:19:38):
For sure. Yeah, yeah. Iron Man had the dummy robot, that just served as comedic relief for the "Iron Man" movies.
CW (00:19:47):
Right. Right.
EW (00:19:50):
So, CARA. You make videos and you post them on YouTube, about your projects.
AM (00:19:56):
Yep.
CW (00:19:56):
And you are very good at it.
EW (00:19:57):
And you are very good at it.
AM (00:19:58):
Thank you.
EW (00:20:00):
Your most recent video was about making Spider-Man web shooters.
AM (00:20:04):
Yeah.
CW (00:20:04):
<laugh> That was so good.
EW (00:20:10):
You showed the slip-ups in the design process. It could have been a video of, "Here is how my perfect project worked," but you showed how it was not a perfect project to start with. It got to a goal, but it was not a direct shot.
AM (00:20:28):
100%. Yeah, that is something I try to show in all of my videos, because it is true and it is real. There is no project that you will work on for several months that will not have road bumps, that will not have changes. Sometimes I even just have to change the direction of the project, while I am doing it.
(00:20:51):
So showing all of that I think is A, pretty honest, but I think makes for a better video. I think it would be very boring if I just showed you the cool highlights of everything working, because you just intuitively know that that is not actually what happened.
(00:21:05):
Engineering, yeah, it is hard and difficult. The whole goal of engineer is to fight through those challenges, and problem solve and figure out what is wrong and how to fix it. I think if I can show that, then A, I am showing how I think, but B, I am giving you just a much cooler video, something that is more entertaining. Because, like any movie, any story that you watch, there is always a problem and there is always a resolution.
(00:21:42):
I try to model my videos in the same sense that a story is told, where I have characters or character, just me. A setting, my bedroom. And a problem, I am trying to get this thing to work and it is not working. And then the solution, I eventually figure out some sort of hack to get it to work.
EW (00:22:03):
You do a lot of your engineering in what seems to be your room. How much is that- Is that a studio, or is it really where you live?
AM (00:22:16):
Well, yeah, it is just my bedroom. Right now, I am in a two bed, two bath, so I have a roommate. But pretty much all the shots of my videos are inside my room. I have two printers at my office space, so sometimes I will record there, just showing the printing.
(00:22:36):
But yeah, all the assembly, all of the testing, that is just in my really small San Francisco apartment bedroom. That is really all you need. You do not need a fancy studio to record YouTube videos. You just need a desk and a couple tools.
EW (00:22:55):
A couple tools!
AM (00:22:58):
Yeah, maybe I am selling it short. I have acquired a lot of tools over the years.
CW (00:23:03):
Lighting seems to be the big difficulty to learn.
EW (00:23:06):
We had to record a video recently. Lighting was difficult.
AM (00:23:11):
Yeah. I have two ring lights and I have a pretty nice balcony out here, so natural lighting is good. But yeah, I have a ring light and then I have this really cool overhead desk light, that I think is just a lifesaver for me, because my room actually does not have any lights. So if I did not have it, I would just be in pure darkness at night-time. But it also helps with recording.
EW (00:23:36):
There are a lot of things that go into your videos. I made a list. I am just going to splat out the list and hope you can tell me which one or two are the hardest parts.
(00:23:49):
There is the physics and the planning and the equations. We are all going to bunch that as one. The mechanical design, and then there is the mechanical implementation. The electronics, the software, those are all normal robotics things.
(00:24:03):
But then you have a project write-up, the video editing, marketing the video, and getting the ads. Which of those- Did I leave anything out, I guess is the first question?
AM (00:24:15):
No, that pretty much covers everything.
EW (00:24:18):
Which part is the most fun?
AM (00:24:21):
Oh God, it is hard to pick. It is all fun to me. If I had to pick one, I would say editing probably has a slight edge, just because that is the culmination of the project. I am sort of doing everything at once when I am editing.
(00:24:38):
I am having to remember all the mechanical stuff, and I am having to present everything in a way that is understandable for people. I am having to think about are people going to find this interesting, the marketing aspect.
(00:24:51):
Editing is when everything comes together. I am encapsulating my four months or three months of work, into this one 20-minute video.
EW (00:25:06):
In some of your videos, I have noticed your T-shirt changes. Does that mean you are spending multiple days? Or is that just simulating the-
CW (00:25:14):
The passage of time.
EW (00:25:16):
The passage of time? <laugh>
AM (00:25:20):
<laugh> Yeah, definitely multiple days. I think any project I work on probably takes a good three to four months. I am definitely not just recording everything in one day and just finishing it in a weekend. It takes a while. So yeah, the passage of time is very true.
CW (00:25:41):
So they are more project journals. You are recording as you go instead of, "Okay, I finished this project, now I am going to recap it."
AM (00:25:49):
Yeah. It changes. I try to record as I go, but inevitably I always just hate what I recorded and want to record it again. So yeah, there is a bit of that. But yeah, generally I just try to get the flow as I am going.
EW (00:26:07):
Over the years that you have been doing this has the video editing, the putting it all together, the rush of endorphins of finishing something, been one of the highlights? Or has the highlight of it changed?
AM (00:26:21):
Yeah, for sure. I think the highlight for me is definitely finishing the video, because I am showcasing everything that I have done and it is always the most fun part. Even before posting a video, I am always just super happy to have finished the project and to have showcased it in the video.
(00:26:42):
Even if no one saw it, I am just happy that I have done it and I was able to accomplish the goal that I set out to do months ago. That is always going to be the most fun part for me, is finishing editing a video.
EW (00:26:57):
Which of the parts has gotten easier over the years? Has anything just like it used to be hard, and now it works on its own? Or I went to college and now I understand all the physics, so I do not have to think about it as hard?
AM (00:27:11):
No, the mechanical and the physics stuff, that always has its difficulty. Depends on the project. If I am making another robot dog, it is going to be a lot easier, because I have just done that so many times.
(00:27:24):
But if I am doing a completely new project and I am not very familiar with the engineering or the physics that goes into it, then it is a whole new set of learning that I have to do. Which is fun, because it is always fun to learn new stuff.
(00:27:39):
I would say the part that just inherently gets easier with every video I make, is probably just the editing and the marketing. Figuring out how am I going to- I have developed my own flow with the editing. You will see the similar beats with my videos, and similar corny jokes, and similar editing and similar transitions and whatnot.
(00:28:07):
That stuff gets easier, just because I have done it so much. I know how to flow with my own personality. It is easier to write scripts, because I know how I should go about writing it. I know what parts to expand more on, and what parts to maybe cut out because it might not have any relevancy or interest for the audience and whatnot. The editing for sure gets easier.
CW (00:28:42):
Do you ever go back and look at your earlier videos, to see how much you have improved? Or is that too difficult?
AM (00:28:50):
Not really, to be honest.
EW (00:28:50):
How often do you go back and listen to our old episodes?
CW (00:28:52):
I do not. I do not do that at all. For very important reasons. I like myself and I do not want to feel sad.
AM (00:29:00):
I am the exact same way. It can hurt to watch sometimes, watching an old video. For me, what I really do not like is when someone is watching my video, right in front of me.
CW (00:29:11):
Oh no, no, no, no, no.
AM (00:29:13):
Yeah, that always- The sound of hearing my voice with people around, just- I do not know. It just does not sit well with me. I try to leave the room if someone is watching my video. I have had family members who will put my video on a TV with a bunch of other people watching, and I am just like, "Guys, what are we doing here?" It is not fun.
(00:29:43):
I really do not watch my old videos. It is very painful to watch.
EW (00:29:50):
But you were in high school when you started. You started with a Nerf gauntlet.
AM (00:29:55):
Yeah.
EW (00:29:57):
Did your channel help you get into Purdue?
AM (00:30:02):
No. I did not actually mention my YouTube channel in any of the essays that we had to write for college applications. I just talked about my love for 3D printing, and my love for engineering.
(00:30:17):
But I did not specifically mention my YouTube channel, just because my YouTube channel, then, was nothing. I just posted clips of things that I had built, did not really have much of a following. So it was not really worth highlighting, I felt, at that point.
EW (00:30:34):
Fair enough. When or how did your channel get more traction? Was there a point where you said, "Wow, I am a big deal in the rebellion now"?
CW (00:30:48):
Was there one video everybody was like, "Oh!"
EW (00:30:50):
That went viral.
CW (00:30:50):
That got noticed and then you picked up? Or was it slow growth?
AM (00:30:55):
I think the video that got me the really big boost, was a video where I made a robotic actuator. I made this robotic actuator with this off the shelf drone motor that I got. I 3D printed this planetary gearbox to put with it.
(00:31:16):
Yeah, I made a video on that and it got I think a hundred thousand views in, I do not know, a week or so. That was really huge for me at the time. That was I think the video that put my channel on the map. Ever since then, my channel has just skyrocketed.
EW (00:31:40):
Do you go out and get ads? Or do people just come to you and say, "Please use our 3D printer. Here is a free one. Also, can we give you money if you will mention it?"
AM (00:31:52):
Yeah. That is the cool part, is I really do not have to go out and seek sponsorships. They all just come to my inbox. So I am in a cool position, a very fortunate position, where I can just sift through all of the sponsors and pick one that I like, and it is not the opposite.
CW (00:32:11):
I want to talk about your college experience, because you obviously have a broad skillset. Watching your videos, you are doing all the mechanical stuff, all the electromechanical stuff, a lot of the electronics and software.
(00:32:22):
How much of that was out of coursework at Purdue? How much of that stuff did you get exposure to there, during your engineering degree? And how much of that did you either pick up on your own, or have since learned on your own?
AM (00:32:40):
Yeah. This is an interesting question, because I get asked this sometimes. I would probably say that pretty much all the skills that I am showcasing in my videos, and everything that I use to build my videos, is self-taught.
(00:32:54):
School is good at a very specific thing. It is good at teaching- Engineering school, that is. It is good at teaching you the physics, the reasoning behind a lot of the practical application stuff. Coursework is mostly just solving problems, where you have to figure out the loading on this bridge or whatnot. You are doing a lot of things like that.
(00:33:23):
But in terms of actually learning how to build robots, it is hard to learn that by sitting in a lecture hall and taking notes on dynamics. You just have to go out and learn to build that stuff. Equations can only take you so far. At the end of the day, if you are a mechanical engineer and you are building physical hardware, you have to just learn to build physical hardware.
(00:33:54):
I really do think that you can be a self-taught mechanical engineer. If you are building physical hardware, again, you just need experience. College is needed for engineering, for sure, but it teaches you a specific thing. Practical application is just something you have to learn on the job or doing internships.
EW (00:34:20):
You have, I believe, credited YouTube to some of your knowledge about electronics and 3D printing. Do you have favorite channels?
AM (00:34:30):
Oh yeah, got a ton of favorite channels. God, how much time do you guys have? "3D Printing Nerd" for 3D printing. I have been following him for a while. Posts a lot of great 3D printing content. "CNC Kitchen" also does a lot of 3D printing content. "Made with Layers," again, a lot of 3D printing content.
(00:34:52):
Jeremy Fielding, I think you guys have had him on the podcast.
EW (00:34:57):
Yes. He is great.
AM (00:34:58):
I have learned a lot from him, just about robotics and mechatronics and mechanical design. He is a great example of someone who- He did not go to school for engineering. I think he went to school. I forget exactly what he went to school for, but it was not engineering. He became a self-taught engineer. If you just met him, you would think that he went to school for this, but he did not.
(00:35:22):
James Bruton, another great roboticist. I have taken a lot of inspiration from his robot dogs when making mine.
(00:35:32):
Mark Rober, great inspiration for the storytelling aspect of making YouTube videos. He is a great engineer as well. But I think most people gravitate towards him because he is just so good at crafting stories, and getting you excited about engineering, especially the younger audience.
(00:35:54):
"Stuff Made Here," he makes just these insane robots and these insane gadgets. He is really good at a wide scope of the engineering discipline. He just knows so much about the software, the hardware, the machining. It is very rare to find someone who can do such a broad scope of engineering. He is really cool.
(00:36:24):
But yeah, I will not ramble off too much, but yeah, those are just a couple of the guys that I think are really cool.
EW (00:36:31):
Looking at CARA and one of the recent videos. Actually, I think it was your capstone project for school. You did a video about- Did you turn in that homework twice?
CW (00:36:41):
<laugh>
AM (00:36:45):
Yeah. So CARA 2.0 was my senior design project. After making CARA 1.0, I was like, "Cool. I think there is a lot to improve here." Senior design was coming up and I am like, "You know what? What if I just use senior design as an excuse to keep working on my YouTube project?" And yeah, it got approved and I was able to do that with my friends. So that was pretty fun.
EW (00:37:14):
The video you made for that CARA project, for CARA 2.0, it was a little strange because it had a lot of cat meows. What were you trying to say? It was not just one cat meow at the beginning. There were a lot of cat meows.
AM (00:37:30):
Yeah. Yeah. The reason I believe that I am really interested in robot dogs, is because I myself wanted to have an actual dog, and my parents were like, "No, we are not going to do that."
EW (00:37:45):
<laugh> Nope.
AM (00:37:46):
So I think I just developed this obsession with-
CW (00:37:49):
I will build one myself, then.
AM (00:37:50):
Yeah, exactly. I just wanted to build one myself, because my parents said, "No." They eventually got me cats. And so I am like, "All right, this is not really what I asked for, but okay." But I love my cats. They are cool. They grew on me. Yeah, that is why I had all the cat meows, just because I have cats now. So yeah, that is the reason.
EW (00:38:15):
Would it be easier to build a robot cat or a robot dog?
CW (00:38:20):
Cats sleep more.
AM (00:38:22):
Yeah. In terms of mechanically, I think it would be the exact same. I call it a robot dog, but-
EW (00:38:34):
It is just a robot quadruped.
AM (00:38:34):
Yeah, it is just a quadrupedal robot. It just has four legs.
EW (00:38:39):
It could be a wombat.
AM (00:38:40):
If you wanted to make a robot cat, maybe you would just program it to not be able to jump, and not be able to- Or no, actually you would want it to jump, because cats can jump very high. Just program it to sleep.
CW (00:38:51):
How do you program sullen and haughty and off-putting?
AM (00:38:57):
<laugh> That is I guess where you have to deploy some reinforcement learning policies to give it a bit of emotion. Yeah. I guess you just got to program it to hop off your bed when you bring it to your bed, or just walk away from you when you want to pet it.
CW (00:39:15):
Roll over on its back, accept three or four pets, and then hit you.
EW (00:39:18):
<laugh>
AM (00:39:19):
Yeah, exactly. Yeah, yeah. There you go. Robot cat.
EW (00:39:24):
All right. All right. Okay. So that actually answers that question, kind of. Why do you keep doing your channel? It is really cool. I am glad you do. I do not want to discourage you, but-
CW (00:39:38):
What keeps you going?
EW (00:39:39):
But it is not easy.
CW (00:39:40):
Yeah.
EW (00:39:40):
It would be easier to just do the projects.
AM (00:39:44):
I like the showcasing of it. Whenever I post a video and I see that people are gaining value from it and people are like, "Oh wow, this is really entertaining. This is really cool. I learned something new." That just motivates me to keep doing it.
(00:40:00):
Because yeah, I could just build the project in my bedroom and show it to no one. But I think it is a fun end goal to be like, "Okay, I am not only going to just build this thing, but also create a story around it, create a video around it and show it to other people, because I think people can get value from it."
(00:40:22):
And it is just fun for me. I actually like, again, the editing process. I like crafting it and creating something, a video from the project and not just doing the project. So a little bit of just, it is fun for me, but also I think a lot of people get value from it.
CW (00:40:39):
Does knowing you are going to present a project, in video form or any other form, change how you approach the project itself?
AM (00:40:48):
Yeah. Yeah, a bit. I would say the presentation, definitely. I feel like it has to be more polished, if I am going to show it to people.
(00:40:58):
So even just designing a lot of curves into the part or making it look presentable or adding some color, like a duo tone style or whatnot. All of that, all of the aesthetics, is very much because I know people are going to watch this. I know people are going to judge if it looks really, really crappy.
(00:41:26):
So yeah, I think a lot of the aesthetics is driven by the fact that I know people are going to watch, and I have to make a video about it. But yeah, the engineering part, not so much. It is mostly just the presentation.
CW (00:41:42):
That is industrial design, right? That is a totally different discipline. You could make your robots act the same way and as functional as they are, while at the same time being made out of, I do not know, popsicle sticks or something not aesthetic. Is that something you just absorbed? Or something you consciously went out and spent time learning about?
AM (00:42:07):
I think that is just something that I have just picked up upon, watching other people's stuff and just using my own creativity. I am definitely not an artist. I cannot draw to save my life.
(00:42:20):
But I think I can make something that may otherwise look super mechanical and super complex. I think I have the ability of making it look a little more soft, a little more presentable. Something that people can look at and not think is a super industrial, super high-techie thing with a bunch of wires poking out of everywhere.
(00:42:48):
I think that is just something I have picked up upon while doing a lot of projects.
EW (00:42:53):
It is funny. You say you are not an artist. I understand, because I always say that myself. It makes me think about identities. I definitely am happy to say I am an engineer. But I will not say I am an author, unless somebody really tries to get me to say it. Do you...
CW (00:43:17):
You are an author under duress.
EW (00:43:18):
I am an author under duress.
CW (00:43:19):
Okay.
EW (00:43:24):
You are an engineer. Okay. Definitely, I think you are going to own that one. That one is not going to cause a problem. There is YouTube creator, storyteller and educator. How do those fit with you? You are like, "Yeah, yeah, that is all me." Or are you like, "No, I am just an engineer, who sometimes has stories to tell"?
AM (00:43:46):
Yeah. That is a good question. I do not know. It is hard to put myself in one specific box, or even multiple boxes. But I think in general, I just say I am a creative. I like to make things, and I like to share that knowledge with other people.
(00:44:04):
So yeah, I think YouTube creator, educator, I think that definitely I fall under it. Yeah, I wear multiple hats, and I think you are just going to have to, if you make YouTube videos.
EW (00:44:24):
I get that. I never introduced myself as a podcaster. So. As long as we are talking about identities.
(00:44:30):
Okay. Graduated high school, you went to Purdue, you graduated Purdue last winter semester. Now how many people have asked you, "What do you want to be when you grow up?" Am I going to be the next person to ask you that? Because I do not really want to, but I also want to know. So, what do you want to be when you grow up?
AM (00:44:56):
No, yeah. Yeah, I get asked that a lot. It is a good question.
EW (00:45:00):
No, it is not! <laugh>
AM (00:45:00):
<laugh> Yeah, everyone should be future driven. I understand, at the very least, why people ask it. They are like, "Oh, you are doing these school stuff online. Where do you think it is going to take you? Is this the thing that you want to do? Or is this a pit stop towards something else?" So I understand the question.
(00:45:28):
Yeah. Before graduating, I thought I was going to just go to grad school. So I had my mind set on that. But I realized that I did not really want to go to grad school. I just said I was going to go to grad school, just because, I do not know, that seemed like the next logical step. All right, get a master's now, after you have done your bachelor's.
(00:45:55):
But when I look back at my entire undergrad experience, I realized that everything that I learned was not actually helping me do the thing that I was passionate about. Gave me some of the behind the scenes knowledge, some of the first principles knowledge.
(00:46:16):
But if I am making a robot, I am using the skills that I used to make the previous robot. I was not using the learnings of my dynamics class. Everything that I was using to build robots was everything that I was self-learning, and everything that I was learning from watching YouTube videos, and doing research.
(00:46:43):
So yeah, after thinking more about that, and I was already talking with Aaron, who is my co-founder now, and we were talking about just the robotic space in general. I figured that the best next step for me is to just go out and work on a startup, because why not?
(00:47:03):
I think this is really the only time in my life that I will be able to just work super long hours. And to just completely take that risk of setting out to do a startup, that may or may not work out. So yeah, I figured that that was the best next step for me.
EW (00:47:29):
So CEO and co-founder of the startup, is going in the identity column. CEO? Or are you-
AM (00:47:37):
Yeah, so I am not CEO. My co-founder is CEO. I do not really have a title at this point. I guess I am just the mechatronics-
EW (00:47:45):
Well then your title is CTO. You get a C title if you are the first two. It is just those are the rules.
AM (00:47:53):
Yeah. Yeah, it is still very early on But yeah, I guess I will take the CTO title.
EW (00:47:59):
I think, "Where do you want to go with your career?" is a question that is asked because it seems like you have so many possibilities in front of you. Everybody who asks it is just a little bit jealous.
AM (00:48:15):
Yeah. Maybe. Or if you are my parents, then concerned. <laugh>
CW (00:48:18):
<laugh>
EW (00:48:22):
It sounded like the separation between the theory that you get in classes, and the practical applied knowledge you have gotten through just doing things and learning from other roboticists, it did not get any closer in college. It actually got further apart. Did I understand that part right?
AM (00:48:49):
I would not even say it got farther apart. It was just sort of indifferent. Sure, I picked up one or two things here and there in college, that maybe helped my understanding of physics or of just mechanical design.
(00:49:04):
But if I did not go to college, I think I would still be doing what I am doing right now. I would still be able to make the robots that I am building. So it did not really hinder or enhance my ability to make robots. It was just sort of indifferent.
EW (00:49:26):
Did Purdue have a robotics club you were part of?
AM (00:49:29):
Yeah, Purdue has several robotics clubs. I think robotics clubs are great ways to get practical hands-on knowledge. I was part of the VEX robotics team for a semester. Then I also did Purdue Prosthetics, which is not- I do not know if I would classify it as robotics. Robotics adjacent, maybe. I did that for a semester.
(00:49:51):
Then I felt like I just wanted to do my own thing in my dorm room, and just devote all my time to working on personal projects. So yeah, I quit those two clubs and just did my own thing.
EW (00:50:06):
It is interesting whether you decide to follow your own vision, or whether you decide to have a larger team. Robotics is often considered a very team-oriented thing, because of the specialties of mechanical engineering, electrical and software.
(00:50:20):
And yet, you are doing most of it yourself, or with a small team. Do you think there are benefits with that? Or do you think it is mostly you can only make a robot this big?
AM (00:50:37):
Yeah, there are definitely benefits and drawbacks, for sure. I was on a robotics team in high school. I did FIRST Robotics, FRC. With the team, it is easier to divide labor. You do not have to-
EW (00:50:52):
You specialize.
AM (00:50:55):
Yeah, exactly. You can specialize, and get just really good in one discipline. I was doing all the CAD design for my robotics team. Did not touch any of the programming, did not need to. Someone else could take care of that.
(00:51:11):
Yeah, that is the main benefit of working with the team, is that if you have someone that you trust and that is good at what they do, then they can fully own that stack. You can fully own another stack. You guys can work in parallel and get things done much, much faster, than if you were just doing it by yourself.
(00:51:28):
Working solo, you do not have that. But what you do have is freedom to make your own decisions about everything. You do not really have to be beholden to someone else's creative vision. You can just go at your own pace. Make the robot exactly how you would want to.
(00:51:53):
It is going to be a lot more work, because you have to do everything, and you might have to learn skills that you do not know. But you have a bit more creative freedom. So yeah, there is good and bad. There are benefits and drawbacks.
EW (00:52:09):
Are you familiar with Robot Operating System, ROS?
AM (00:52:13):
A little bit.
EW (00:52:14):
What are you using for simulation, for reinforcement learning?
AM (00:52:17):
Oh, we are using Isaac Sim.
CW (00:52:20):
That is NVIDIA's thing.
AM (00:52:22):
Yeah, the NVIDIA software.
EW (00:52:24):
I am not familiar with it, so I am just trying to figure out-
AM (00:52:27):
Yeah. Yeah, it is NVIDIA's simulation software. You can essentially load in your robot and train it to do what you want. All of those rewards that I talked about earlier, that is where you do it, in Isaac Sim.
CW (00:52:48):
Brief pause while Elecia looks this up, figures out if she should be using it for work.
AM (00:52:54):
Yeah. Isaac Sim I think is- Especially because NVIDIA is leading the semiconductor industry right now, their software is very important.
EW (00:53:07):
So Mecha is making quadrupedal robots, and you are looking at reinforcement learning for behaviors. Is this going to be another one of those military dogs?
AM (00:53:25):
Definitely not. We are doing the exact opposite of that. The goal is to make home robots, in general. We are starting with quadrupedal robots, because we think that is the easiest form factor to integrate into homes right now.
(00:53:39):
Yeah, like you mentioned, pretty much everyone who is making quadrupedal robots right now, is targeting more industrial applications, more military applications. Spot from Boston Dynamics, that is an industrial robot. I think the only other company that is making quadrupedal robots in the US is Ghost Robotics, and that is a military robot.
(00:54:07):
Our goal is to make a friendly robot. Something that is not going to be shipped off to war. Something that can be in your house, and be this all-in-one smart device for your home. Something that you can go on walks with. Something that you can ask questions with, that can be a personal AI assistant. Something that can help carry in your groceries. Something that you can play with, with your kids.
(00:54:31):
We want something that feels real, something that is expressive, something that is emotive. That is the goal. We want a cool home robot. Because we really do not have any robots out in the world that are doing that, that can be in the home right now.
(00:54:52):
A lot of people are focusing on humanoids. That has been the big buzz in the robotics world. But I do not know, humanoids just are not quite ready for the home yet. They are a huge safety hazard. These things are like 150 plus pounds.
(00:55:06):
What their people are promising that they are going to do, is just realistically another five to ten years away. There is no humanoid right now today that can go into any home and do a task. Like wash your dishes, or do your laundry autonomously. There is no humanoid that can do that today.
(00:55:36):
We think we will eventually get to that point, but there has to be something before that, something before the humanoid. We think the quadruped is the perfect form factor for that.
EW (00:55:50):
I like the things you listed, carry your groceries, play with your kids. Those are two different weight classes, though. Carrying your groceries, it has to be massive enough to withstand my groceries. But playing with kids, I do not want it to be more than 20 pounds, at that.
AM (00:56:08):
Yeah. Great point. That is the goal weight we are targeting as well, around 20-ish pounds. With a 20 pound robot, you can carry some groceries, maybe not all of the- Probably is not going to be able to carry 20 pounds of groceries, the same amount as its body weight. But you could probably carry ten pounds of groceries. Especially if you are not doing it for super long distances, you can carry some amount.
(00:56:42):
But yeah, that is very important for us, safety. We do not want this thing to be something that can harm you if it falls over on you. It is low to the ground, which is good. But yeah, we want it to be low weight. Something that you could pick up, or a kid could pick up, if you had to.
EW (00:57:01):
I spend quite a lot of my time trying to figure out how to make sure my robot will not fall and kill somebody. It is very distressing. It is very top of mind for me. Is it a toy or a tool? I am just going to add that to all of my questions now. Is it a toy or is it a tool? It came up recently on a different show, but I think it is relevant for you. Is it for fun or for usefulness?
AM (00:57:32):
Yeah, that is a great question. The goal is both. We want this to be something that can be used in research labs, for AI and reinforcement learning. But we also want this to be something that can live in the home. We are targeting that gap, because right now there is just a huge divide.
(00:57:53):
You either have- I think you mentioned the Sony AIBO earlier. So that is a pure toy, something that cannot really do anything other than sit on your desk, and do a couple of cool pre-programmed actions. And then you have something like Spot, or the Unitree Go2, which is just an industry tool.
(00:58:14):
I think the intersection between those is something really cool. Something that can serve very useful practical applications, like being an AI assistant. Being something that could even be a security device to monitor your house. But could also be something that you can go out and play with. Go to the field with and run with you on runs, and play with your kids.
(00:58:41):
So we think that we can target both ends of that spectrum, because the goal is to essentially make a platform that developers can create apps for the robot. We want this thing to be something that you can just use for so many applications, and all those applications will just live in an app store.
(00:59:06):
Maybe you want an app that can have the robot go and check up on grandma, and see how she is doing upstairs. Or maybe you have an app that can follow you around and record your TikTok or whatnot. And maybe there is an app that can go on a run with you. So we want to target all of those applications, through an app store.
EW (00:59:34):
I like that idea. But from an entrepreneurial perspective, I do not think you can keep it that open forever and you will have to make decisions. Making it a platform, the research, there are a couple other robotics platforms for research institutions. So there is a market there, not a huge one, but at least it gets other people to have time on your robot.
(00:59:59):
I like the idea of a pet most, because that is what I would like most. I am thinking, "Okay, what would I pay $2,000 for?" Not that I have any idea how much you are aiming for, but that seems like a not impossible price.
(01:00:18):
As a pet sitter for our dog to be gone all day, but then it has to open the door and that is kind of hard, so maybe not that. Would not take it to the beach. Probably would not take it on walks, because we have an actual dog and she would get jealous. I do not know. What do-
CW (01:00:40):
The things you mentioned like security, having it wander the house while you are gone, and be aware of things or be able to-
EW (01:00:49):
Fetch me water.
AM (01:00:56):
The Sony AIBO I think sold 150,000 units the first two years. It did really well.
CW (01:01:01):
Really! That is more than I expected.
EW (01:01:01):
It was so expensive. But it was so cute.
CW (01:01:04):
That is a lot of units.
EW (01:01:06):
Well, and it was programmable.
CW (01:01:06):
Yeah.
AM (01:01:08):
Yeah, it did extremely well. And again, this was in the early 2000s. So accounting for inflation, I do not even know how much that would cost today. But people were super excited about it.
(01:01:20):
The reason it failed is because it just was not feature rich. People would play with it for five seconds, and then they would exhaust all the features. There are only so many times that you would want your Sony AIBO to bow, or- I do not even know if it could roll over. But bow or do whatever pre-programmed sequence it could do.
(01:01:43):
Our goal is like, "Okay. There is interest. People will pay a pretty penny for a cool robot. But how do we get people to keep that robot, and stay interested in what we are doing?" It just has to be feature rich enough. So that is why the app store idea came about.
(01:02:00):
It is like, "We want a place where you can just do just a bunch of different things. We want you to be able to use it for very practical things. But also just use it for fun, playful things." I think that is really the only way right now, that you can get people to want to adopt robots into their homes, is if it does just a whole lot of things.
EW (01:02:29):
With the platform, you give people the opportunity to make their own applications.
AM (01:02:32):
Exactly.
EW (01:02:32):
Although there is only a small percentage of the population who is able to do that.
AM (01:02:38):
Yes. It is funny you mentioned that, because we are trying to close that gap. We want this to be easy enough where just anyone can create a cool application for whatever their specific use case is.
(01:02:52):
The goal is to make an SDK with high level APIs, where it is just as simple as typing in natural language for the robot to go do something, and having it go do that. So we want it to be very simple to make apps for developers, and make commands for end users.
EW (01:03:15):
Okay. I have an app idea. Christopher, you have to think of an app idea while I explain this one.
CW (01:03:19):
All right, great. I will-
AM (01:03:19):
Let us hear it.
EW (01:03:21):
Our dog sometimes does hysterically cute things. It is just who she is. I have wondered could we outfit the house with cameras, so that we could have her have her own YouTube channel. This is never actually going to happen. It is just fantasy.
(01:03:39):
But if I had a robot that could just follow her around, and tag whenever she is moving her toys from one place to another because they are in the sun or whatever. And then building on that, instead of just follow my dog around, you could have follow my toddler around. Alert me if she leaves this room. So the follow somebody around would be pretty cool.
AM (01:04:08):
Yeah.
EW (01:04:08):
Okay, Chris. Now your app idea.
CW (01:04:13):
Well, I like the idea of having something carry a bunch of tools and stuff, from wandering around the house doing stuff.
EW (01:04:20):
Oh, so it would be a- But he already has a mobile toolbox dog. You saw that in the video, right?
CW (01:04:25):
That follows you?
EW (01:04:26):
I do not know. But I did see a CARA that was a toolbox, which was hilarious. It was so cute. Did that toolbox actually carry anything, or was it all full of batteries?
AM (01:04:39):
Wait. Which robot are you talking about?
EW (01:04:41):
In one of the CARA videos. It looked like it was built into a- I want to say a Home Depot toolbox. But a standard toolbox shape.
AM (01:04:50):
Yeah, I think that is just how the robot looks. It has a handle, and it looks very toolbox like. It is definitely not a toolbox. But it can carry ten pounds. I tried it with- Or I think it was 15, actually. No, it was like 15 pounds. I do not know how long it could carry 15 pounds for. Did not test it for long durations.
(01:05:09):
But yeah, it could definitely carry a couple tools for you, and follow you around. CARA did not have any vision, so the follow you around part would probably be a little hard.
EW (01:05:23):
Are you familiar with OpenMV from Kwabena Agyeman?
AM (01:05:28):
I am not. No.
EW (01:05:31):
He has this amazing board that is MicroPython. It does all kinds of intelligence with recognizing faces, recognizing specific faces.
CW (01:05:48):
Objects, people.
EW (01:05:50):
It is a whole camera intelligence system built into one thing. It is really cool.
AM (01:05:57):
Neat. Yeah. We are trying to figure out the right perception stack, right now. We have a lot of different options, the robots. LiDAR is of course the main one, the go-to for being able to sense maybe obstacles on the ground and step over them, sense the terrain. Cameras.
EW (01:06:21):
But it is so expensive.
AM (01:06:26):
Yeah, exactly. So that is the thing. Yeah, LiDAR is very expensive. Yeah. Yeah.
EW (01:06:27):
Cameras are so cheap. We can get four cameras for the price of a LiDAR.
CW (01:06:34):
They have small LiDARs now. But yeah, they are expensive.
AM (01:06:38):
Yeah. So cameras are great with maybe recognizing faces and stuff. But using a camera for depth perception and obstacle avoidance is harder. But I think if you can do it, it is a lot more rewarding. Because that is what the Teslas do. And as you see with the Cybercab, it is a fraction of the cost of a Waymo.
EW (01:07:02):
If you had one camera on each ear, or eye- I guess you do not have a head, so on each shoulder joint-ish area.
CW (01:07:10):
They might have a head eventually.
AM (01:07:10):
Yeah, we will have a head for what we are doing with the startup. Definitely we will have a head.
EW (01:07:14):
Oh, good. I am glad. I do not want to say that CARA is creepy, but.
AM (01:07:18):
No. You can say it. Yeah. I think there is definitely an uncanniness with quadrupedal robots that do not have heads. I am definitely aware of that. I am a very harsh critic of my work. CARA, for example, even the legs, they are five bar linkages, which are not biological at all.
EW (01:07:37):
Huh.
AM (01:07:37):
But yeah, there is definitely this uncanniness with a quadrupedal robot that does not have a head. That is something we want to do, is we want this to look like a dog, really look like a dog. A lot of people will say, "Robot dogs," and they have just this quadruped with no head and the same front and hind legs.
(01:08:02):
But we want something that actually looks like a dog. So proper front and hind legs, that actually look like a real dog's front and hind legs. And an actual head. A curved body, not something that is like a toolbox, or that is just very boxy and very simple. We want something that looks intricate and biological.
EW (01:08:27):
What about a tail? And would it be just a tail that goes back and forth maybe? Or would you go ahead and make it a full tentacle, that it could grip things? Prehensile tail, or regular tail?
AM (01:08:38):
A tail is something we are also figuring out now. I think it would be cool if we had a tail that could actuate. I do not know if we would be able to have it be a full manipulator to pick up some stuff. But I think a tail that could just at the very least wag and move a little bit, would be really, really cool. It would be a small detail that I think would add to the user experience.
EW (01:09:01):
Oh. Yeah. Between the tail and a heartbeat- If you had some small vibramotor that gave a heartbeat indicator that you could feel, or a purr indicator if you decide a cat, you might not need a head.
AM (01:09:19):
I think we will have a head no matter what. Just because I think it will just look really cool with a head, because no one else adds heads to their quadrupedal robots. So I think that would be really cool.
(01:09:29):
But yeah, all this other stuff you mentioned I think would also be really cool. Like a heartbeat, that is not something I have even though about. But if you can touch the robot and actually feel its heartbeat, a thud, I think that would be really cool.
(01:09:46):
It is one of those things that you would assume that Apple would do, if they made a robot. These small intricate details that you were like, "Wow, that is really not necessary, but I am glad it is there."
EW (01:10:03):
Yeah. I have put heartbeats in a few small non-commercial things. It is really nice. It takes it from something that is clearly inanimate, to something that borders on cuddleable.
(01:10:16):
Oh, wow. I have so many ideas, but we are pretty much out of time. Chris, do you have any other questions you want to...
CW (01:10:26):
No. No.
EW (01:10:27):
All right. It has been great talking to you, Aaed. Do you have any thoughts you would like to leave us with?
AM (01:10:34):
Yeah. Yeah. My one thought is that robots are the future. I think we are going to see a very interesting wave of physical AI coming up next. We have seen the AI wave, and we are still in it, but the next step is physical embodiment. That is why I am doing the startup now.
(01:10:53):
I think the future is going to be very, very interesting. I think we are at a time for robotics similar to the early days of computers. We are going to just see robots everywhere in the next five to ten years. So brace yourselves. Yeah, it is going to be a fun time. I think robots are really going to change the world.
EW (01:11:16):
You are telling me I really should read that [Embedded AI] from [No Starch Press] book, and be prepared to start riding that.
AM (01:11:26):
Yes. Yeah. Watch all the sci-fi movies that you can. It will not be that terrifying. But I think robots will help with our day-to-day life.
EW (01:11:38):
We did not ask one of the lightning round questions. "Murderbot" or "Iron Giant"?
AM (01:11:43):
Iron Giant. 100%. Yeah, no Murderbots.
EW (01:11:48):
Our guest has been Aaed Musa, co-founder of Mecha and YouTube creator. Check out his YouTube channel @AaedMusa or check for the link in the show notes.
CW (01:12:02):
Thanks, Aaed. It was really good to talk to you.
AM (01:12:04):
Thanks. Yeah, it was great talking with you guys as well.
EW (01:12:07):
Thank you to Christopher for producing and co-hosting. Thank you to our Patreon listener Slack group for all the weird things that have happened in the last two weeks. And thank you for listening. You can always contact us at show@embedded.fm or hit the contact link on the embedded.fm website, where you can also find show notes, transcripts, and assorted other things.
(01:12:29):
And now a quote to leave you with. This one is from Miles Morales, one of the Spider-Men. "Everyone keeps telling me how my story is supposed to go. Nah, I am going to do my own thing."
