Embedded - 88: Science Is a Lot Like Quilting

Episode Date: February 12, 2015

Same day PCBs?!?  Danielle Applestone (@dapplestone) chatted with Chris and Elecia about desktop CNC milling using @OtherMachine's OtherMill.  OtherMill links: features tools and materials (neat!) store (kits!) instructables (chocolate spaceships!) kickstarter page miniature mocha pot brass, err.. aluminum turbine (also: what Elecia heard) stories of people using OtherMill Synthetos TinyG controller (also see the Make write up about TinyG) BANT (budget, authority, need, timing): more info

Welcome to Embedded, the show for people who love building gadgets. I'm Alicia White, here with Christopher White. Our guest today is Danielle Applestone from Other Machine Company, the makers of Other Mill, a desktop CNC machine. Hi, Danielle. Thanks for being on the show with us. Thank you. It's good to be here. Could you tell us a bit about yourself? I am... You already said my name, which I forgot about that part.

I was like, oh, yeah. So I'm actually a materials scientist parading as a Silicon Valley CEO. I love it. I'm from the middle of nowhere in Arkansas and sort of got to where I am today by learning how to build things and how to talk to people, essentially. We have a PhD in materials from MIT, is that right?

Okay, yeah. Chemical engineering, I studied it at MIT and then PhD in materials from UT Austin. So, you know, you kind of know what you're talking about. Yeah, well now. Took a while to get here. But yeah, I was fascinated by battery materials and just, you know, I wanted to be the person that someone would come to and say, hey, I need a piece of something that can do this, this and this. Can you make it for me? And so that's essentially material science in a nutshell. And then now you're a Silicon Valley entrepreneur, one of those rare breed of people who come here and ask for lots of money in order to make product.

Yeah. Was that what you expected when you got a materials PhD? No, I just wanted to make batteries. I wanted to make better batteries because I'm like, no matter what happens with this, um, you know, with the future of renewable energy or anything, we're always going to need batteries. So let's, let's do that. Um, and then, you know, I came to, I came to San Francisco to apply for a job at a couple places. And then an old friend of mine said, hey, I think there's something you might be interested in. How about running an education program to develop desktop-sized CNC machines and software for the classroom so that, you know, there can be thousands of engineers just like you? Wow, I didn't realize it was developed for the classroom.

Yeah, originally, yeah, we did some work for DARPA, about a year of work, development work for DARPA. Well, it's the Defense Agency. Yes. What are they doing educating kids? I'm afraid. Well, they actually, Department of Defense did a study, and they decided at the end of their study that the biggest threat to national security was not like chemical warfare or anything like that. It was just not having enough trained people to operate the military infrastructure.

And that means machinists, mechanics, CNC operators, engineers. There just weren't enough of them. And they said, well, let's reinvent shop class for the 21st century. But this time, let's add digital design and actually CNC manufacturing, sort of like modern manufacturing tools. But there really wasn't a class of machines that was right for that environment. Software was too hard. It was going to require too much training for teachers. They were too expensive. And so they paid us to dream on that a little bit and do some building of new types of machines. Did you discover that program first, or did you have the idea and then go look for the funding and discover DARPA?

Well, by the time I showed up, all of the ideas had been had. DARPA put out a request for proposal, so it was their idea. And then the person that I ended up working for, Saul Griffith, he had a team of people who applied for the grant. And Saul Griffith isn't part of your current company. He runs more of an incubator sort of thing? He does. He's our board chairman and one of the original founders of Other Machine Company. But he runs Other Lab, which is essentially, you know, just like we started as a DARPA funded project, there's all of these other little companies that start with government, essentially on government grants. Okay, so now the CNC machine. Yes. How big is it? It's 16 pounds. It's about, it's kind of like a 10-inch cube. So 10 inches wide, 10 inches tall, 10 inches deep. And it's white, and I carry it around with me every week.

Well, it's white but open. And so it looks a little Escher-esque at times because of the way that the lines meet, but they all like kind of shade and it's kind of cool. But the size of a bread box, I mean, you know, we said, is it bigger than a bread box? Everybody's like, when was the last time you had a bread box? But this is about the size of a bread box. Yeah, it's like exactly the size of a bread box.

Maybe people will start saying, is it the size of another mill instead of a bread box? Maybe we'll get to that point, hopefully. So for the mechanically uninitiated who maybe didn't take shop class, or if they did, they used antiquated tools, what exactly does this thing do? What is a CNC milling machine? A CNC milling machine is a tool that you can control with a computer. That CNC is computer numerical control. So with your computer, you tell it what to do. You essentially program it. And all it does is has a very high-speed rotating tool.

The tool is very sharp. And the tool will move through a block of material like wood or metal and cut an object according to how you've programmed it. And so you've used a router in the shop, Christopher. And so it's sort of like a router except sort of more controlled. Yeah, it's just like a router. Okay. And you cut with your cnc milling machine i personally make jewelry with it so i cut metal um brass mostly because it's cheap i can get it from the from the hardware store um and but a lot of people do it or use it for printed circuit boards. So you can cut really precise circuit boards in your home, which is something that people haven't been able to do before. So usually, you know, people who are

super enthusiasts, they will use a chemical etching process, which is kind of a pain, and then you have this chemical waste at the end of it. Or you send your board file to somewhere else, and they make it for you, and they send it back. But both of those processes are really time consuming. And so a lot of people found the other mill, and that's the original use for it. They find that they can just make prototypes of circuit boards in the same day. Because you take your PCB material, which isn't like PCB I might get if I did a fast turn at a shop because that stuff's kind of poisonous. But you have your PCB boards and then you route out your connections

and then you pour copper in it? What is that next? Oh, yeah. No, it's actually really, really fascinating. So the type of circuit board material that you start out with for milling is FR1. And like you pointed out, typically you will get back from a board house, it'll be fiberglass based, which you wouldn't want to mill. So we start with FR1, which is just this essentially resin with a thin layer of copper metal on top of it. And so when you cut out the traces, essentially where do you want the lines to go, you're left behind with your circuits flat on top of this resin material.

Oh, so I'm cutting away the copper. Yes. I don't have to pour copper on later. It's already there. I just have to remove everything else. Yes. Ah. Yes. Yeah, and then people usually will use a solder mask or paint-on solder for wherever they want their components to be,

stick on their components, and then put it in an oven or a hot plate if they're doing surface mount. Otherwise, people just— Yeah, reflow ovens are pretty common. A lot more people have those than I expect. Yeah, in our lab, we essentially took a Cuisinart and just suited it up. Yeah, toaster ovens are pretty common. A lot more people have those than I expect. Yeah, in our lab, we essentially took a Cuisinart and just suited it up. Yeah, toaster reflows. It's quite inexpensive. And so what about the other side? Can you do two-sided boards?

Yes. That's so cool. Yeah, it's nice. Most of the difficulty with two-sided boards is getting things to line up. But we essentially, we ship the machine with an alignment bracket. So you do one side and then you pick it up and flip it over and push it to the other side of the alignment bracket and everything's perfect. And so you drill holes and you have vias, but how do you connect the copper on both sides? Yeah, that's a fun problem too. So some people will just do solder wicking into the hole.

But then we also found these super cheap tiny metal rivets, which you just push into the hole and then you can solder around the edge on each side. It's like a really, I mean, I think probably Asia has solved all of these problems. Tiny metal rivets for doing double-sided boards. Okay, but you make jewelry. Uh-huh. And I can see you've got a bracelet that has words carved into it. Yes.

I guess milled is what we're supposed to be saying, not carved, because there's no carving here. Yeah, well, it's nice to, you know, some people call it carving, but I like to connect it back to the entire history of manufacturing, which it's called milling, so we're going to call it milling. But that's not the only kind of jewelry you can do. Oh, no. People do sort of like 3D wooden jewelry. Also, a lot of people just engrave just about everything. You can take a pretty inexpensive piece of metal and write something special on it, and then that becomes something that someone will cherish a lot.

So we think about jewelry not just being wood or metal or in 3D or casting plastic buttons or things like that, but also a way of infusing story into whatever it is that you're making. Well, I saw you had dog tags up too. I'm thinking all the things that you go to an engraver is for, you could have everything engraved. You could. And you know, one of the things that's interesting to me is, you know, how do I make things in my life more special or how do I make gifts more special for people? And things that you get from the store are often very nameless. You don't have a connection to the story of how it got made. It is also not exactly what you would have made maybe if you could have make it yourself. So I make jewelry because, you know, I couldn't get this brass

bracelet with what I wanted on it anywhere else for affordable. So I just made one. Well, I was actually excited about the jewelry part because you have wax as a material you can carve. Yes. Mill. Yes. Not carve, mill. Sculpt.

Sculpt. Well, yeah, at that point it is sculpting. And so when you talk about boards, you're just milling out a little bit or engraving. You're just getting the top edge. How much, if I was doing wax, how big is my throw here? I mean, can I make, I don't know, a ring? Yeah, so you can cut up to an inch and a quarter thick. So this is one of the ways that it's different from a router so you have three axes and you can

cut a piece that's five and a half pieces long or five and a half inches long by four and a half inches long by one and a quarter inch tall so most of anything that you would have on your body like wearing or in your pocket or or you know all kinds of like wearable devices and things that people do pretty much all of it fits in that space so those are the that's a size that you can mill on the other mill so that's kind of a bar of soap that's been flattened yeah like a really big bar of soap big bar yeah okay um i i did lost wax casting at one point and it's really fun but you have to make your wax mold. And that can be really fun, but it can also be really tough.

And there were times that a machine aspect to it would have been really nice. Well, one of the advantages is that you could do carving of the wax by hand, but if something happens to it or if it messes up or you want to change it slightly, you have to start all over. Lost wax casting means you only get to make one of them. Yeah, totally. But if you just have a file on your computer and you're just cutting these things out, you can modify really easily and you can iterate in a way that there's just a barrier to now if you're carving things by hand.

From your website, it also looks like you could mill chocolate. That seems important. You can mill chocolate, but what is better than that is mill the shape that you want in wax and then pour basically a silicone, make a silicone mold of whatever it is that you want. And so we have people who are using it for... That makes so much more sense than milling chocolate. You don't want the chocolate to have a little machine oil flavor.

I definitely tried it, though. Well, and it would heat up, and so it would melt a little bit. Yeah, well, you know what's interesting? And it's just really cold chocolate. Yeah, if you start out with cold chocolate, the biggest challenge is actually just figuring out how to stick chocolate down to the machine so that it doesn't move while you're milling it. Because if you make the chocolate cold, it sort of sweats as it warms up. Things like that become problems.

But we did mill chocolate once just to see. What about glass? So we've talked about metals, aluminum and brass, wax. I didn't think about wax and then casting. That makes more sense than the casting material. Glass? Yeah. Is that just etching or can you carve glass?

People have etched glass before. It is an advanced sort of, you know, you can also mill stainless steel, but both milling of glass and milling of stainless steel are kind of advanced techniques because you need to use a cutting fluid. So a lubricant or a coolant. And so we don't recommend that for people, though very advanced people have done that before. So using diamond tip blade to cut off, you know, shave microns off the top of a piece of glass is really fun and very exciting to us. But it isn't a common thing. Though, you know, one time someone milled graphite.

So when you mill out graphite instead of milling out a mold in machining wax, you can actually fill it with glass shards and put it in the kiln. So milling graphite where it's really dusty, you have to have a pretty sophisticated vacuum system going, it's possible to then, you know, then you have molds that can be fired. You know, if you could put clay pieces in them and this person was really making ingots for like tessellations in glass. That is so cool. it is very precise i mean that was when i was looking at 3d printers which is the big thing to compare it to because everybody's like oh

do it yourself little box on your kitchen table that makes things it's the opposite of a 3d it is the opposite you're taking things away instead of adding. But all the 3D printers that I was willing to pay for were pretty imprecise. Yeah, totally. With a milling machine, precision and also accuracy is the name of the game. So when you're doing something, you want features to be very small. And then you also want, when you flip your board over to do the other side, you want things to line up exactly. And so the positioning tolerance of this machine is 15 micron.

So you can go to the same spot over and over and over, and you're going to be within 15 microns, which is, that's the accuracy part. And the precision is really like, okay, I need to be able to do small features. So you can do features that are a thousandth of an inch, essentially. that accuracy part. And the precision is really like, okay, I need to be able to do small features. So it's, you can do features that are a thousandth of an inch, essentially. That is very, very small.

Yeah. I mean, for something that's, you know, around $2,000, it's, it's pretty incredible. So $2,000. I saw you did a Kickstarter where the units were like 900 to 1200.

Uh huh. But now it's 2000. Yeah. You're really punishing us non-Kickstarter folks, aren't you? You know, the people on Kickstarter are really just gems. They are the people who support you when you're making a thing that doesn't exist yet. It's going to arrive at their desk, you know, in a variable amount of time. And they're the guinea pigs. So I think you should really reward those people. Also, we had no idea how much it was going to cost to make, you know, we had ballparked it, but we didn't know. And, you know, you make an estimate about like, okay, well, how are we going to run a business off of this? What should the price really be if we're going to fully support people for the next 100 years of desktop milling?

So that's the reason for the difference in price. So that sounds like the on-top cost, the support and the packaging and other later things were added. But did the BOM cost change a lot? It did. We had to make a machine that was fully certified, which meant we had to add some things that weren't there to begin with. We also made... Fully certified by?

The FCC. So just to make sure that the mill, when you plug it into your wall and you turn it on, it's not emitting strange frequencies and interfering with your other electronic devices. So that's what the United States requires. So we did that, but we also added some features based on feedback from our Kickstarter customers. You know, it's fully enclosed now. It has windows, which, you know, basically lockout windows so that when you open the window, the spindle will pause, things like that.

It has an external power supply for fire safety. So we did a lot of engineering. We also made the parts of the machine that used to be made out of plastic are made out of machined aluminum now, which makes them sturdier. And in addition to that, we made the spindle speed higher. So the spindle now is 16,000 RPM versus with Kickstarter it was 12,500. How does that change anything? Well, it changes the type of materials that you can cut and basically like the quality of the cut that you make in them.

So there's a couple variables. Whenever you put a new block of material in there and you want to cut a file with a certain size of features, you have to choose a tool. And there's essentially a giant book, which is, you know, it's called the Machinery's Handbook, and it tells you for any given material and tool, how fast should I be moving my tool through that material? How fast should, you know, how many RPMs do I really need to be spinning this tool at to be able to make the highest quality cut? And so there's all of these calculations that go into it. Essentially what we did was we took that book, we, you know, squished it down and we put it in the software. And so now what the user sees is I'm going to select a material from a drop-down menu.

I'm going to select a tool from a drop-down menu. And, you know, now I'm good to go. All the defaults are in there. So making the spindle a higher speed basically unlocked a couple of, you know, basically like performance. So we can do stainless steel and things like that now and then does it also tell you what bit to put in as well as once you have the materials and it'll yeah it'll recommend a tool to you um a lot of you know a lot of the the challenge with a cnc machine is just the software like what we're talking about now is like how do you interact with this machine um so we've made it very visual.

It'll show you what it'll look like with a suggested tool, and then you can either say, that's not quite the level of detail I want. I'm going to add a tool that's smaller, and it'll update it and show you, okay, this is what it'll look like with a higher detail tool. You're interacting with the software in a visual way, just like, is this the part that I want?

Is it going to turn out like I want? And sort of just clicking through some choices, essentially. Does it do a two-tool sort of solution where you rough out stuff and then you smooth the areas that need smoothing? Yeah, you can do that too. And it has a tool-changing sequence where it will just pause and tell you which tool to load. And so your company has done a lot of software. Yes. Because just making the Millie machine wasn't hard enough. You support a bunch of different files.

I saw SVG, which I think is the same thing i use on the laser cutter when i use it it's vector format yeah yeah and that's for etching and and for like pcbs well you know honestly a lot of things are are just um extrusions it's essentially two and a half d so i can have you know i could make make one layer with a 2D shape that's a certain depth. And then I can have another shape overlaid on top of that that's a certain depth. And so there's a lot of forms that you want to make that are just sort of stacked 2D images. And we find that it's easier to get people started with 2D design, you know, in Illustrator or Inkscape or something like that, draw an SVG, cut that, and then you've taught people essentially 80% of what they need to know, which is a digital file can become physical.

This is how I set up the machine. Oh, it's going to recommend tools for me, you know, and I can cut, you know, I'll cut one. And yeah, it's awesome because you don't have to start out with 3D design. You can just use SVGs. But we also, you know, we import G-code, which is like the standard machine programming language. So if you are a more advanced user and you use SolidWorks and, you know, Mastercam or whatever, you're used to those kind of programs,

you can also just import those files for doing 3D objects. Can you plan, like, multi-step uh programs i guess where you mentioned with the pcbs you can flip them over and do the other side for other things could you do that to make say overhangs or you know flip the piece over and do the underside and have that all be sort of one program or is that would that be considered i need two separate sets of uh steps to do that well if you're doing a full 3D object, for example, we did on our website, you'll see it's a 3D coffee pot, a solid maple mocha pot, I guess. And that required two files, and you essentially, in the middle,

you cut the first file, and then you flip over the block of wood, and then you cut the first file, and then you flip over the block of wood, and then you cut the second file. So it is a seamless full piece, but it was done in two steps. So in that case, you wouldn't be able to sort of stack up files because you have to interact with the machine in the middle. But what I like to do is if I am going to do a series of 2D cuts, like there's another one on our website, which is a brass turban, like a micro turban. That's just a stack of 2D files that you just load all of the files in and press go. It'll just cut all of them.

What do you mean? Brass turban? Yeah. A hat turban or? No, like a turbine. I'm from Arkansas, so I've sort of worked. Mot hat turban or? No. What? Like a turbine. I'm from Arkansas, so I've sort of worked. Oh, okay, a motor-y thing. Yes.

Sorry. Yeah. Yeah, we were doing some, one of our team members was doing some experiments with passive cooling and structures for that, so he was making microturbines, turbines. Either way. Sorry.

You mentioned cooling, which they were making an object to do cooling but uh-huh uh for certain materials do you also have to provide uh some sort of i know what i'm using the dremel tool which is a terrible example to do things with wood oftentimes it tries to light on fire and that's probably because i don't have that book and because you've got the little thing up too high that's why that's why i said about the book but are there other materials that you have to do extra work to keep them from damaging themselves under the stress oil yeah but we don't recommend it that it's it's it's a dry tool essentially so when you're doing when you're doing cuts in wood you're you're moving so fast through the material it's not spending very much time in a place so it doesn't heat up that spot very much um also you know you have to think about these tools are

pretty small so you know you're not even you don don't have a whole lot of contact, like the surface area of the point of contact isn't very big either. So we don't, we have had that problem. Sometimes when, you know, when we're like pushing it to its limits and, you know, milling out big pieces of wood and not vacuuming it just to see what it'll do. But most people, if they're doing that, they'll just open the window in the middle of the job, vacuum it out, close the window, and then resume the job. You also do PCB files direct. I don't have to try to do something strange in Eagle

where I save it to some format that I can then use SVG or other weird, you read Eagle files direct. Yeah. Yes. We import directly BRD files and also Gerber files. So, you know, the idea is that, all right, we want people instead of mailing away to get these boards to make them in house. So let's just take the files that they would be sending out and import those and let them just cut without doing anything else. So having this really very different turnaround time.

I mean, software is one of those things I talk about. One good reason to get started with web software is because you can immediately see the effects of what you're doing. One of the reasons to get started with Arduino as soon as you're willing to touch something is because blinking a light is stupidly easy and you can immediately see the effects. But PCBs have always been something that you make and then you spend a little longer on it because once you send it out, you don't get

it back for a week. And there's this risk associated with a bad trace or incorrect things. And that doesn't seem like that's true if I can just mill it up on my desk in a day. Yeah. You know, the other thing is you can be more adventurous. So things that you wouldn't, you know, you feel this burden of, well, I'm going to drop a couple hundred dollars to get this board back. And is it really going to work? I'm going to, you know. I might as well get 10 because that's not any more expensive than getting two. Yeah, exactly.

And you have to wait. And so I think that what you were hinting at is like making hardware design, especially with electronics, more like software design. You know, you sit there and you have an idea and you test it out electronics, more like software design. You know, you just use it there and you have an idea and you test it out and you're like, nope, I'm gonna do something else, you know, and you keep iterating on it. The other thing is, it's, it's unreasonably hard to get a PCB in a weird shape. So if you are making some sort of new wearable device, and you only have Yeah, you've got some small real estate, you want to figure out like, okay, how do I make a board that fits in a regular shape? It's not a rectangular, you know, prism at all. It's even more expensive to get boards like that. But for

us, you can just draw the, you know, draw whatever shape it is that you want. You can use even like the CAD for the device you're trying to put it in and make something totally weird. So people also sometimes will use the circuit board as a structural component to the devices that they're making, like making little robots, but where the back of the robot is just the circuit itself, which I think is awesome. Yeah. Uh-huh. Make that processor into a backpack. Yeah, exactly. Well, you know what's also cool is you notice when you watch people's projects, they start to see the PCB as not just like this

unchangeable object. It's decorated. So when you have the ability to make the traces look whatever, like however you want, it doesn't even have to be efficient, but maybe it's more artistic. People are really taking it there. They're, you know, they're like the way they arrange the resistors, the way they arrange the lights in the circuit boards that they make are maybe slightly less efficient from a packing standpoint, but they're more beautiful. And that's been an unexpected thing that we've seen. Yes, as you were talking about board shapes, I was thinking, well, you could do little stars and put an LED in the center

of the star, and you can actually use the copper at the edges so that you get a little more sparkle. Yeah, I can see that. But who is your audience? I know Arduino was originally more artist-based, and it's gotten more into education. And you said you started as education, but now who's your audience? It's really product designers, and specifically product designers who are doing electronics design. As we grow and we can create more resources for people who are new, eventually we'll start gathering more of the Etsy entrepreneur crowd. People who are making custom objects and jewelry and things like that, printmaking and letterpress. But for now, it's people who want to make either, you know, 3D components out of aluminum or metal, and they're doing prototyping essentially of that,

or they're doing electronics. So you are going more for the professional folks and not as much for the hobbyists? Well, that is basically just a function of trying to run a business. So there's a thing which I learned from one of my good friends who's been in sales for a while. She's like, you've got to go for the bant. She's like, you've got to go for the people with the budget, the authority to buy, the need, and on the timeline. And so we want to be around for a long time. And so you also have to think about who is our market from a business standpoint. And for now, it means learning who really needs this machine now with the budget to buy it.

And so are you still working with schools and tech shop-like organizations that are into education? Yeah, absolutely. I think that we'll never really stray from our core. It's very important. If we want the future of manufacturing to be distributed, to be smaller batch and more custom and less wasteful, we need a generation of people who know how to run these types of machines and know how to design for them. And for that reason, you have to interact with schools and support students who are getting exposed to CNC machines for the first time. This is nothing like the milling lathe or any of the tools I used in any of my shop classes. It's so small and cute. Yeah, that's the point. You have to feel like this can be a part of your life

and that it's not intimidating. It's a thing that anyone can do. Well, I liked the fact that the machines were all intimidating and people would leave you alone. If there was a chance that they'd get their finger cut off if they were close enough, that was awesome. But you're making it so approachable. The machine doesn't require a foreman to supervise you it's just not real yeah and i would i would wager that you're probably a little outside the the norm when it

comes to approaching technology um there's a lot of people who are they are brand spanking new and it it you know not only needs to feel accessible from, you know, the physical aspect of it, but also you need to have ways of people, you know, so that they can achieve success easily, which is why, you know, you can just go to the internet and download an SVG of your favorite animal or whatever it is and cut that. And that's basically all you need to know. You put a piece of material in there. You tell it what material, what size it is, and then what is the file that you want to cut.

And you can choose from the different tools and stuff, but that's as simple as it is. And I think that when people have an idea for something, even it's just like, I want to combine my name with this shape. And then they see it happen. Even if it's just an engraving, it's like, oh yeah, it's just an engraving. It's just a dog tag, except for it's my dog tag. And that was like the thing that I wanted to do, I did. And I think that you have to make it that easy if you want way more people to participate in any technology, really. I agree.

I wholeheartedly agree. So what's inside? The machine? Yeah. Yeah. As soon as you said that, I pictured the bottom of the machine with the power board and the motor controller board. And yeah, so we, inside the machine, sort of in the bottom part where all the electronics are, it's run by a tiny G motor controller board. So this is a piece of open hardware made by a company called Synthetos. Thedos. And essentially, we thought that the easiest way to get started was to have an off

the shelf component that we could program, you know, put our own firmware on and things like that. And that also, you know, the brain of the machine would be something that anyone could have access to. So we write our own software that runs the other mill. But because we have this open, you know, this piece of open hardware as the motor controller board, anybody could write software that would run the other mill as well. But isn't the tiny G usually used in 3D printers? Yeah, but it's all just motion. You know, it's coordinates basically interpreting G-code and translating that into the way that the motors are going to move and at the speed, you know, changing the speed that the motors are going to move. So it doesn't, it's multipurpose.

You know, that motor controller board will run any style of CNC machine that you would want, you know, within reason. But it's designed specifically for desktop CNC machines. I just designed the worst performance art machine in my mind. It's a 3D printer and mill. It prints and then it just erases what it printed. Yeah, very existential. Yeah, it's just called the futility printer.

I like that. So the tiny G is what you've built the product around. And then the difficult parts of actually building the product are the mechanicals. Is that right? Or are there other weird and cool things? Oh, yeah. So one of the interesting things is, you know, there's a few reasons why there are not very many, in fact, zero, desktop CNC mills that can make really precise cuts in aluminum, essentially. And one of those is you have to have a really rigid machine. So when the tool is cutting and it's moving through the material, it's vibrating a lot. And so if you're vibrating, you're not going to be right on point.

When I was talking about the precision and accuracy, that's basically going to go out the window. But HDPE, high-density polyethylene, that's what our frame is made out of. It's amazing at dampening vibration. And so we have made all of the components of the machine, you know, the spindle assembly and, and the Y bed and everything is super rigid, very well balanced. In fact, if you have a spindle assembly out of a machine, you can just set it on the desk and run it at 16,000 RPM and it will not

budge. It's perfectly balanced. But even that, even when you have balanced all the components and you put them together in this frame, HTTP is just magic because it takes all of the extra vibrations that you can't really account for and dampens them. Well, like the environment, I mean, a truck goes by and your house rumbles a little bit that that's going to affect the precision. Well, the only thing that really matters is the relationship between the position of the cutting tool and the cutting surface. So if you have those two things tightly coupled and very rigid, then it doesn't matter. You can, I've in fact dropped a machine on the ground. Basically, you know, you walk by and you like hook a cord on

your foot and then, you know, a CNC machine falls, that is cutting, falls off the desk. It was totally fine. And it's because you have to have those two things coupled and the rest of the world can be chaos. But if those two things are aligned, you're good to go. And so, you know, in fact, the last thing that – the last sort of paces that we put the machine through before it leaves the factory is we make the machine cut the surface of its own bed. So the bed is aluminum and that's where you put your, you know, whatever it is you want to cut. So we put in a quarter inch tool and with a flat end on it and we mill out the surface of its own, you know, it mills out the surface of its own bed and then it's perfectly aligned. I feel like you should talk to the Navy. I want one of these on big boats that move around because I remember doing some software for those and they move around a

lot yeah well there's like a the Cal Maritime Academy you know they it's all ship based and so we've been talking to them actually about oh what if you could have like the mobile CNC machine on your boat um yeah send it to the ISS they have a 3d printer now they should have a mill exactly you know they already have a mill do. I think they've had a milling machine up there for a really long time. It was just like, you know, milling is so boring to the media. Maybe not you all, but. Since it's based on open hardware, do you worry people are just going to take your idea? No, you know, hardware is going to be commoditized no matter what we do. So might as well just make it all open and give everybody access and do you expect to make money with that you know um i think that it's for us it's not just

about the hardware it's about how well the hardware works with the software and how well we support the learning process of our customers. So, you know, there's really no substitute for caring about your customers. And people recognize that. And so, you know, our goal is, you know, in the end to really sell knowledge. So it doesn't really matter what happens with the hardware, you know, We are going to educate people and make sure that they grow up to become mini machinists. I'm not convinced about your selling knowledge because you're giving it away on Instructables with all of these really neat little kits of information and how to build the… It's true you know in the in the in the beginning it's all about you know it's it's valuable for us to see people making things and have our machine in front of people so we are

gathering as much value from our customers as they're gathering from us by using this machine because nobody's ever tried to do this before you know know, there's been, there's a machine by Roland called the iModela, and it's sort of like a portable milling machine, but it only does foam, and, you know, the software is pretty rickety. So this is kind of like the biggest experiment in really small, affordable desktop milling machine that can do high precision in in in you know aluminum and and

softer and it's good for it's actually i mean it could be argued that it's good for you for other people to enter the market because it brings more awareness to the entire uh product area of affordable milling that probably doesn't exist before oh yeah totally i mean i was told you know a couple weeks ago by a potential investor, it takes two points to make a line. He's like, if you were just out here with no competition, that would be really weird. And it's true. Our competition is not other CNC mills.

It's people not knowing that CNC milling is possible. Just people not knowing that they could have access to these tools. I would say your other competition is 3D printers. I disagree. I think they're synergistic. I think people who want to have 3D printers would look at this and say, I want to have that too. Because you can do a lot more having both than just one or the other.

And it goes to this whole moving away from mass production to being able to do kind of craft production and just a few low volume things easily in an affordable manner. It's kind of like self-driving cars. What does that mean for society? Yeah, I think, you know, what we see is often people who buy an other mill also have a 3D printer. And it's because a 3D printer, you get so much complexity for free. Like you can have a 3D object that you can print and it's in any form and it can be biological in nature or whatever it is. Whereas that would be very difficult. You would have to, you would have to be pretty sophisticated to cut, you know, the inside of an ear canal, for example, like on a, on a milling machine that you couldn't do it, but it would be very difficult. Um, whereas with a 3d

printer, you're just like, cool, we scanned that print it, you know? Um, and so I think for making things that are more organic, more biological 3d printing printing is huge. That includes watches or rings, things that you want to just make in plastic, just prototype. It's very quick to make stuff that interfaces in a regular shape with a 3D printer. And then you can make all the aluminum pieces and wood pieces and electronics with the mill. Oh, I see.

So you really are advocating we get both. Yeah. But another mill first. Because it's easier to start out with. Well, a lot of the 3D printers we've looked at require so much care and feeding. They're like little pets that need constant attention. In your Kickstarter, it looked like you were trying to say that Other Mill was not going to be like that. You were going to go against that trend.

Was that intentional? Yes. The reason is, you know, the world is not helped out by a bunch of machines that just sort of sit on shelves. Like you have to have something that's easy to use. It's going to last forever and it's easy to maintain. And for us, you know, highest ticket item on the bomb essentially. And it's for a reason you want that part. You just can't, you can't re-engineer that part. You can't like do anything to make up for, you know, a crappy motor or whatever. Um, so, uh, yeah, so we put a lot of effort into that because I want this to be a hand-me-down tool. That's how I think of it, is like my mom handed down a sewing machine to me and I want to be able to hand this tool down to my kid.

Not that he already doesn't use it, but it should be like that. It's not a thing that's thrown away. It is a thing that you use and it will be there. Just like the old sturdy Bridgeport mills and the bottom of old machine shops that have just been there forever we want this to be that way too and you know that's another reason why the cost was higher than on kickstarter is because we invested

in really high quality components that's such a different take on it than i hear from 3d printer folks which is like this will be great you should one. And then you should buy one in two years when we have our crap together and it actually works. And then you should buy one in four years because that one will be better. And this idea that you're trying to build for the long term is pretty cool. Yeah. And even, you know, we've essentially made two versions of the other mill. One was for Kickstarter and then one is the sort of commercial version. And we you know, the first version Kickstarter machines are still going strong and we use them around the office because we have more of them around. And it's part of what we care about to

make a tool that is going to be supported into the future. Where do you do your manufacturing? We do final assembly and quality control in San Francisco in the mission. We're in the old Schoenstein pipe organ factory. It's historical landmark number 99, I believe. So we've essentially retrofitted it into a CNC machine assembly place. And then all of our components get made, well, I mean, hundreds of people. It requires hundreds of people to make another mill is the truth of it.

So things like folded metal parts and the wiring harnesses and the power boards and the machined aluminum components are all made by contract manufacturers here or in Asia. But essentially, you know, it's like hundreds of people within a 200-mile radius of us are all required. Even like packaging, like printing and cutting the boxes that happens, you know, on the peninsula. So it's pretty intense, this sort of machine that we've got going. Because you build, you know, you design a piece of hardware, which is a machine, and then you have to design the machine, which is the factory that makes it and all of the moving parts of the people that are involved with each different section and quality control for each of those. It's, you know, you got to love systems,

essentially. You've chosen to do a lot of that stuff locally. It seems like at first blush that that would be much more expensive than even just outsourcing to the Midwest for, say, packaging or something. Have you found that not to be the case, or just the advantage of having stuff so close by outweighs the extra cost? Yeah, well, you know, in the beginning, you're learning a lot, which means you are failing a lot and making a lot of mistakes. And being able to drive to your contract manufacturer and sit down

across from them and be like, I know we're making batches of 250. And that does not represent a large portion of your revenue for this year. However, this is going to grow and, you know, we're going to make mistakes, but we're going to be here and we're going to fix them. So it's a lot about relationships, but also being able to iterate quickly. Like if you are, you know, if I, if we send a part away to China, it, it takes way longer to get that thing back. And because we are always improving the product, you know, we just can't afford to wait that long. And, you know, like I said before about relationships, it's very important to be able to sit face to face. And if I had to go to, you know,

our 20 different component suppliers, and they were all in different countries, that would be super expensive, not, you know, to say nothing of the cost of, of waiting on parts to arrive. So, you know, we went local with, with pretty much everything. Likely if you were building it in China, you would be spending a lot of time in China because these sorts of relationships matter there as well. Yeah. It's not just here, go build my part. Oh, totally. That never works out the way people think it will. Yeah. And, you know, when you're a startup and you have limited resources and limited people on the team, you have to make that choice. You're like, could I afford to have one or two of my engineers living in China for the next two years?

I really need them here. It's more efficient for us to have people close. How big is Other Machine? We are 20 people. Still pretty small. Yes. So it's about a third hardware, a third software, and a third community in sales and marketing and me.

You launched with a Kickstarter and you wanted 50K, but you got 311. Yes. Why did you ask for so little? Well, it seems like it would be silly, except for we had no idea that it was going to work at all, but we wanted to at least make 50 machines we're like okay if we can at least get 50 people interested we can build 50 machines and it'll make sense and so that's why it was like you know 9. Yes. And how would it have all been different for you? I mean, would you be at 10 people now? Do you think you would have succeeded?

Well, the real story is that you run a Kickstarter campaign, but for something as complicated as the other mill, you're really building a company as well. So we used Kickstarter as our seed funding. And after that, we went out and got additional funding. So we raised $3 million over the next six months, essentially. And would it have succeeded? Well, you know, we would have definitely built those 50 machines, because my word is by bond. So I would absolutely have delivered on that no matter what it took. But if we were not able to convince outside investors of the value of what we had created, you know, that would have, that would have been the end of it. Kickstarter does help convince people that not only do you have a product you can ship and that there's a market for it,

but that they should give you money because other people have. Yes. It's really amazing how nobody will give you money if you actually need it. Yeah, you know, the other thing is we ran a funny campaign. So we had 33, we have 33 investors, which is unusual. That's a lot. It's a lot of investors. But they really, it's all people who really care about this space.

You know, they're working in similar areas of technology and they, you know, they believe in the future that we want to make, you know, make happen. What's next for Other Machine? Well? More precision? Cheaper? Oh, I mean, geez, the feature list is pretty intense for what we would want to do. But right now, we just have to prove that we can make

very, very many other mills and make them all high quality. What's the lead time for them? As of this week, I believe it's seven weeks. Okay. So you're still, if I order one, you still are building it. You're not keeping a stock yet. No. Well, that's the other thing you need money for.

If you raise money and you're using it to operate, you know, a certain portion of it can become inventory. But to be honest, you know, our lead time was eight weeks because our longest lead time part was also eight weeks. So you get in these like interesting, interesting challenges there. But yeah, our goal this year is to bring that lead time way down. And I wonder, since we started the conversation with you expecting to build batteries, do batteries seem so exciting as they did? Have you converted to the Silicon Valley entrepreneur? Or do you still see some part of you going back and working on batteries? Well, I mean, you know, I'm a lady with a 10-year-old and a PhD in material science, so I don't really fit in anyway. I don't think I'll ever convert to

Silicon Valley entrepreneur. But, you know, I think that batteries is, it's just so fascinating. And I really, you know, I looked at battery design and battery materials chemistry kind of from a ground up perspective. And so I went way back to like, all right, let's start at the periodic table. Okay, let's look at all the like, old phase diagrams and old, you know, x-ray crystallography data and really tried to dig out new solutions from research that had been done in the early part of the 1900s. And I would love to do that again. You know, I found it, you know, because it's like, it's kind of like an Indiana Jones movie in a way you're like in the library and you're digging around, you're trying to make connections. It's kind of, yeah, like part spy movie, part Indiana Jones. You're like, you see connections that nobody else has

seen before. And you can actually go make that into a real material and, you know, a little mold. It hasn't even been patented yet. Holy cow, you know, like that is so exciting. You know, and one of the reasons is when you're a scientist, you publish research and sometimes it just disappears or it'll just be in that library. But it's really awesome to be able to take work that other people did that maybe would never see the light of day and turn it into something that now is relevant to us. And it's, you know, it's, I kind of liken it to like, you know, have you ever gone to thrift store and you buy a, buy a quilt top? It's like a quilt top that has no quilt insides or bottom, but someone made this quilt top. And it's basically like, whenever I see one of those,

I must buy it and complete the quilt because another human being put so much effort into it, I feel compelled to take it further. So anyway, I see science as a lot like quilting. Title. Yeah. Do you have any more questions, Christopher? I have one question. How many fires did you start during development?

And can you describe the best one? You know, the funny thing is people ask me, they're like, oh, what happens when you break a tool? And it's really boring. Basically, you break a tool and the machine just mills into the air with its stubby thing that no longer has a sharp part on it. And our fires are, when we have had something like- I thought she did say fires. Yeah, no. Just small electrical things. I mean, we're trying to really push the boundaries of what's possible with different different electronics and, and this machine in a small space.

But yeah, no, not nothing exciting. We did have a lightning strike though. Once while we were, we were milling out frames. This was like one of our, like,

you know, tragedies happened during your Kickstarter campaign that slow things down. Well, one of those was a lightning strike, literally that, that burned out one of the boards in our large manufacturing machine that makes the frames. So we were down for two weeks and we had to, yeah, we had to call in the people to replace that, which was kind of

intense because it was, you know, single point of failure. Don't do that again. What are the odds? Yeah. They're exactly the same as they were before the first. intense because it was, you know, single point of failure. Don't do that again. What are the odds? They're exactly the same as they were before the first lightning strike. I've really enjoyed speaking with you, but I would like to free you to your normally scheduled afternoon. Do you have any thoughts you'd like to leave us with? The thought I would love to leave everybody, everybody with is,

you know, my parents definitely told me that anything was possible and it sounds really cheesy, but it's definitely true. And we come at that angle from, you know, we come, we come at that topic from the angle of empowering people to make physical objects, but it's really the power to make something change in your environment.

And I think that when you do that with a physical object, you have an idea and you prove to yourself that you're capable of making it happen, that really just infects all of the areas of your life eventually. I can make this thing. Well, what else have I been kicking around on that I could do? Maybe I want to start eating healthier. Maybe I can change my career, things like that. So for us, it's not only about digital files become physical things. It's about people becoming fully realized versions of themselves, which to me is definitely worth working on. That's great. Thank you.

Thank you. Thanks, Danielle. Our guest has been Danielle Applestone, CEO of Other Machine Company. If you'd like to check out Other Mill, go to othermachine.co. That's right. I didn't mistype Mill, go to othermachine.co. That's right.

I didn't mistype that. It's othermachine.co. And it will be in the show notes, of course. Thank you to Christopher White for co-hosting and producing. If you'd like to say hello to us or to Danielle, hit our contact link on embedded.fm or email us show at embedded.fm. If you'd like a sticker, send us a note. Tell us you told somebody else about the show and include your address. I'd be happy to send you a sticker.

And thank you for listening. Also, thank you very much to Warren Young. I'm finding Stephen Wright to be quite amusing, though I think I've accidentally stolen some of his jokes. Now, I do have a final thought for you. From Brian Selznick, The Invention of Hugo Cabaret was his jokes. Now, I do have a final thought for you from Brian Selznick. The invention of Hugo Cabaret was his book. And in it, he says, I like to imagine that the world is one big machine. You know, machines never have any extra parts. They have the exact number and type of

parts they need. So I figure if the world is one big machine, I have to be here for a reason. And that means you have to be here for some reason too.