MacroFab Engineering Podcast #149
MacroFab Engineering Podcast #149
On this Episode, Parker and Stephen have guests Aditya Bansal and Mijael Damian of Kinetic.
Visit our Public Slack Channel and join the conversation in between episodes!
Jeff Doong and Matt Glazer of KINETIC. In an ideal world Matt will send Jeff to terraform Mars.
The “GORE VENT” that enables pressure equalization between the outside and inside of the enclosure. Keeps the relative waterproof rating of the enclosure intact.
KINETIC’s Lego style assembly instructions. If only more product designs had these!
Parker is an Electrical Engineer with backgrounds in Embedded System Design and Digital Signal Processing. He got his start in 2005 by hacking Nintendo consoles into portable gaming units. The following year he designed and produced an Atari 2600 video mod to allow the Atari to display a crisp, RF fuzz free picture on newer TVs. Over a thousand Atari video mods where produced by Parker from 2006 to 2011 and the mod is still made by other enthusiasts in the Atari community.
In 2006, Parker enrolled at The University of Texas at Austin as a Petroleum Engineer. After realizing electronics was his passion he switched majors in 2007 to Electrical and Computer Engineering. Following his previous background in making the Atari 2600 video mod, Parker decided to take more board layout classes and circuit design classes. Other areas of study include robotics, microcontroller theory and design, FPGA development with VHDL and Verilog, and image and signal processing with DSPs. In 2010, Parker won a Ti sponsored Launchpad programming and design contest that was held by the IEEE CS chapter at the University. Parker graduated with a BS in Electrical and Computer Engineering in the Spring of 2012.
In the Summer of 2012, Parker was hired on as an Electrical Engineer at Dynamic Perception to design and prototype new electronic products. Here, Parker learned about full product development cycles and honed his board layout skills. Seeing the difficulties in managing operations and FCC/CE compliance testing, Parker thought there had to be a better way for small electronic companies to get their product out in customer's hands.
Parker also runs the blog, longhornengineer.com, where he posts his personal projects, technical guides, and appnotes about board layout design and components.
Stephen Kraig began his electronics career by building musical oriented circuits in 2003. Stephen is an avid guitar player and, in his down time, manufactures audio electronics including guitar amplifiers, pedals, and pro audio gear. Stephen graduated with a BS in Electrical Engineering from Texas A&M University.
Special thanks to whixr over at Tymkrs for the intro and outro!
Welcome to the macro FEHB engineering podcast. We're your guests Matt, Matt Glazer,
and Jeff tau.
And we are your hosts Parker, Dolman.
And Steven Craig.
This is episode 149.
Matt Glazer is the lead electrical engineer at kinetic, building IoT wearables for the industrial workforce by day and building electric long boards by night. He's obsessed with space and can talk for hours on end about the current state of space technology. Jeff Dunn is the head of manufacturing at kinetic in charge of making sure current products are manufactured correctly, and new products are designed to efficiently scale. If he wasn't doing this, he'd be a farmer back home in California.
I got a little question.
Yeah, today's or
both. So Jeff, you said you'd go back home to California for for farming, but would you go farming on Mars?
Oh, good question. Yes. Although I would probably send someone like Matt out first. I had this year that I would just end up suffocating to death in some way like, asteroid you know, puncturing some seal. And don't get me wrong. I like farming, but not not that much. I wouldn't I wouldn't die for it. So I
owe you an asteroid resistant Greenhouse on Mars. Yes, but
then I think I would also get trapped there. So there would have to be a way back. And not just like Elon Musk sending more people to Mars to keep me company. It's like, I need a way to get back home. You know, I already complain about weather in New York. Planning, but weather Mars would be a whole different whole different aspect.
I had to be home by Christmas. Right?
Exactly. Right. So So you guys are basically talking about recreating the Martian in real life where Matt sends Jeff to Mars and then Jeff plants potatoes there, right.
I think he planted potatoes out of something that like, like human excrement.
He was a little desperate.
I would assume so. Can you even get all the nutrients you need from that soon?
Just just on record, Jeff had a bet with me to see how quickly I could talk about space. And he beat me, which is impressive.
Yeah. Well, well, Parker asked.
I think you saw me like the the $5, though. So that was
it. So um, this is not the first time we've had kinetic on. Episode 53. Before we hit three digits, I feel like ages ago, we had a deja and Mihail on and what are their positions at the company?
A DTS is the CTO of the company. And Mihail is currently the director of analytics.
Oh, that that's right. That's right. And that was on in February 2017. So it's almost been two years. So for our listeners that have not listened to that podcast yet, what is kinetic?
So kinetic, one of the like you said, we're a company building IoT wearables for the industrial workforce. And kind of what we're aiming for is a the device we build the reflex is used to reduce back injuries. So any form of warehouse, you picking up a heavy box, you're moving it, you're probably not moving at correctly. And from there we have, we kind of do some behavioral training kind of thing to teach workers how to lift properly, thus reducing back injuries, thus saving companies, extremely large sums of money on workers comp, and so on and so forth.
So it has a name now, because was the treadmill while we're on the podcast, it was it was just called the device, the kinetic
device or the wearable unit. Now, it's actually the reflex
reflex.
Where did that name come from?
A couple of weeks of continuously, generating as many name ideas as we possibly could. And that was the one that we all settled on, that we liked the most.
And how does the reflex do like modified or not modified, but how does it influence people's behavior? Does it like pop out little treats?
I kind of wish it would do that? Rolls. So what it does is the device sits on your hip, either left or right side. And when you do an improper lift, when you return back to a standing position, it'll buzz and it's strong enough buzz that you'll feel it but not enough to be over the top. And it also it's got a display on the device. So you can see I see I call the device. It'll show step in it has step counter it'll show how many high risk postures which we call what we call the bad lifts. per day, and then it's got a clock. So let's see how many you do per day.
So the device clips to your your belt correct. So it does it does it generally work off of like dispensing angles or
so? Yes, yeah, we have an IMU and a altimeter. And between those two, we have some and then on top of that, we have a bunch of machine learning algorithms going, that can infer your back angle, the angle of your back. And based on that, we have a point where we consider a bad Iris bad left high risk posture. And then once you come back up, it'll let you know that you've done a high risk posture. And then obviously, you don't want this thing to buzz at you. You prefer not to have it buzz at you. So the wearer starts being like, Hey, maybe I can, I'll squat instead of bending my back or like, maybe I can lift that. Lift my table up a couple inches higher, so I don't have to bend over as much to grab boxes.
Cool. The I think also, it basically reminds me of a the devices like a pager device. You hear that a lot. Yeah. Yeah. And so we'll put some pictures on on our blog in and go check out episode 53 know more about the device, or the reflex. Well, now, so let's expand a bit more what y'all do at kinetic. So Matt? So the lead electrical engineer, what's, what's that entail?
So I do pretty much all of the hardware development at the company. So if it has to exist in reality, and it's electrical, I'm the one working on it. So boards, PCBs, initial bring up of practically everything, new designs. Yeah, pretty much everything. And then I also kind of, I handle a lot of logistics within the company itself. So like getting things out the door, getting things in the door, ordering toilet paper. Yeah. And office manager seems to be one of my roles. So
next, so it's important to stay on his
Snackmaster. Cover pecans are a special treat at our company.
It's called everyone just getting away.
About how about you? Yeah.
Yeah, so I run the manufacturing side of it. So everything sort of I sort of imagined earlier on. So everything from, you know, making sure that units get built, designing the mechanical side that goes with what Matt does. And then making sure that for future revisions that things are designed so that they're easy to assemble, so that as we scale, then, you know, these issues that we see with like quality control don't sort of appear over and over again. And a lot of that has to do with managing vendors, managing suppliers manage making sure that you know, we're able to hit the timelines that we are. So when sales comes and says, Hey, you know, Jeff, we need, you know, 500 units for deployment, that we're able to go and say, hey, yeah, we can have those either in stock or in X amount of days.
So since you do the you do the physical mechanical design, then have like the enclosure, is that right then? Right? Okay.
So the mechanical enclosure, we actually had someone else design it. But moving forward, it would be something that, though I wouldn't be designing it, I would have input in such a
gotcha. So it's not like, it's not like you and Matt, like argue like, I need that mounting hole moved over point one millimeters, and mats like that runs into our 13.
Not yet, soon, soon.
The mechanical design of the device is actually very elegant, and pretty complex, actually.
It is, I mean, to be honest, I think a lot of it can be slimmed down, a lot of it was in the very beginning, when you have sort of your minimum viable product and you run with it and sort of it takes on a life of its own. Whereas if you had more time, you could then go back and and put a little bit more thought into it. But if it works, then then you know, and there's demand for it, then then you were
really happy with how it looks. It's just Yeah. Be nice if we could assemble it faster.
You know, back in my days at the FAB, I remember going through your product and looking at the bill of materials and there was one particular item that I always laughed at. There's a thing called a gore vent that goes on the side of it and we always joked around like, what needs a gore vent. You know?
What's in Gore vents? I actually don't know about this part. It's like a rubber grommet. Right?
It's a it's an air permeable seal. So we have a semi waterproof device that we the pressure sensor obviously needs to be able to get some access to the outside world. I think The initial samples we got were literally core brand, which is why we always call it in core vent. So thankfully, it's just air that we need to pass through the vent. Nothing else.
giblets everywhere. So actually, the pressure sensor is, Can y'all tell us the like, is it sensitive enough to know if it's like at standing height versus like on the floor? Which is like what four feet a difference?
It's more when it's more an absolute value. So we can tell if you move. Gotcha. Okay, knowing the height of a person would be helped, like, it can definitely be helpful. But in our case, we mainly just need to see how much you move. We don't even know like, Oh, you're four feet off the ground? We can kind of
that's all I'm saying. Is it precise enough to know, that amount of difference? That's actually impressive from that whoever manufactures that part?
It's conveniently the same sensors and pixel.
So the so the pixel that I have right here has got one of those. If it's the pixel one, yeah.
So Parker, so what that saying is that we could take your phone, take it apart one day when you're not looking and I just pull that one part
out? Exactly. Now, it'd be like, our manufacturing floor would just say, hey, we need one more of these parts. And I'm like,
no, no, no, no, we will volunteer your phone for you. Yeah.
For overage a part of
the shipping box of pixels we found on the side of the road, so you can tear them apart and shove the parts of your product.
Alright, so since since February 2017, what has been the major engineering challenges? Have y'all been scaling up? This product? Jumping right into it? Oh, yeah, I went from like, super goofy to just like, straight up like the hard hitting.
That's a gut punch right there. Yeah.
That's what Matt's been working on for the last, what, seven months?
Seven or eight months now. So we you see us an Intel Edison as the processor as our some in our device, and that conveniently, totally conveniently went ELL in like about three months, which is like a third of the time that you normally get for an ELL part.
Yeah, usually usually accompany we like, Yeah, we're gonna get rid of these in about two years. And until we talked about this on the podcast until was just like, yeah, we're just not going to make these anymore. What's on the market is left was what's there. Oh, they
dropped the floor.
Yeah, we, we had the opportunity, we literally got an email saying order what you can the fam shuts down in December. That's it. And as a startup, we, I mean, we can't just go and be like, Alright, I want 50,000 of these things. So we never have to worry about it. Again, we bought what we could and then started the mad dash to find a different processor to put in our device. And we ended up finding a module. I think it's a Digi module, Digi part that works well. But if you don't expect to have to do a sidestep, just to make your product work. And it's an interesting. The Add on the electrical side of things we went, I had to completely redesign the board. There's always the IFS of like, there's always the weird quirks of bringing up a whole new thing. Took a couple of reasonable revisions there. But also, it's the switch from x86 to arm and all the software requirements that go in there.
Now, I can't imagine you basically you probably had to rewrite a lot of the software. Yeah,
one of the big things that we had an issue with is the Edison the Edison for what it was, is actually pretty great. It's a dual core processor, with, I believe four gigs of memory and a giga RAM, which, when we were shopping around for parts that fit in the same size ish, was impossible to find. We're not running that much memory and RAM anymore, which I mean, in the long run is great for us, because we can improve how we run everything, but it was not expected. So it was it was fun design challenge.
Yeah, I mean, it's fun, but then you also realizing that hey, you're spending six to seven months of you know full time work just trying to get your device not not improved, but to behave the same. So you know, this is time and money that you could spend elsewhere to getting the product better, but inside your head just just to exist as for the company, for the company to exist, you have to spend this giant chunk of time and money on so by then sort of them the mechanical side is already made and so We do have these size constraints. So it's not like you can swap out any old part just to just for it. So
you already got your molds made for your enclosure and stuff like that. So you have to fit, basically redesign the core of your product and six months and make it fit the existing enclosure. Yeah,
I definitely, I'm sure Jeff can attest to this. I am a broken record. And we're like, can we make a mold change? Can we make a mold change? I want a little bit more room, and some more room.
That's what we were talking about earlier. So it does happen. It does happen. The opposite way, I need to move this mounting hole point one millimeters over. So I can fit our 13
Yeah, there's a there's a there's a row of components at the very top of our board that I had to go from. I think I tried. So like everything's pretty much Oh, 402. And I'm just like, I can't fit this. So I have a whole row of components as Oh, 201 just because I couldn't fit anything else. That's great. That's fun.
So So did you guys, you did pick up some Edison's to hold you guys over? Have you exhausted those and you're on to the new device yet?
Yep, the the Edison's are all gone. And they actually were Maghreb is building the new version now. So.
So you guys have rolled over? What? What was your plan for doing that? And how did it how did it all come out?
So a lot of it, at least on the electrical portion. I worked heavily with a DJ, to a DJ was doing most of the research into the modules themselves. Because he had a bit of a better connections with that, that realm. And once we kind of settled on a couple of them, we were talking about, like, maybe the manufacturer could make their board make a board change to fit all of our parts, or like, we'll have some design help from them, or like, either, we'll just implement it ourselves. What ended up working out was just, we can buy the module, we do our carrier board and then go from there. We ended up finding the pretty much the only part that would really fit on our board was the one we settled on. And then it was just half mad dash two. We have a little bit of time here in there to just kind of churn through revisions. The you can tell the revisions that I have from the amount of bodge wires that, like just spawn off the board. It's pretty great. I mean, like, Oh, hey, that boot mode doesn't work. Let's switch it. So grab the exacto cut the traces and run a bunch of wires everywhere.
Yeah, I mean, I think Stephen to answer your earlier question more. It's also from sort of like a supply chain side, it's an ideal world, you build as many as you can to, to have a buffer, and then you have that buffer stock while you then switch over to start your Digi run. So that you have some time in between to work out any of the kinks. For us, though, it ended up being build as many as you have on stock that you were able to buy earlier. And then just switch over your manufacturing your next run of our dejes and work out the kinks as you go along. So it wasn't ideal. But I mean, when you're forced to do it to do so then you you you make it work.
So, I mean, you Matt said that this was like this Digi part was like the only part that you could replace the Edison with? Was there any other considerations for this part? Like, are they guaranteed, like years down the road for the part now?
So I think one of the things, it's an NXP processor it said i MX 60? Well, I believe and I have I think those are guaranteed for five or six years, the processors themselves. But again, it was just kind of, if we don't have things, we don't have a company and we need to get something out the door. Yeah. It's like pivoting your company to do the exact same thing.
Well, in that were you able to potentially enhance or fix things that you weren't really thinking of doing with the Edison.
So what were you able to actually we were able to, I've we were able to add some extra connectivity to device in the sense that we have a couple we have some pins on the back. So now I have a couple more interfaces, they're mainly USB and then we overloaded just beef up our battery management a bit so we have we had we had a decent battery management to begin with. But now I've got an add in an actual proper BMS chip and some other stuff.
And so so this is this actually happened during Urals process of scaling up. So what kind of design changes that y'all make to the board to make that easier.
The board
With this, so one of the things is that the process, the module that we got is surface mount versus the Edison module, which kind of I mean, it connects on our connector, and then we had them bolted down. So we actually, I think it was in a roundup or two ago, those pin Pentek standoffs, we actually use those for the Edison. And then we'd screw the board down just that way, it never came off the carrier board we have. So that kind of simplified the part we inadvertently simplify the process of assembling these boards. Because that way, you can do the entire SMT assembly just on the SMT line, instead of having to build the board, bringing bringing the Edison module, plug it in and do mechanical assembly as well. Some of the things, we learned some of the other things like the placement of the battery leads and the motor leads and so on, we were able to change based on feedback from the assembly line as we were building these building Edison units to the Digi units. But other than that, not a huge amount change just because we are extremely space constrained. And we did like most most of what we do is within the processor itself. So we didn't have to add a huge amount. Otherwise.
Yeah, I think from a mechanical side, I think it might have even been a step backwards, just because the antenna had to be a separate part. And then that had to be added in. So I mean, it would have been great to be to say that we had the chance to redesign the mechanical portion to make assembly easier, but really to happen, at least for for that side of things.
Like in terms of changing the wearable to scale. That's something we're working on now. Because we've we've pretty much gotten over the hurdle of Okay, now we have a working hardware. Now it's and Jeff can talk more to this is just how do we a bring it from hundreds of units to 10s of 1000s? Plus, eventually? And how can we reduce the amount of time it takes to assemble? Do we need these complicated parts? What What can we do to simplify this process kind of thing?
Yeah. Can you expand on that? Jeff? Yes, sir.
So a lot of it is. So when the part was sort of when the wearable was first designed, it was I mentioned this earlier, it was sort of like a minimum viable product, it was meant to be more of a proof of concept. And you know, use it were very nice proof of concept. And so people didn't bought it. And, you know, once the ball started getting rolling over to the ball started rolling, it's hard to pull it back. So there are a lot of things that we learned along the way, in terms of, you know, how these things should be assembled, are the major pain points that we're now looking into, for our next revision to sort of take these into account and say, how do they how do you build a product now that you know you're selling, not just hundreds, but you know, 1000s, or 10s? of 1000s? And so, with those things in mind, you know, like, Are you paying attention to, you know, assembly, are you paying attention to certain things that you you now have more time to think about? It, if that makes, if that makes sense, it
makes total sense. I mean, you spent the last, you know, eight to 10 months, basically redoing what the company did the previous couple of years, just to catch back up to getting around a part basically falling off the market. And
just feel like feel like as if you had a sort of where you you're going from, you know, putting out one fire to and another and so there's always something that's burning in your element is sort of managing managing priorities and saying, hey, you know, all of these things, as it relates to scaling up, you know, of these things, which things are the most important, which things are the lowest hanging fruit? And if you, you know, only have a set amount of time before you have to, you know, deploy your second generation, you know, which things is it most worthwhile to focus your attention on?
Yeah, I that's, you know, we have had the same thing out at macro fab as a startup is you have efforts to do X, and then you also have the impact of x. And so you want to maximize impact by lowering and also lowering the amount of effort to do x. And if that's the right balance, then you do X
right, but anyway, hit there's also so many features that man, I would love to add on to the wearable, but at the end of the day, it has to still make business sense. Like from an engineering perspective, it'd be so great if we could add you know, XYZ functionality, but in the end if the customer doesn't need it, or it needs working space, right? Yeah. Then you have to put it aside and you know, man, I can just build our dream wearable. As like a fun hobby. But for this, it has to it has to make sense. Yeah.
It's one of those moonshot interview questions just to kind of see how people think. And it's quite literally like, what would you do to make this work in space? Is the gist of that question. You know,
let's actually go, let's go down that route. How would you make the device or read the reflex? Work in space?
I would just build it into a suit.
See, I'm not sure. During my interview, so no, I haven't I don't have anything,
I think is really cool. Like the like, I guess it depends also what you define as space. Is it like, it's, it's definitely zero gravity when people just leave space, but it's it's in the ISS? Or is it actually exposed? The elements in the vacuum? Yeah, you
on a space station? Are you outside? Are you on a planet? Are you on an asteroid? There's so many different options. One of the things with that question that was it was mainly, like, how would you communicate with many devices kind of thing? So it was kind of like, what would you use to communicate? And it's like, trying to see like, what do you use? Stuff like low raw or what what can you apply to actually think about? I don't even know, like, I'm just trying to think of how I would truly answer that question.
Also, how long do you want it to last? I mean, I'm sure we can last for a few seconds. But yeah.
Oh, LED display boils out. Yeah. How?
Does it off the shelf? lithium ion battery handle? Negative 300 degrees Fahrenheit? Do I want it? for that?
I was about saying you gotta get
back and let's just dump the entire thing and silicone. See if that'll work. That'll probably off gas to I think silicone off gases pretty heavily.
Yeah. Even regular fr for out gases. So you'd have to use a special space rated fr for probably.
So how much? Are you willing to pay for this?
Starting at the low low price of $100,000? In advice, plus internet fees?
And then not not to be honest, I'm not even sure astronauts need it. I mean, I mean, are you are you worried about high risk Foster's when you're out in space, you probably have other other more pressing things to worry about, then, then you're back.
We'd also like tell our operator robots, to the robots get bad lifts?
No, no. Great.
So we're gonna we're gonna stay in warehouses. And we'll see about the reflex face addition, sometime in the next couple of years.
One day, one day, when when Elon Musk is building his Mars colony, and those workers need it, then we will we'll have that discussion.
What I mean, they're gonna have space warehouses. So
yeah. So when that happens.
So with the lead electrical engineer and the head of manufacturing at kinetic, I'm sure you guys both have a bit of sway on new designs, right? So you can put all these in, right? All these requests of they'll probably get shot down, down. But you could put
imagining or what happened if we presented the idea of the space wearable?
Like, from a complete outsider's perspective, other than the fact that I've spent maybe like an hour or two with the DTN Mihailo. They have a sense of humor, but I don't know if they have that much of a sense of humor.
They're used to me talking about space nonstop. So they'd be like, okay, Matt, that sounds like a good idea. You know, go back and continue working on the charging dock.
Think back on and say and so
if Jeff says that, yeah, they would definitely think you've gotten saying,
or, or that I've spent too much time with that closed environment. And
actually, you know what that brings? That brings up another question that I had earlier. How, on a day to day basis, how often do you two work together? And what what do you guys do together?
I usually turn my chair and go, Hey, Jeff, I've got a question about every 10 minutes.
Right now. So I mean, I think of the people that work together in office, I think Matt and I work work pretty close together. Just because there's so much overlap between mechanical and electrical, there's so much overlap between sort of the supply chain stuff and sort of, because me even when I'm ordering parts for, you know, the next run, I mean, a lot of it's still like half the bomb, it's still all electrical. So a lot of it needs mats input and a lot of times when we and this is one of the issues that we wanted to add Walmart scaling as well where you know, you'll buy out, you know, all the parts. I would like to do GQ and so you know, what, what alternates are there and Are the V times for those alternates? And? And hey, you know if you have no part it's like, what's the plan? Yeah.
Is this 0.1? microfarad capacitor. Okay, the sub in.
Right, right. And then I can't answer that question on my own. So it's a lot of it is me as Hey, Matt, hey, can we can we do this or by him saying, Hey, we have Mr. Barnes. Yeah,
that happens over the summer. That was, there was a point where I had found an alternate went to go buy it wanted to check with a co worker, just to make sure like, we were all set for ordering things. Got to my desk, and it was gone.
That was passives. Yes,
I am pretty sure I have bought out Digi key and Mauser combined on at least four to five parts. So if something disappeared to zero, that was me. Oops.
Yeah, I mean, I think I think even for mechanical stuff, we've had suppliers then reach out and they and then being like, hey, we noticed you put in a really large order do you want us to, you know, prepare ahead of time in the future, you know, so it does happen as you as you start ordering more and more parts. So it's nice to start developing those sort of vendor relationships as well, then you can negotiate more prices you you for that sort of like more personal relationship, you can then get more sway and sort of like when they do have very long lead times, you can can then go in and sort of negotiate with them on either partial shipments or
pay a little more and get yours put ahead in line stuff like that.
Right. Right. You bribe them with cookies or whatnot. Chocolate covered
pecans. Yeah, exactly. I think you said Pete cans earlier.
Yeah. Did beacons Is that is that not how you say it?
They say P cans up here in Colorado.
But technically we are. You're accepting Matt's rightfully pronouncing
it. Yeah. One of the other things like on the in regards, like buying parts and everything is one thing we've been learning is just managing lead times. It's interesting going from oh hey, I can just pick up 10,000 of these and educate to now I need 150,000 I should probably make sure I've got 10 weeks at minimum. So that's definitely something you have to watch out for. As you go from hundreds of prototypes or even building by hand to Okay, now I just put in this giant multi 1000 In order and we need a million components.
Yeah, it's one thing when it's like a passive part or Jelly Bean like transistor where you can find substitutes anywhere but when you need this one specific part that's where not self cups comes into play. I remember um, or building boards for pinball machine and I we need one specific microcontroller and the lead time was like 12 weeks and just like no yeah 12 weeks later have to wait.
Yeah, I mean, because it also affects sales right when when the sales guys come and say hey, we have this really large order and that takes out you know, half of your inventory you know, then you need to replace that and so you can't and then so they come back later you know week later so hey, we have this a second large order you came to like okay, wait 20 weeks.
We our customers are amazing, but sometimes they have immediate needs. They want their they want their things quickly. So they want it now they weren't
yesterday. That has happened. I amazing idea. Amazon yesterday prime. Exactly. box shows up the day before you order it.
Jeff Bezos is already on it.
Probably. He's using AI predictive technology to figure out what you want to buy. So is that going to happen before the device goes to space? Probably.
Amazon will already have a hub on Mars for your farming, you can just you can get your stuff right there.
Amazon will ship Amazon's new shipping service will ship our device our awareables to Mars so we can sell them on Mars.
Can you do that prime to date, Maurice? I think that's available now.
I would say with enough energy you can do anything
enough fusion reactors and you can get yourself there in a couple days.
So Jeff, you're in charge of the supply chain, is that correct? Right. Right. So it sounds like you were quite a quite a few hats because you got your hand in talking about design work, especially the mechanical side. You're also in charge of manufacturing but but supply chain at the same time. How do you handle all that?
Lots of Excel sheet It's, there's a lot. But at the end of the day, because manufacturing and supply chain is so closely intertwined, it's, it almost seems like a natural extension for that part and, and supply chain, it's a mean that in an ideal world, once you locked down, at least, at least on the mechanical side, it's relatively straightforward, you know, your lead times all your parts have custom anyways. And so it's just a matter of placing them in and the longest lead time components are electrical. And so if you know if Matt's like, hey, I need 20 weeks for this lead time, then, you know, 20 weeks, I have plenty of time to get all the mechanical side in order. And so it's just managing those shipments and making sure that everything arrives in time. And I guess one of the other issues that we've run into as we scale up is managing space in the office. It's, it's, it's one thing to get, you know, five boxes of, you know, six inch by six inch cubes, you know, and they will fly tiny components each now that we're, you know, taking in 1000s of units, and you have 1000s of components, where are you storing them? Right? And so do you have enough space in your office to store everything and so, so our hardware area in the office is quite, quite a mess,
the criminal about 15 by 20, it's great.
Yes, a lot of cardboard boxes, you have them stacked all around your desk. And so it's, it's, it's something that you should be aware of. And that's one of the hard things of being in New York City. I mean, it's nice, but you're also even more space constrained. And it's not like you can receive palletize shipments easily and ship them out again.
I remember before macro fab. I was working with Chris church at at a he's our, he's our chief product officer working with him at a company called dynamic perception. And we would get shipments in like in bulk like that. And you'd get all the aluminum and basically the entire manufacturing floor will be full of aluminum extrusion for basically the next four months. And you couldn't do anything else besides like walk around aluminum. But that wasn't even the worst. The worst was when we would get the packaging in. And it was just the foam inserts. And we would fill everyone's offices with foam inserts pallets of foam inserts. And so I would come in one day and be like, Okay, now I can basically wiggle through my office and get to my desk. Because there's just phones everywhere.
And then you walk out and you have no phone pebbles just like stuck to your shirt everywhere.
Yeah, I regularly established for kinetic which consists of inbound materials, outbound inventory, and whatever prototypes I'm working on they go on top.
So it's a little cardboard cardboard box for so like
child monitors lighting it like the the torches that you put in Minecraft kind of thing.
Your own little cave. Yep.
The best part is the office area that we're in. It's very well lit except the corner we're in because we're under a bunch of H fac equipment. So they tried they did a good job, but it's definitely a little bit darker than the rest of the
and so it's like the stereotype of like having the the engineers in the dark corner. Facts I've learned from all the normal people.
Yeah, that was that's actually the thing yet at macro fab the developers like it to be really dark. And so I was come in and flip the lights on and wake up. I calm the trolls. Gotta wake the drawers up. That's the office.
Yeah, I think ours might even be worse because I think our software team is in a whole different location up in Saratoga Springs. So we've we've managed them there, which we've managed.
Winter's coming.
The best part? Like they're always complaining about, it's snowing, and we're like, oh, it's still like 60 degrees in New York City and sunny. There's no snow on the ground. It's great.
It's that much warmer in New York compared to Saratoga.
And see 60 degrees is a little high right now. But like the weather warm it is.
It's cold. I mean, it's cold because they're all there. And we have subways heating our ground, I guess.
Yeah, we just have 20 million people making it 10 degrees warmer.
They Yeah, all that little space heaters. Yeah. Imagine
all that dead skin flakes. body heat.
Well, let me bring it back. So what's what's new, are you guys working on I guess, Rev. 2.0. Because you I guess what you just did was 1.5 in a way right or or 1.0 plus. So So is 2.0 in the mix right now.
While we are talking Well, our kind of primary focus is doing the Value Engineering thing. There's no reason we can't make things look better, or like, Oh, hey, it'd be really cool to do that. Because, I mean, one of the things is, right now our dock is made of plastic, it's really hard to assemble, we've learned, we actually had a great example of that we had a customer, one of our bigger ones be like, hey, we need something special kind of thing. And we came up with is one of the on top of this is a, we build a 10 unit dock normally, and we decided to build a 50 unit dock, it looks a little bit better, it's a lot less cabling, and so on. And while we're going through this, assembling this, because this is the equivalent of 100 docks that we had to rapidly put together, I'm pretty sure we found every possible pain point that we could. And one of the things is like, the value engineering we're doing is more towards like getting rid of a lot of those. And in the process, we can make things look better, like making it look better. And part of like the new product idea is actually making it way simpler to assemble.
Yeah, so So just imagine, you know, Matt, and myself, sitting in a room building these giant dogs over and over again. And yeah, I think I think a lot of the learnings come from, you know, when you actually build these units, by hand by yourself, and then you, you are sort of intimately aware of all the, you know, places where you can do better next time. And so, like what Matt was saying, a lot of it is improving it so that it is better designed for assembly. And a lot of the cost savings then come from it being easier to assemble.
Really interesting. Like when we started working with Makerfaire, we had actually built 200 ish, a little bit more than that devices in house, like that was one of the first things I did at the company. And that's how we built our work instructions. That's how we, we didn't just show up with a pile of parts, we were actually able to present something and be like, Hey, this is how we built it. What do you think of that, and go from there. I mean, it's changed dramatically since we two people building it versus a couple more people and a much larger scales changes things, obviously. But I think that was incredibly valuable for both sides of the parties.
And and getting more people will look at it and be like, oh, you know, might be an easier way to do this one step, or put this step at a different spot, because it makes more sense.
Also, I think it's a I think it's important to note that I don't think I've ever seen work instructions from a customer as beautifully done as kinetic has done them. They're akin to a very well put together Lego set, they actually have like a legend where it's like you need 1x of these parts and 2x of these parts and every image is 3d rendered on how to do that particular step. It's incredible.
I need to send you the new ones, because they've gotten even better, I feel oh, have they really? Yes. So the beautiful thing about those instructions, is our designer is also he also builds his own 3d printers. So he knows both ends of the spectrum. So that's how you get those wonderfully made instructions.
Oh, they're fantastic. Like, you never ever get anything like that from a customer. So that one time that you do, you're like, oh my gosh, and it's a complicated build. So it's nice, it's very nice to have something like that.
Yeah, so it's one of those things where documentation is a pain to do. So it does take up a lot of your time. But once you do have these things documented, it's much easier to maintain them. And whenever you have, you know, times in which you do need to drop, you know, give people work instructions, or you know, then you you've already have you already have these made and so you don't need to spend, you know, an hour or two describing how to do this one step when you can just give them the documents that you've already invested time in beforehand to make.
And so before we sign out, is there anything y'all would like to say to yourselves like two years ago, before you started to scale up? So someone, y'all were in your prototype phase, building these like one off kind of units? What would you go back and tell them and it can't be don't use the Intel Edison.
When I joined the company, I was actually an intern. And what I didn't realize at the time that those were one of those are like our original pilots and customers and first gen hardware. What I would have told myself is pay more attention to the fact that you're building the first ones. Because I think quite Yeah, you don't realize until you're like oh, wait a minute, we could have done this differently. And it's like, oh, that that assembly could have been done differently because you're I had the preconception that this is all thought out. And that is not something that was true.
I think I think we yeah, we get what you're going.
Yeah. And I think I think for me, though, it's really take the time to think about how you want it to scale, or how what issues you'll have with the current design. If you're earlier on, like, what issues you'll have, as you scale. And, you know, sure, you can build, you know, one offs. But if you're doing things at scale, what issues are you going to run into? And how are you going to solve those, solve those problems, and sort of really rely on the network of people that you have around you. And I think both startup founders are aware of this, and they have a great network. So being able to draw on those and you know, ask people for advice. And sort of have them review your work would save, you know, a lot of pain points down the road.
Jeff said my statement much more eloquently.
I had the time that you were thinking to think about what I was going to say.
Apply those tomorrow for a year, two years from now.
Exactly. Well, in two years, we'll come back and then you come back, Jeff, so Did you listen to what your own your own advice? And I was like? No. It's easier to say that in hindsight, of course,
they would have ignored the advice and built the reflex.
This one, one reflex that's covered in silicone clubs.
So Matt, and Jeff, where can people like get in contact with you or find out more about y'all,
we have react, we have our website, which is where kinetik.com And if you're interested in learning more about the product and potentially deploying it, we have a contact page on that. And then we have a slight social media presence on Instagram and Twitter, although they're sparse, and I think the most recent Instagram picture is actually have a lot of charging.
I guess we should also note that Matt is also our social media manager right now. So if you reach out on Instagram, you will reach bat.
He will Yes. Can people suggest snacks? On social that is a
slippery slope, but I am completely open to that you are we have a macro fed podcast please candidate snack ordering? Well, I like this, like please, please poke them Ah, but ordering kinetic snacks
making that so they can order them and they can pay for them. We might just not give them back to them and we might not ship to them. If you want to buy an annex I'm all I'm all for. Jeff I need to show you TwitchPlaysPokemon I'll understand Twitch I'm sorry. Why would people watch people other people play games?
It's neuroscience. We are on opposite spectrums here.
It's like sports Jeff it's like because like became
even I'm also probably an outlier but it's like you know why watch baseball on TV where you can just play baseball you know or it's I don't understand
it's but you can play baseball on TV. See,
on Twitch I can understand if you if you're actually playing it, it makes sense. But if you're just like in the stands, watching baseball, the main drawl is like the soda and the beer and the hotdogs but that's you're not even watching the game then.
We can agree to disagree on this point.
If you want to find us on the internet, it's where kinetic and then I am completely open to snack suggestions.
So if you tweet snack, kinetic that's where you get a hold of them on on Twitter
that's available that handle
probably so it's where kinetik.com Yes. Cool. And with that, oh actually just one more thing. There's also a case study about kinetic on the macro fab website. So if you go to macro.com boom there's a case study there that talks about scaling up and all those hardships that we just talked about, you know running as snacks and that
kind of very important thing is as you scale up you also will
you can't can't run out of chocolate pecans because
and with the pecans pecan moment, do you all want to
so that was the macro fab engineering podcast, we are your guest
Matt Glazer,
Oreos, burger Dolman
and Stephen Curry.
See you later everyone take it easy Thank you. Yes, you our listener for downloading our show if you have a cool idea project top Uh, let's see you in and I know Tweet us at McWrap at Longhorn engineer, that's with no O's in Longhorn or at E and G. I think that stands for at No, my bad at analog E and G. And I think that E and G stands for engineer, or email us at podcasts at macro app.com. And we have a new mailing list for the podcast. So go to macro.com/blog/podcast There'll be a little pop up, punch in your email address, and bam, you will get our new mailing list. It's kind of like subscribing to the podcast in a different way. Interesting way. Also, check out our Slack channel, Steve and I hang out there all the time commenting on the latest podcasts and we always chat with our fans. I think we're getting to about 300 people and that slack channel now. So yeah, go join the fun, push it over that 300 number. And if you're not subscribed to that Podcast, the podcast, our podcast if you're not subscribed to our podcast yet, click that subscribe button like in iTunes or whatever podcast app you're using. So that way you get the latest map episode right one right when the RSS feed lands and please review us wherever you listen as it helps the show stay visible and helps new listeners find us. It also makes Steven and I feel really good thanks again guys.
On this Episode, Parker and Stephen have guests Aditya Bansal and Mijael Damian of Kinetic.