Embedded - 201: Accidentally Incredibly Dangerous

Episode Date: May 24, 2017

Shaun Meehan (@logiclow) joined us to talk about robot arms and stealth rocket companies. Shaun’s rocket startup is hiring; information about the job openings are below.  Shaun’s robot arm is an ABB IRB-2000 (video of Fred). Elecia was reading How to Choose the Right Industrial Robot when Shaun emailed. He convinced her that the MeArm Pocket Size Robotic Arm is the likely best choice for her machine learning typer project (which needs a better name).  All this led to a discussion of inverse kinematics, robot operating system (ROS), and OpenAI. SparkFun has a nice guide to selecting the right motor if the DC, servo, stepper section went by a bit fast. Elecia mentioned the TI Piccolo line as good motor controllers, assuming you aren't building an FPGA controller from scratch on your own. Repair cafes are a thing. Shaun was on The Amp Hour 220: Doctiloquent Dove Deployer where he talked a lot more about his robot pets. For more about Fred, the robot arm, check out LogicLow.com. Also, see Shaun's github repo, Fun with Flip-Dots (on hackaday.io), his intended page for big servos (Not Your Hobby Servo, also hackaday,io)  His personal site detailing new projects, motors, and fire-breathing dodo birds is ShaunAndKelly.com.   Shaun recently enjoyed The Professor and the Madman: A Tale of Murder, Insanity, and the Making of the Oxford English Dictionary.  Stealth Space Rocket Company Hiring Information We are a small, highly entrepreneurial team of rocket engineers with deep technical expertise who love to build things and relish the idea of a grand challenge.  Building on over a decade of technology development in rocket propulsion, structures, and avionics funded by NASA and DARPA, we are applying a fast-paced, hardware-focused, agile approach to space launch. Are you an engineer, hacker, maker, or physicist who has always dreamed of building rockets? Come help us build the hardware and launch the services that will open the frontier of space to the next generation of entrepreneurs. The company is in the San Francisco Bay Area, California. If you want to apply, email Shaun: space at logiclow dot com.

Welcome to Embedded. I'm Elysia White. My co-host is Christopher White, and our guest this week is Sean Meehan, a guy who spent two years in Antarctica, had his electronics orbit the Earth many times, and keeps a large robotic arm as a pet in his apartment. Before we get to talk to Sean, who I am super excited to talk to, I want to remind you we have a survey, a survey. Please fill out the embedded survey so we know more about you and what you like about us. It helps us figure out how to direct time for guests and the blog and all the other things we do. So please, please take three minutes, fill it out. The link will be in the show notes.

Hey, Sean. Hey. Good to talk to you today. Thanks for having me on. Okay, so I did give you a little introduction, but maybe I should ask you to tell us about yourself. Yeah, I am a happy hardware hacker is the way I like to describe myself. I have worked on everything from 100,000 volt, 20,000 amp laser systems to satellites. I've worked in Antarctica. I'm now working on rockets. I have a bunch of robots that I tend to and Sean's home for broken robots.

Yeah, I'm someone who sleeps very little and loves to work on hardware projects. Cool. Well, I know you listen to the show some, so I'm not going to explain lightning round. I think we're just going to get to the most important question, given that you did live in Antarctica. Sure. Have you touched a penguin i have not nobody has i think you're not gonna find anyone yeah i mean if he didn't if he hasn't no i haven't actually seen a penguin in the wild well in antarctica i've seen them other where other places traveling but yeah it's against the uh antarctic treaty to uh interfere with any of the wildlife so laws that's what's that's what's hard harming yeah yeah back in the day they definitely they did more than touch the penguins okay i'll just leave that

they uh they turned them into uh to bowling uh pins at the uh at at the McMurdo station. Yeah, they had a set of 10 stuffed penguins that they would bowl with. But then, that was back in the Navy days. The rules kind of changed when it became a scientific research station. All right, moving on. Favorite movie, book, or other piece of fiction which you encountered last year? In the last year, I read a fun little book called The Professor and the Mad Men about the history of the Oxford English Dictionary.

I thought that was kind of fun. Oh, that was a great book. Yes. Because Webster, I mean, he was, yeah, he was crazy. Yeah, he was legitimately crazy. Yeah, so I would say that's one that jumps to mind all right that's not fiction but it was so much fun i i totally will give you that oh yeah sorry uh preferred voltage uh anything above a thousand volts i would say jesus favorite fictional robot oh oh man it has to be it has to be fred uh my robot i i give him a lot

of uh personality that he doesn't really have okay i was gonna say he's real but you know if he's part real part imaginary that that works yeah I talk to him a lot for something that's not real. What is your favorite orbit? Oh, man. There are these basically Lagrange points. Yes. Yeah. Those are so cool.

Those are fascinating. I don't know why we don't have a base out there yet. Yeah, it's not really like an orbit, but it's a cool point, I guess. Well, technically it's an orbit. I had to solve for those in school, I remember now. Maybe you should explain what they are. They're these weird resonant points trailing an orbiting object or between an object and what it's orbiting where you're stable and some of them are metastable so you you can kind of be knocked out of them easy but some are very stable and you can just sit there there's some between say the

earth and the sun there's some trailing the earth and leading the earth i think i can't remember them all and then there's another one uh away from the earth but in line with the sun on the earth yeah i think there's five five per every two orbital bodies or something like that i could be totally wrong too i think it's five yeah and what's weird about them is i think you can orbit them so they act like virtual there's sort of little gravity centers that don't make any sense because there's nothing there. Yeah. And it all is Kepler's fault because it has to do with ellipses? I don't remember.

That's all gravity's fault. Okay, next question. Do you have a safety tip for us? Always keep one hand behind your back. Because that's the one you like?'s the one you only got one left you only got one left you're you're totally fine to work on the high voltage but yeah um that's the uh the safety tip for um anytime you're working with a high voltage or i even do it with uh fairly low voltage stuff is um you are less likely to encounter a current path across your heart if you're holding

right one arm behind your back so and you've already mentioned your robot fred but if you got a robot dog what would you name it oh maybe sparky all right all. I'm into that. So I heard about you for the first time on the Amp Hour, episode 220. And it was really cool because here was this guy with this robot that he somehow got. And now his whole life was maybe I'm not really doing a good job of explaining this but that episode was really fun and it had a lot to do with fred yeah so you have a robot tell us about fred uh so fred is a one ton abb robotic arm it's a abb's a ABB IRB 2000. I found Fred on eBay. I actually wrote a little pearl scraper, and I found basically looking for orange robots because when I was a kid, I had seen one of these robots operate, and I'd fallen in love with it. And so when i got back from antarctica i had some disposable uh cash because i'd been living

on an isolated research station for a while no amazon there yeah exactly and i uh found found a broken uh a broken arm on ebay and i think i put in a bid at seven hundred dollars uh like the within an hour of it coming online because my scraper had notified me that there was something cool. I put in a bid, it was mislisted in eBay. And so a couple days went by and no one else made a bid.

And then a few more days went by and it was November 11th, 2011. And then it hit. And I realized that I now owned 2,700 pounds of freight in Connecticut. And I lived in Colorado. And so then I had to find a logistic adventure to get it back home. I mean, these are like $20,000 robots. Yeah, brand new.

They're very expensive. Fred is from 1987. He was very broken. And yeah, so they've gone obsolete and they're being removed from factories all over the place. A friend of mine a while ago sent me a photo of a field in, I think it's New Jersey, where there's just a pile of these things

that some factory basically dumped out back as they were upgrading the line. So they're around if you want to find them. And the mechanisms inside are beautiful. It's built like a watch. The gearboxes only need oiling once every five years of continuous operation. The bearing cases are... Everything about it is just beautiful inside.

And you take apart the electronics and they smell like a new radio. I mean, these things were impeccably built and they've just been replaced by, I really think, a new communication interface and higher efficiency transistors. Where were they made? So Fred was made in Sweden. Okay. And then I think, I'm not sure where Lefty was. So Fred, before I had him, was a jet engine welding robot in Connecticut.

And so he welded jet engines for Pratt & Whitney before he became my buddy. And Lefty is the other one? Yeah, Lefty used to be a packaging arm at a Kodak factory. They're both identical models, both from 1987, except that Lefty is the higher resolution version of FRED. So FRED and Lefty can both move at 300 degrees per second on all axes. And that's with FRED is sub-millimeter accuracy, and Lefty is, I think, 0.1 millimeter accuracy. And lefty has absolute position information, whereas FRED doesn't.

How often do these run? I mean, are they just sitting in your living space waving to you when you come home? Yeah. Oh, I know. So I took on the project of building their controller from scratch. So the old controller ran on three-phase 480 uh these phase converters that a rotary phase converter it just runs a three-phase motor off single phase power and generates some pretty nasty uh three-phase off of that basically running the motor like a motor generator and the transformer was bought

from a guy who definitely stole it from somewhere and it was designed to step down from 480 to 220 and i just ran it backwards and so stepped up to three phase 480 in the garage and and that was pretty exciting for a period of time um but it wasn't a practical uh system for the home hobbyist. So I... And when you have a multi-thousand pound robot arm, you want it to be practical. Exactly. Two of them.

Yeah. So what I did was decided that I was going to throw away the old controller cabinet, and I was going to design the highest efficiency controller I could and I was going to have the entire thing run on FPGAs. I didn't want to have any procedural code inside the controllers or was kind of the idea behind it. And so that was a very quick project that I finished in a weekend. Or, no, no. It's a project I've been working on for about five years,

and I'm nowhere near done with it. I was about to take issue with that. Wow, he's really fast. No. I have a terrible VHDL repository on GitHub about, I think it's called like the Fred controller or something. And yeah, I have, I've built like PWM blocks and dead time blocks and commutation controlling blocks, which is how you control the permanent magnet motors inside him. But those are such a pain and resolver to

digital converter blocks and all this. And, you know, I realized many times during this project that if I had just thrown this into, like, an STM32, I would have had it running in a couple weeks. Yeah, I didn't want to say. Yeah, it's really sad. I would have actually suggested the TI Piccolo line. They have a lot more motor controls with built-in code and commutation libraries.

And then you have Dave Wilson on making all those amazing YouTube videos, like stepping you through how it all works. And yeah, that would have been a much better decision. But I don't know. It was kind of fun, too, because part of the challenge was I wanted to understand how every single

block works. That's what I really enjoy, is finding out how these subsystems that we normally implement through register addresses are actually implemented in silicon. And that's the cool thing about FPGAs, is they force you to understand every waveform, every transition. Yeah, I remember doing a one-wire bus master or working alongside somebody who was doing a OneWire bus master in VHDL. It was really cool to see it from that aspect instead of the code aspect where, okay, I just tell this thing to go or whatever. Yeah, and I really love the fun thing about FPGAs too, which is if you only need two bits, you only have a two-bit bus.

If you only need three bits, you have a three-bit bus. And that's just kind of a... It's kind of fun to see all the optimizations inside that you can do and kind of ways you can play with math inside that. It's funny to hear you say that you wanted to understand all of the pieces about it, because when I'm going to ask you all my questions about robot arms, my goal is to just get to the end so that I could play with other things.

Yeah. But before we do that, I actually, you emailed me and had a question, and I emailed back and said, oh, good. I have a whole bunch of robot arm tabs open. I'm so glad you emailed. And yet, I feel like I should actually maybe work on your question a little bit, which was, you were looking for people to work with. Yeah, I'm trying to build out a team. So I have just taken on a role at a new rocket company.

We're building a launch vehicle in the Bay Area. And we're looking for avionics, electrical. We're looking for some mechanical. We're looking for embedded. We're looking for controls engineers. Yeah, so we're trying to find great people. And in the past at Planet, we reached out through the Embedded podcast

and we found some really great candidates there. So I wanted to do the same thing. I wanted to find some happy hackers who want to work on a rocket. So it seems like it's a pretty great project. And so I'm trying to find the kind of people I want to work with, which are the hacker maker community.

And it's in the San Francisco Bay Area. Can you tell us what part of the Bay Area or not without an NDA? i don't think i can't say anything i am i am so is it underwater so hobbled yeah it is above water i can say that much um it is in a pretty fascinating facility um it's about there's about 30 30 rocket scientists and me crammed into a room. And we are... Sounds like fun. Yeah, it's really fun. I mean, we set up a pinata downstream of one of the engine tests the other day.

And that was pretty exciting to watch a pinata get blown to kingdom come by a rocket. But yeah, we're a venture-funded rocket company. And we're looking for embedded and controls engineers who want to help build some really cool hardware. Can you believe that we live in a time where you can walk up to someone and say, I work at a rocket startup? Well, I mean, yeah.

It's just... Ten years ago, it was pretty much, yeah, okay, if you worked at Boeing or Lockheed. Rocket startups. Those are startups. And now there's many of them, right? Yeah, there's a whole bunch. And everyone's working on...

The cool thing is there's a ton of satellite startups, too. So the industry is kind of like catching up. Because especially for Planet, like when I was at Planet Labs, we were kind of hamstrung by getting rocket launches three years out. Like that was when you could get a launch. And so there's a lot of kind of new rocket startups coming online that are trying to drop the cadence of time to launch. So the goal for this company is to launch on a much more aggressive cadence. And we talked to Planet a few times. I know we talked to Patrick Yon and Kat Scott.

And then later, not too long ago, we talked to one of the CubeSat founders, Professor Pujswari. And there's this, I mean, Planet is making flocks of imaging satellites, and the idea is that they're not disposable, but not super expensive. They're agile-ish. They're new space. It's okay if they fail because we're putting a lot up and they're small. And it's such a different mindset. Yeah, and that's actually the mindset that

this rocket company has started with. So we're trying to do that same agile approach to rocket design, which is going to be incredibly exciting on the launch pad because the agile approach is to fail fast approach to rocket design, which is going to be incredibly exciting on the launch pad, because the Agile approach is to fail fast, and so we're going to see some

exciting launch failures, I think. I wish you the best. I have no comment on that approach. It's going to be fun. I mean, the technology has advanced so far that, I mean, we have these things like littered with processors that the end telemetry streams that the Apollo teams could only dream of. It really does seem like there's a required minimum number of catastrophic failures before you.

I guess it's just a bathtub curb of rockets or something. Yeah, exactly. And I think the challenge is when you're when you're first starting you're not really concerned about mass you're not concerned about you know how much payload you really have because you're trying to figure out the launch vehicle and then as you start to trim away and make your vehicle as light as possible that's when you start to see more failures later on down the line because you're trying to get the most payload possible into orbit okay this you know there are some people out there going you're trying to get the most payload possible into orbit. Okay, this, you know, there are some people out there going, you've got to be kidding me.

There's no way I would work on exploding rockets. Really? And then there's me going, this is really tempting. I could watch rockets explode and get all the data. That would be so fun. I mean, I've worked on things that burn and things that, you know, I haven't done exploding yet. I know, exploding. Let's yet. I know, exploding. Let's go.

I mean, yeah. I mean, you're more than welcome. Not that you want them to explode. I think we should be clear that we don't want them to explode. This is like talking to Patrick and Polarski and threatening to chop off our arms so that we could have new prosthetic limbs. Right, right. You don't want the rockets to explode, but you do. But if they do, you want a good seat.

Yeah, and you want an amazing data set. And I think that's the exciting thing, is that was what really drew me to Planet, too. Planet early on was almost all fundamental research people, or with backgrounds in fundamental research. And the data set was really what drew a lot of people there and that's the cool thing about you know we're doing rocket engine tests uh several times a week now and the data that's coming off there is just incredible to analyze and it's all up on the

server so you can go and check out the pressure transducers and you check out the thermal couples and just kind of it's just great to see tons of data. I don't know. I love data. Yes. Yes, exactly. And exploding things and data always go very well together. You have to sample quickly when things explode. That's true. Yeah, exactly.

And that's one of the challenges, too. You look at... The data throughput. Yeah. And you can't log anything. If you're logging and then sending down these packet bursts, then the data you really want is halfway in that buffer. Right. We store all our telemetry on Amazon S3.

Yeah, as long as you can get it off the rocket quickly. As soon as you measure it, get it off there. That seems to be the number one telemetry rule, as far as I'm concerned. All right. Well, if we talk too much more, then I'm afraid I will get you to say something that you're not supposed to. And I do know that rocket companies can't say a lot. So let's go on to my problems. Sounds great. Where to start? Okay, so I want a robot arm that I can use machine learning techniques to drive, and I want it to essentially type. And I don't have any real reason to do this other than I think it would be hilarious to make a robot type what I'm telling it.

I fully support this i support more robots doing strange things that is my that is what gets me out of bed in the morning is building building weird things um and so yeah i have some of the machine learning stuff i have i have a jetson tx2 that i got from a friend and I can identify things. It's very good at determining when things are banana or an apple. Or a frog. And a frog, although sometimes it says the frog is broccoli. I'm teaching it

about keyboards and helping and it's learning to detect which key is which. And so I think I can say find the A key and it say find the A key, and it will find the A key, given about six more hours of work on my part. But I'm close to that part.

And I have heard that you can make arms, you can make systems where you don't have to tune the robot, you don't have to tell the robot what to do. You kind of have it learn what to do. So I have that as the idea. Okay. There's a group that's working on this. This is the OpenAI group.

They just published, I think, something today on, I think, robotic stacking of blocks or something along those lines. So you're worried about the interface between the computer and the robot in terms of the drive signal that you want out of that? Or what would be the output to get a motion from the robot? So, no, not so much. Because I'll probably, I mean, depending on where we go,

I mostly want to know what I need from an arm. I mean, they have $50 arms that you can take apart and then rewire to an Arduino. But is that what I need? Or should I get one of those Mii arms that are really cheap but also look pretty cheap? Or we splurge and get a good arm. Yeah. Oh, yeah. I mean, I highly recommend.

Do I need one ton or two? Yeah. I highly recommend. nothing measured in tons please the the old uh car welding arms i think they are they for the dollar you really can't nothing nothing that we have to put multi-phase power into the house the the the bigger concern is you have to make sure that the floor can support it. It can't. Actually, no, my office really couldn't, could it? No. I think those little, for a key press application, the little servo input robotic arms, the ones that just have the tiny servo motors mounted at, like the little hobby servos,

mounted on the axes of the arm would be perfect for that because you don't need a ton of force on your keyboard unless you're using one of the DOS keyboard red key super clicky things that Patrick uses that wakes up the whole office. There's a little underhanded jab there. Those are for private use. They really are. Yeah, you don't bring They really are. Yeah. You don't bring that to work.

No. But I would go with that because it's so simple to integrate to the Arduino. And then it just takes in a, you know, a 0 to 180 or whatever the signal is for Arduino. I forget how that is. PWM. Yeah. And they even have this great little servo library. And that would be a fun thing to plug into that. And I think part of the joy of working with robots

is getting one that is either drastically oversized or drastically undersized for the application, in my personal opinion. And so I think one of the little laser-cut hobby servo robotic arms would be perfect for that. Okay, some questions before. Sure. Okay, servos or steppers?

I mean, steppers are bigger and better, aren't they? Servos are just the hobby things. Well, you should say that. So I just threw up a hackaday project I'm kind of working on in the background, which is a multi-kilowatt servo motor that's designed around like the hobby servo one. But I can link that one in. That's kind of a strange hacker project scaled up a little bit. But stepper motors are fun. They are big, they're heavy. For the amount of force that you need to generate i think they're they're a little bit overkill but um yeah the the stepper motor control is is good as long as nothing skips

a beat as long as it's uh if it doesn't have any kind of feedback that's your that's your main concern but i we should walk back i mean what is the difference between a stepper and a servo oh so um a servo motor is actually if you if you rip it apart inside you have a a tiny dc motor and then you have a potentiometer which is uh like a variable resistor mounted underneath the gear train on the output. And there is a tiny circuit board inside there that takes in the input PWM signal and then will either run the DC motor forward or backwards until it gets to the position indicated by the potentiometer or requested by the PWM signal coming in.

A stepper motor is actually just a series of coils inside that you step in a direction. So let me see how I can describe this over the air. I really love drawing pictures. I love answering this question. So stepper motors, I'm making a lot of hand gestures here. So the stepper motors, uh, if you think about it as a compass needle, um, and you put four coils around your compass needle and you energize the North compass needle, or you energize the coil at North, and then you energize the coil at West, you're going to have counterclockwise rotation of that, of your compass needle. Until you get to west you're going to have counterclockwise rotation of that of your compass needle until you get to where you're pointing to the second one yeah exactly and then it's going to stop and you have to energize the next one in order to continue

spinning you have to step through yeah exactly and you're actually you're energizing both the north coil and the south coil in opposite polarity to to generate that um that field that you step through and that gives you a push it gives you a pull and a kick you can do just one but you but almost always you pull yeah and a lot of times they're wired in such a way that you have when you energize uh one path it does energize the other path as well um and so yeah and that's the um how a stepper motor will that's why they call it step stepping it literally steps through the as each tooth increments inside the inside the the motor housing there that's why if you feel a stepper motor stepping you can feel it actually locking between these teeth yeah it's like whereas on a dc motor you don't feel that yeah exactly and it can hold really strongly too it can hold because it will stay where it's it's stepped to

and someone could do micro steps and i don't remember how that worked but yeah that's great um that's actually if you take if you were to take the north and the west coil and you were to energize them each at the same time you'll generate a a field where the compass needle will be resting between northwest okay yeah and then uh as you vary the the intensity of north and west you can vary the position and this varying intensity is uh is part of the commutation thing we we said commutation and then we just like zipped right by it but it's this uh pattern of energizing things and it's not that you just turn on north and then we just zipped right by it. But it's this pattern of energizing things. And it's not that you just turn on north and then turn on west,

because that hard turn on is kind of bad for the motor. Usually you have a sine wave. And so you hear about sine commutation a lot if you end up having to drive motors like this. Yeah, especially if you're dealing with brushless DC motors. That's where commutation really is important as well. A lot of times you can, like with a stepper motor, you can drive it open loop,

which is without any kind of feedback system. But if you're working with a permanent magnet AC motor or brushless DC motor, you actually, it's critical to know where your motor uh rotor is which is the center part of the um the motor mechanism and then you're energizing your stator which is the outer portion based upon a a phase leading or lagging of your current position and that's um and that's my like hacker way to describe it i'm sure there's sure there's

people listening to this saying i'm off by an order of magnitude but um this is kind of the way i think through these things no no it's it matches my mental model pretty well okay so so stepper motors which like i said i usually think of as a hobbyist thing it's something you use in airplanes and models servos sorry servos servos are are cheap um not necessarily actually pretty expensive yeah but since they're based on a dc motor which really is cheap because it's dumb it doesn't have any feedback the stepper the the servo the servo um it has it's just a dc motor with a little bit of feedback a little bit of smarts and it isn't as intensely powerful with holding as the stepper motor is but the stepper motor doesn't come with position i mean sure yeah if you want to go 200 steps and you know about the motor you

might be able to say well it went 30 degrees around um but unless you have unless you know where it started you don't know where you are yeah and also the the other thing is you never know whether or not you've missed a step. So let's say your robot bumps into a wall or bumps into your keyboard, and it's continuing to step and continuing to move through the excitation of the coils. However, the motor is no longer rotating, and that's something that your open loop controller would have no position information for. And so while you think you're at a position, you're actually offset from the expected position. And so usually you get an encoder,

and that's how you know where you're located, or you have to do some homing thing at the beginning and then hope you don't miss any steps. Yeah, exactly. Okay, so that's the motors. The next thing about the arm that I'm still having a little bit of trouble with is apparently 4-axis and 6-axis don't mean the same as they mean in inertial measurement units, which is where I think of things. Okay. So what does a 4-axis one do versus a 6-axis?

And are all of the axes labeled the same across manufacturers? Not that I've noticed. So, FRED is a six-axis arm, and what that usually means in robotic arms is the number of actual rotation axes on the arm. So, it's a count of the number of points that the robot will rotate around and so fred has the first three axes are the base which will rotate um in the i guess around the z axis so you have xyz at the base and then you have xyz at the at the wrist and so that's a six axis arm and you can have seven axis arms. And those are used a lot in conditions where you can't have any axis failure. So if you're operating inside, say, a nuclear

reactor or something like that, if your robotic arm loses one of the axes, you actually can no longer fully position the tool within the entire environment, within the entire reach of the robot. And so if you have an extra axis, then you can rotate around and make it work with, with six. So, yeah. So,

so it does the gripper count. Like if I have something, usually they show claws at the end. Yeah. And so they're going to call the claw an axis probably. Okay. yeah and so they're going to call the claw um an axis probably okay and usually the first axis is the one that uh it's sort of the waist it it turns the arm back and forth wow this is not going to be possible without visuals is it um it might not be i i feel like i'm i'm dropping the ball in

terms of being able to describe this um yeah flashing back to the calma filter episode sorry yeah oh my goodness yeah um okay so i want to be able to mount my my base on the desk and i want to be able to put a keyboard in front of it and so it needs to be able to turn back and forth so it can get to the different keys yep and needs to be able to go forward and back and up and down so it can can put on so it can push one key at a time that sounds like three axes so yeah that should be all right forward back up and down well you might want to risk to hold it so it's always so it's angled so the pusher finger thing is always angled properly yeah yeah okay so four axis so that actually there i saw a lot of hobbyist ones that were six axis and a lot that were four axis so that that makes sense

yeah and so really the the benefit of six-axis arm is that you can hold your, basically your tool surface or the outside plate of the robotic arm in any orientation within its range of motion. And that's important if you're basically, if you can think about the robot moving like a mime across a surface. So you could have the arm holding a plate and moving across without breaking that plane and with a three-axis arm or four-axis arm you wouldn't be able to do that maybe if that made any sense did it make sense to you sort of will you explain it to me if you

think about uh like a mime moving inside a box oh my god we're really we're off the rails uh okay um okay so i have i have what was his name marcel marcel yeah if marcel marcel was trapped inside a glass box and he's kind of moving his hands along that glass box uh to keep his hand against the outside surface of that invisible plane he would need to have a six degrees of freedom at minimum oh i see you can think about this i see if you yeah and then the top of the box and the and the bottom of the box and each oh that's a totally different way to think about axes mine if you were mine if you were to hold uh hold a glass of water this is, here's something you

can do at home. I can make people do the visual I'm trying to make. If you hold a glass of water and you bring it across your chest and you're trying to maintain it along a flat table that's in front of you, say, you can actually see your arm moving around to try to maintain that plane in a way that wouldn't be uh that you wouldn't really think about your arm moving until you actually do that behavior so yeah i guess it's both my elbow and my shoulder need to move don't they yep okay all right let's stop doing that. Let's go on. Okay. Okay, so some of the arms have built-in controls. And I know for right and lefty, you're rebuilding their controls. And for the ones I saw online for the sub-1000 dollar ones,

the controls didn't look that good. Yeah, I don't think they... Well, I don't know. There are open source tools you can use. You have to write a fair bit of code, but I don't think that's a big barrier to entry for you. To interface with them, like the Open... I think it's Open ROS?

Like the Robot Operating System? Yeah. And in has inside of it all the inverse schematic solvers and everything you need to actually move a robot within uh an area if you're trying to i guess it depends on what the control inputs you want like you could output from your um your neural net or whatever you're building a PWM signal. And it could just be 0 to 180 on all three axes. And it could try to solve for the motions it needs to make to actually successfully strike a key. I'm totally into gradient descent right now. Totally into it.

So that could be a way to go about it. Okay, you said inverse kinematics, and that was on my list of things to ask you about. Mm-hmm. What is that? So this is my understanding of this. It's the study of how you get a robot from point a to point b and so you build a a model for your robotic arm which is a lot of uh basically a lot of vectors

or not a lot of vectors a lot of um a lot of matrices to say that this is the point in space that you are at and this is the point in space you want to be, and orientation. And so an inverse kinematic solver will basically take your current position and your desired position, because to move a robot from point A to point B in space and orientation A to orientation B, there's an infinite number of paths. And so you can solve for minimal energy, you can solve for

fastest speed, you can solve for like any number of ways to move between these two positions. And that's kind of the job of the inverse kinematic solver is my understanding of it. Okay, so if I want to press the letter H, and then I want to release the letter H and then I want to press the letter P. As a human, I see H, I press it, I pick it up, and I move over to P a little bit. I move over and a little bit up to P. And all I care about is the tip of my finger. And with my robot, that's actually all I'm going to care about too. I don't care if it puts its elbow up or has bad posture or anything like that.

I only care about the pressing surface. And this inverse kinematics is the way that I don't have to care about its shoulder, elbow, joint 1, 2, or 3. It's only caring about that one point in space and moving it from where I am to where I want to go. Yeah, and it takes into account the model of your robot. And so the positions, kind of the output of this solver is a, you can think about it as a degree that each of your axes will rotate. And it's also important to say the path that those, or the speed at which those axes rotate. So if you wanted to move from depressed H to depressed P, if you try to just directly move in the line of shortest travel,

you'd rip apart your entire keyboard if you did it with Fred, which he's done to me in the past. He put a refrigerator through a drywall which was pretty exciting and you don't have video of that i was planning on destroying some keyboards but yeah that's that's a tragedy that's almost something you need to reenact i i have no desire to reenact that. That was exciting. Not even under contract? Well, maybe. I was trying to teach him to take beers out of the fridge and open them, which he did. I did get a video of him successfully opening a beer with a bottle opener.

But yeah, he basically opened the refrigerator door. He had his little suction cup and he reached into the refrigerator and um he he got to the beer can and he um drew the beer can onto him and then just decided to keep going forward and then keep going forward and keep going forward and uh when a multi-ton arm decides or when a one-ton arm decides to keep going forward while inside a refrigerator the refrigerator uh is no longer a stationary object in the outside world and so it uh continued into the wall and then uh fred through one of the the best errors which was uh over current on like axis three it's like oh i tried to go to where you asked me to but it you know there's a building and it's i i just love how casually this robot will destroy things and then throw a very calm

error saying like oh i drew a bit too much current um something may be in my way uh what did your neighbors think of the refrigerator entering their apartment or leaving your house with a giant robot arm attached to it? When I was illegally living in a warehouse in Colorado, and my neighbor was, ironically enough, a safety company, which I just thought was great. They sold all the traffic cones and everything exciting there, but they got a kick out of the robots. I set off their security system once or twice by accident by doing silly things on my side of the warehouse. But yeah, everyone loves the robots, I got to say. People either love them when they meet them or they just kind of don't get it.

Have you met Simone Yacht yet? Oh yeah, actually. I met her through Alvaro. Yeah. Yeah, so have you introduced her to Fred?o. Yeah. Yeah. So have you introduced her to Fred? I don't think so. I don't think she's met the robots yet.

I'm just thinking that that might go horribly and you should video it from the very start. Oh, my goodness. I know. When she started posting her kind of, I don't know't know derpy robot videos i guess it'd be the right way to describe it um i really wanted to make like a series of like fred doing the same activities but kind of with a buster style uh dummy just because the toothbrushing robot um but with fred would, would be so exciting. Yeah, those robots are just, they're so wonderful.

And they mean so well. And yeah, they are, they're just the best. I really love my robots. They're kind of like my 69 chargers, the way I describe them to people who don't get it. As I say that, they're this project that will sit in my garage for the rest of my life that i'll constantly hack on and constantly work on and the people who get it will absolutely love it and just think it's the best while it like drains all of my income and you know yeah destroys all my relationships but no it's uh no they're great i i do i do love the robots dearly

but okay so you were teaching them to fetch you beer oh what else have you you taught them to do i one of the coolest things i think i did with them was i put a uh a bike light on them um one of those like really bright bike lights i turned off all the lights in the warehouse and i had them go through a path and it would basically draw this like sphere around the robot and light and uh with long duration photography and I thought that was just the coolest thing it was so simple but um that was something that when I got my controller up online I really wanted to build was uh a way for them to kind of draw pictures and, uh, and things like that in three space, because you could build a really cool like map of the world or something

like that in three space while you're moving this robot around. Um, so, uh, that was a, that was a fun project. Um, I gave him a paintball gun once, which was the worst decision I ever made. Um, I, uh, for, for some reason i got i got hooked up with a a think tank trying to find like non-military applications for a super high resolution thermal camera um which i i still don't know like i was in a room with a bunch of uh a bunch of incredible engineers and uh i was the youngest one by by 20 years and at the end of the first meeting there was like does anyone want to take this thing home and play with it and like my hand shot through the air and they're like give it to the kid see what he does with it and and so uh and after the after the uh the think tank the last meeting um

i i asked the guy, I was like, do you need this thing back? And he's like, oh no, here, wait, yeah, bring it to the hotel. And I went and met him at the hotel and he signed into it over the web server access and basically opened up every feature it had and gave it back.

And he's like, have fun. And I think that was his way of telling me that everyone else on this group got paid. But so then I took it home and it basically had the ability to like detect, it had some really simple kind of AI or image recognition built into it. And so it would like detect a person. And I was like, Oh,

how cool is that? I was like, Oh my gosh, I could go borrow a paintball gun from a buddy and i was like oh how cool is that i was like oh my gosh i could uh go borrow a paintball gun from a buddy and i could give fred a thermal imaging camera that would detect people and then uh he could uh i don't know dude just shoot paintballs this is how ed 209 from robocop got made right did you forget you were people i mean did no i i it was just like it was i am often blinded by the excitement to build something without really thinking it through um which might uh kind of

describe how i got to wherever i am now but the uh so yeah the robot came online and he, this is before I got like a thermal camera integrated. I had like little patchworks of things to wire it together. And I finally got the paintball gun wired up on a solenoid that the robot would trigger. And he just kind of like spun up and spun around and fired a bunch of paintballs against the wall. And I looked at this robot holding a gun and I was like this is the worst idea i've ever had and i just stripped it all apart and i threw the paintball gun in a box and i just i just every now and then i find this paintball gun in a box and i was like

that was that was a really bad idea that was a terrible idea but uh so i did that. I think there was another accidentally incredibly dangerous one, which was I put a whistle detector on him, which was a video I posted up on YouTube a while ago, Whistlebot. And I would whistle and the robot would come over. And that was like kind of the gimmick. And I'd whistle and he'd go away. And it was really fun, but I'm like, at my heart, a very happy person who's always wandering around whistling. And I would forget that the robot was on. And so I'd be kind of like walking around and I'd whistle and the robot would like spin and knock trash cans over and, or would like spin towards me and when you know the the main bulk of his arm

weighs about 700 pounds and when that's spinning at 300 degrees a second it'll it'll teach you a lesson about knowing where to whistle you know how when the dogs get super excited and they leap on our legs and in our laps and we're like kind of pushed over but they're not made of metal they're they're they're only like 30 pounds yeah i'm surprised you're still alive uh i've i've done that a lot that was that was one thing that um like i have many friends i consider wearing a helmet in the house yeah exactly told me that's like oh yeah i had this dream you built this really cool thing but then it killed you and i was like what what kind of a dream is that's like, oh yeah, I had this dream. You built this really cool thing, but then it killed you. And I was like, what, what kind of a dream is that? I was like, oh, my main concern was like, was I nice? You want to make sure that the other versions of me are

still being nice to people. Um, but yeah, the, I don't know what other, what other robot tricks I had him pour champagne. I think that was the first thing he did was I basically duct taped a bottle of champagne to him when I first got him up and running which took about two years to get him to to move for the first time and uh yeah when he when he poured champagne that was that was pretty great that was a fun moment but yeah it's it was fun when i was teaching him to open glass beer bottles, I had a church key on him, one of those big like bartender bottle openers. And he pushed it down into the beer bottle so hard he bent the thing in half. And the table was bowing underneath it. And this glass bottle was sitting there with all the force of this robot pressing down on it. And that was another time where Fred was like, oh, too much current. It's just like there's a bomb sitting in front of me and I'm not quite sure how to disarm it.

I was just like, okay, this is going to be a learning experience no matter what. But yeah, that was fun. It was a bunch of fun projects with a robot like that but now he's uh so i left his old um his old brain in colorado and i've been working on the new one out here and so i only have uh about one axis of control up and running and so i have uh he's moving again for the first time in about two years. He's up and running, and that's pretty exciting. Even though it's only on one axis, it's fun to see him move

and to kind of develop the FPGA code. How loud is he when he miffs? He's dead silent. It's incredible. There's a four kilohertz hum, which was actually this really incredible moment. I remembered from when I was a kid and I saw these things move for the first time at a factory that they moved so quickly, but they didn't make a sound. And that just astonished me that something that was so big could move so precisely and so quietly. And there was, I had this memory in the back of my mind of this four kilohertz hum and the first time i got the uh the resolvers which are fred's um

position feedback sensors up and running uh you excite them with a four kilohertz sine wave and so i turned on the uh and i you know, I was writing all the code to do this. And I threw the switch on and I heard the four kilohertz hum again. And I knew what it was because I was generating the raw signal from the op amp. And it was ringing through these six resolvers on board Fred and resonating through the whole structure. And that was a really incredible moment to have known how the actual core of the robot was working and where that 4 kHz hum came from. And to generate it without even knowing what I was really building. That was probably one of the greatest FRED moments I've had.

Yeah. I forgot about that you're in the home that's so cool yeah most of the arms i looked at which weren't nothing like fred realistically were quite loud though how why is he so quiet i think it's just the, he is a really a precision instrument. The ABB arms that there's only one or two models, I think that came before Fred. Fred was really one of the very first robotic arms and the, or one of the first commercialized robotic arms. But the first ones that were put out by ABB are still in service.

The first two arms are still doing the exact same job that they were built for in, I think, 1985 or 86, something like that. And so these arms have been operating for nearly 30 years, and they are designed to have no downtime. The gearboxes are impeccably designed. It's all, I think, drastically over-engineered for the application, but that's kind of how the first prototypes are. And so I think it's just a testament to engineering. And that's one of the reasons I love them too. When you take them apart, the attention to detail and the care inside of it is so clear. And that's something that I look for in engineering is the clear and that's something that you know i look for in in

engineering is the attention to detail i think you can if you're not a subject matter expert or if you're not um you know used to working in a certain field as long as you have attention to detail usually you can you can make it through but yeah okay switching subjects a little bit although sticking with that idea you did not go through the traditional four-year university engineering degree to get where you are you've done a lot of learning on your own yeah absolutely how do you do that i mean do you read a lot of books do you watch a lot of videos do you just try everything how how do you a how do you do that? I mean, do you read a lot of books? Do you watch a lot of videos? Do you just try everything? How do you, A, how do you motivate yourself? And B, how do you just do it?

I would say that, yeah, I mean, so I could give like a little backstory on me. For those who haven't heard my Amp hour podcast uh when when i was in high school i started working with the uh the national science foundation uh developing these massive laser systems um so i was just kind of like a research uh assistant at the national science foundation lab and uh we went as like part of a high school trip and they said would anyone want to come back this summer and help out? And I lost it. It was the most exciting thing I could think of, is walking into this massive engineering facility and working with these amazing PhD, postdoc students, these incredible engineers and scientists building X-ray lasers.

That just sounded like a really, a really weird futuristic thing. And that kind of got me in the environment of people who really loved what they did and love science. And, and I've always had a love for science and for figuring out how things work. So I've always taken things apart. And I think the, the challenge is taking it apart in such a way that you can put it back together and then the next step is to upgrade it and then the next step is to build your own and uh so after after high school i went down worked in antarctica which um i went down to to shovel snow basically with the ice cube neutrino observatory and i spent the whole

uh summer in the uh radio shop basically fixing all the broken band equipment and all the old like nintendo 64 i reverse engineered the uh kind of the audio and video output channels on the n64 and wired it into some cabling so we could have the uh the game system wired in which was really fun because they used to have this big block on the back of it that they had lost and uh and they asked me to stay on for the winter as the the senior communications uh tech which was uh really fun i went from a snow shoveler to like second in charge on station which was great just out of high school yeah i had my 19th birthday at the south pole wow yeah that was that was pretty fun yeah and so uh yeah i went down there in november

of 2007 i think and i came back in uh december of 2008 and uh so i spent a year down there and the last plane leaves in November, and the first one doesn't come back until, or the last plane leaves in February, and the first one doesn't come back until November, and so for those nine months, you're totally cut off from the outside world, and the people who are down there are amazing. I had really incredible mentors down there, and you're locked in and so i had people who i had a co-worker who was working on the ice cube neutrino observatory and he had the vi keys tattooed on his forearm and so he was he was a linux god basically to me and i didn't know anything about Linux or anything about Unix and he

basically made me blow away the the US government required Windows install on my desktop and gave me an Ubuntu install and kind of made me fight through why every time I rebooted the graphics drivers didn't work and and I had other coworkers who were amazing network administrators. And there's very few people who will get how bizarre this is, but we had six 6509 Cisco switches at the South pole for 47 people. And I think PCI or, uh, one of the big telephone companies use three of these to route all the phone calls in the United States.

So the network infrastructure was insane. And so he gave me like the CCNA books and, uh, basically said like build a network in the shop. And so I used the little CCNA books to try to build networks or understood what I could. And there were amazing scientists who were wintering over working on the BICEP and IceCube neutrino observatories. And I spent a lot of time out there with the telescopes talking to them. And one of them gave me a uh an atmel chip and the little like big or not little

it was a big pcb that had the uh like a parallel port programmer on it and so for the entire winter i had one chip because there was no mail in or out so if i blew up that chip that was the end and i have no idea why it survived like to this day like thinking back up that chip. That was the end. And I have no idea why it survived. Like to this day, like thinking back on that chip because I was driving, I was driving, uh, like DC motors off of the pins, you know,

PWM driving. No, definitely not. Yeah. That was a, that was a terrible thing to do, especially if you only have the one chip, but it was amazing.

And it worked for, uh, for the whole winter. And, um, you know, that being said too, I had $600,000 worth of government test equipment that I was there to, I was there in the event that something went wrong. I could fix the, um, the radios and the, and the copper networks and I could patch in fiber if they needed it or I could fix broken fiber. It was kind of the purpose for having a comms tech winter over and work with the HF equipment and all that. And I was basically surrounded by a bunch of graybeards who had been working on the continent for a number of years and would help me through these problems. And so I think a lot of the reason I got to the skill set I have now is both working in the National Science Foundation lab and working in Antarctica and just being able to surround myself with incredible engineers who are doing really cool things. And I try to do that same thing with

every opportunity I have to bring on interns and things like that and get kids from high school into our lab and work with things like satellites and rockets that they might not otherwise be exposed to and just see how much fun it is to engineer. How much of a spare parts warehouse did you have down there? You can't just pop down to... Two or three Xs. You'd have to have at least two of everything.

Fries and pick up a collection of stuff. Yeah, exactly. Crossover cable. so is there like an electronic shop with just you know okay we've got a million of every resistor we need or do you just have crimpers and you can make any cable in the world cables cables are yeah but yeah um so we had you know tons of cable tons of wire to make cables i spent a lot of the winter making Teflon network cables for some of the equipment out at the NOAA facility. Because in the winter, if you think about a Cat5 cable

outside, it's not really a big deal until you get down to negative 100 degrees at South Pole and that plastic cable is now a brittle thing that as soon as the wind blows will shatter so uh we were making teflon network cables for the outside of the building which was pretty cool um but the comm shop is kind of a strange holdover so there's a ton of equipment to repair equipment that is no longer there um there's also uh some spare parts to replace uh like comms equipment and a lot of what you're kind of there for is

to do a entire unit swap and so early in the winter if something fails you you swap the unit if you're getting towards the end of the winter, you have to swap out, you have to take your broken unit and try to figure out how from the two broken units to make something work. So, NOAA lost a computer. There was a there was that series of Dell computers that had the leaky capacitors. I don't know if you guys remember. Yes. Remember those. And, yeah yeah one of them ran the noah um this is the national uh weather ocean yeah uh-huh national oceanic and atmospheric

administration so it ran their their weather balloon station at south pole and it it fried and there was not like a spare computer and there wasn't a spare setup and so we actually uh repaired the motherboard on that computer and then got it up and running again you know something that you'd normally just pitch we were like soldering on this this crazy motherboard trying to get these you know blown caps off and get some new ones in and uh so there's a lot of equipment like that to to fix uh kind of older tech that happens to, to also fix some of the new stuff. But, um, like a lot of the cabinets just have 7,400 series logic and, uh, yeah, so there's just trays and trays and trays of 7,400 series logic.

Um, and so I built like little things for people, would come in like one guy Josiah really wanted he was a heavy equipment operator and he had his headset on all the time while he's outside in the snow you know plowing around moving the mountains of snow they accumulate or that blow in but

he wanted to still be able to hear radio calls when they came in and so I wired in this I I built him this little box, a little Schmitt trigger box that if the radio signal would pass a certain level, it would cut off his music and patch in the headset from the radio. And then as soon as the call ended, it would patch back in his music in a little delay. And that was, so I spent a lot of the winter building like weird little things like that. And so I put a little sign on the door

that said grandpa's workshop, which was fun because I was the youngest person on, or actually my friend of mine went over with me. And so we were 19 years old. And so the, you know, putting the sign on the door that said grandpa's workshop was kind of a kick to some of the the people around there but um they would bring their broken toys and people would bring in their their

ipod and they would just have tears in their eyes and they'd say like this is all i have yeah this is can you imagine the only entertainment i have and you know it's there's not going to be another flight for for half a year and you don't have we don't have a constant internet connection and there's definitely no bandwidth for music or anything like that so um yeah fixing fixing broken ipods and fixing broken laptops was kind of par for the course and sometimes you just have to write it off like when someone brought me their their macbook that had they had spilled soup on and it was just through the boards and it was caked and blackened. And I was like, oh, man.

Like, I really don't think that this one's going to make it. But, yeah. Do you think having had that experience, does that drive you to go to small companies? I mean, because when you went to Planet, it was really small. And now you're at a stealth company that's you and scientists. Yeah, I love small teams. I mean, that was the feeling at the National Science Foundation labs and at South Pole and at Planet back in the day.

It was 20, 30 people in a room solving problems. And I like teams where the number of problems exceeds the number of people. I think that's kind of my threshold. That's a good way to put it. Yeah, once you have more people, more than one person working on a problem, then I think I'm not really a huge asset to that organization anymore. I mean, Planet's huge now. It's nearly 400 or 500 people globally, which is super exciting. It's great to have worked on a project that is really taking off. But I think, yeah, I like the small teams a lot more.

Yeah. How does your role differ as a company gets larger? I feel like one of the wonderful things about where I am right now is there's basically no meetings. And I feel like... And yet you still talk to each other like and yet you still talk to each other yeah you still talk to each other everyone knows what everyone else is working on um and i think the you don't have people whose responsibility it is solely to communicate between people and i feel like that's that's kind of the hallmark of a larger company is you have

you have managers who are actually acting as as just kind of like communication channels and or you are you're no longer embedded in the problems that you're trying to solve or you know you're trying to you've handed off so much of the project that no one has a clear understanding of the the overall picture anymore or everyone doesn't have an overall picture of the project anymore and i think at a small company everyone's required to know pretty much everything about what you're trying to build. And once you're at a large company, it's very difficult to have that large view because the depth in each of those has become so great. I sometimes get bored because at a larger company, they just want me to write firmware or just want me to be a manager or a mentor. But at a smaller company, they need everything.

They need my marketing opinion. They need me to say if this is a good app and I can use it as a user. They need me QA and firmware and management and product spec. And I just like the variety. Yeah, absolutely. Yeah, I'm the same way. One of the fun things about being at Planet back in the day was we had to build satellites. We had to build boards.

We had to test them. And as you get bigger, you have test engineers and you have production teams and you know right now at the at the the stealth uh rocket company um like we all dropped everything to go and build a new test stand which was great like wiring up a giant test stand and then we had uh everyone had to jump on lab view and write all the lab view and that was that was an incredible pain an incredibly painful process i'd never been exposed to lab view before. And exposed to lab view is a very good way to put it. Yeah. That was,

that was, that was not, not the, I was just trying to find a way to add a code block. It's like, please let me drop a block that is just an equation in here. And it's like, nope,

you have to drag divide. Then you have to drag constant. Then you have to drag multiply multiply and then you have to wire your multiply to your divide i was like this is uh well i mean can i just type at it i don't want to use my mouse yeah i know i am like clearly a very bare metal person i want to be i want to be at the silicon level as close as possible and yeah these you know i i when i when i checked this project into into the repository i was astonished at the size of it i i couldn't wrap my head around how this simple file was so big i i just, it's, it's staggered my mind, but yeah, the, uh, yeah, I love that, that being dragged into, into hard problems that you need an entire team to jump on is, is really fun. And, you know, it's the late nights and it's the, it's the camaraderie that

comes with a, uh, you know, working on a, a very difficult challenge. And also I'm surrounded by people who I have immense respect for and I've never had the opportunity to talk to people about rocket engine design and about rocket nozzle design and about the injectors into the engine and why that's an entire field of study. And it's great. Whenever someone has a a graph on their screen i was i was terrible about this at planet two as i just walk up behind them and i'm like what are you looking at show me the data what do you see tell me what you're reading but

yeah that's uh that's definitely where my heart is is in in small teams building things quickly, I think is important. I 100% understand. And that's the fun part for me. But I should be letting you get back to work or actually this weekend, there's a lot of things going on. Yeah. You're at Maker Faire. Is there anything cool that you're looking forward to or i guess

it's just started so you haven't much seen anything probably yeah i haven't seen anything really yet i am uh i have plans to do just kind of go and check it out uh me and uh my partner kelly are going to be down at uh there's a um there's a repair cafe going on in san jose which i am stoked for so we're gonna go and fix some broken toys and uh try to upgrade some things that's that's kind of going to be our station we're going to try to add like blinking lights and craziness to things um and yeah that's going to be that's going to be really fun but uh no we usually just have a great time wandering around maker fair and running into everybody everybody's there you see you see all the all the happy faces of the of the maker community and it's it's great yeah um so i haven't i haven't locked down a schedule yet

but i do have a ticket i understand that so the repair cafe this cafe, this is a thing. Yeah. Actually, when Kelly and I had left Planet, we were talking about just starting repair cafes and traveling around and fixing broken toys. We bought a busted espresso machine on eBay and we stripped it down and threw it all back together. It was hundreds of pieces. It was such a fun project to get this broken thing on eBay and fix it. That's one of my hobby projects is I buy broken toys on eBay and I try to fix them. Like Fred. Fred really kicked off that hobby. um yeah we got like a vibratory parts feeder the other day which is this fascinating um uh like mass assembly line tool uh which was really cool and we put like dog treats in it and the the puppies were freaking out as these like dog treats were flowing up this

bowl and out onto the onto the side and uh yeah and so that's that's been really fun but yeah the repair cafe is i guess people bring their um their broken things we've signed up for um like small mechanical i think uh like high-end electrical that's not computers explicitly and um and maybe like appliances i think is our is what we volunteered to work on. But yeah, it's the first one I've ever been to. Kelly found it, and I'm pretty excited to go be a part of it. Searching online, there are all different kinds of groups doing repair cafes and repair clinics, and they just seem to spring up.

And yeah, you can volunteer to do stuff or you can ask to take stuff it's pretty cool yeah and on top of that this is this is one of the coolest things i found out uh this morning was that we get an apprentice or we get multiple apprentices so every every fixer gets an apprentice which i guess is someone who's like i want to know how to fix stuff and so you you can have someone shadow you as you take something apart and try to figure out how it works. Because a lot of the time, it's just a wire that's come loose inside. You just plug it back in, and everything comes back online.

And then that's even more exciting, because you can dive deeper. But it's so rewarding and magical to hand it back to them after three seconds. Yeah, and just they're so happy. And then hand it back to them after three seconds. Yeah. And just the, they're so happy. And then they know how to fix it next time. And, and I don't know, that's just really, that's really cool. I, yeah, I love fixing broken things. That is a good note to start closing the show on.

You should let him get out of his car. It is your car. It's so hot in here. It's so hot. No, it's actually, it's out of his car. It is your car. It's so hot in here. It's so hot. No, it's actually, it's my roommate's car. It's even worse. We had a long discussion earlier in the week about going to Maker Faire and recording in cars that were not ours. What might, you know, Honda, green Hondas in the parking lot.

Not only are your lights on, but you have podcasters in the car. Then they're actively reporting. Oh, man. Do you have any thoughts you'd like to leave us with? I think you should take things apart and put them back together with a new feature. And I think that that's just a really great thing that more people should do. You should push a button on your blender and the refrigerator door should pop open. All right.

Our guest has been Sean Meehan, avionics lead at Stealth Space Company. Something about rockets. If you are interested in joining him to work there on rockets, space, mining asteroids. We're making that stuff up. We don't know. Explosions. We have no idea. But if you're interested, where can they send email, Sean?

You can send email to space at logiclow.com. L-O-G-I-C-L-O-W.com. That's it. Yeah, that's the website of the robots. And if you don't recall it and can't get to this point in the show, you are welcome to send me email. How would they know?

Hit the contact link on embedded.fm. Well, because I say that part often. And I will make sure to put you in contact with Sean. And thank you so much for being with us, Sean. I have not met Fred or Lefty, and I really, really want to. Oh, yeah, definitely soon. Yeah, we're going to host a robot warming party soon, once they're up and running. Well, thanks so much for having me on the show.

It was really fun. Thank you. And thank you to Christopher for producing and co-hosting. And of course, you listeners, thank you for listening and for completing the survey. I do have a thought to leave you with. This one from Mark Twain. And yes, I did verify it really was from Mark Twain.

Keep away from people who try to belittle your ambitions. Small people always do that. But the really great ones make you feel that you too can become great. When you are seeking to bring big plans to fruition, it is important with whom you regularly associate. Hang out with friends who are like-minded, who are also designing purpose-filled lives. Similarly, be that kind of friend for your friends. Embedded is an independently produced radio show that focuses on the many aspects of engineering. It is a production of Logical Elegance, an embedded software consulting company in California. If there are advertisements in the show, we did not put them there and do not receive money from them. At this time, our sponsors are Logical Elegance and listeners like you.