Episode 88: Those dweebs Ben and Steve talk about wedding bands. Benjamin is excited that robot software and setup are getting easier and more accessible. Stephen shares some detective work about Dior’s new 3D-printed shoe concept. Ben thinks sustainable aircraft could be in our future. Steve closes with a story on how the USMC outsmarted DARPA’s advanced AI and automation.
Discover the past, present, and future of American manufacturing with https://www.mmsonline.com/madeintheusapodcast
Explore, watch, read, learn, join, and connect at https://www.imts.com/
Tune in to the AM Radio podcast https://www.additivemanufacturing.media/zc/am-radio-podcast
For the latest in Manufacturing Technology news https://www.amtonline.org/resources
Benjamin Moses:Hello everyone. Welcome to the AMT Tech Trends Podcast where we discuss the latest manufacturing technology, research and news. Today's episode is sponsored by Modern Machine Shop Made in the USA Podcast. I'm the director of New Year's Resolutions, Benjamin Moses, and I'm here with-
Stephen LaMarca:Stephen LaMarca, AMT'S Technology analyst.
Benjamin Moses:Steve, how you doing the start of February?
Stephen LaMarca:How's the gym looking for New Year's Resolution? Apparently there's a lot of Tic Tokers who are accusing random people of staring at them, but we won't get into that. We're not about that. We're about the technology.
Benjamin Moses:We've got some future plans that I want to talk about. You're talking about wedding stuff and you got a wedding band.
Stephen LaMarca:Oh man, okay.
Benjamin Moses:I want to know more about your experience.
Stephen LaMarca:So future, sure. Plans, not so much. We have absolutely nothing planned other than that we're getting the rings taken care of now.
Stephen LaMarca:Fortunately, one of our colleagues, Rob, for whatever reason I think he threw a ring at me to pick on me or something. He took his wedding band off and threw it at me as if he was divorcing me.
Benjamin Moses:That's dangerous by the way. I'll never do that for my life.
Stephen LaMarca:But anyway, for whatever reason I decided to put it on, and holy cow, it's a perfect fit. Anyway, my fiance Melissa is like, "Will you..." I sent her the wedding band that I want because it looked really cool.
Benjamin Moses:You going to tell me about that now or later?
Stephen LaMarca:I'll get into that.
Stephen LaMarca:But I send her, "I want this, order this for me now." It's like, "You have to find out your size dumb dumb." I'm just like, "Oh, okay." By chance Rob threw his wedding band at me. I put it on like an idiot and was like, "Oh my, this is a perfect fit. What size are you?" He is like, "10 and a half. I'll bring in a bunch of other ones that you can try too." It was like, "Whoa, whoa. I don't want a hand me down wedding band." But he did anyway and I got the Mitutoyo calipers out and whatnot, and anyway, found out I'm 10 and a half. Took all that information back to Melissa. We specked out the ring. I want a simple wedding band. No engravings or any date and no stupid gemstones or anything, whatever. It's not stupid. If you have a wedding band with gemstones, it's great. Good for you. But a simple rounded wedding band. But I wanted a really cool material and I saw this stuff on Instagram, believe it or not, called Timascus. Titanium Damascus.
Benjamin Moses:Oh, nice.
Stephen LaMarca:There's multiple layers of titanium folded over, the way you'd make Damascus, but because it's titanium and titanium under heat blues really beautifully and it's lightweight and comfortable, but because it blues when you heat it and there's multiple layers of titanium that blue differently under the same amount of heat, it makes this really cool light blue to deep purple tiger stripe pattern. It's a little ... Melissa did accuse me of picking out a tie dye wedding band, but it looks really cool and I thought it was sick. Titanium's lightweight so it's not going to feel like wearing much. Got it at specked out and everything.
But anyway, gave her the information. She ordered the band, and the company that she ordered from is great. It's like they've done this before and they're like, "Okay, you say you're this size." And this doesn't slow down the ordering or production process at all, which is really cool. They get into the production. The company told us everything about how they're made and whatnot. So I got to see a lot of lathe work and that was sick, but they were like, "It's not going to slow down production at all. But in the meantime, we're going to ship you a sizing kit."
Stephen LaMarca:Because it's something like 85% of people. I pulled that percentage out of my bum.
Benjamin Moses:Some very high percentage.
Stephen LaMarca:But it's a very high percentage of people that order wedding bands buy the wrong size.
Benjamin Moses:Sure, yeah.
Stephen LaMarca:So in the two weeks that it takes to make this wedding band for you, you will receive within the next couple days a sizing kit and it will have every quarter size around the size that you think you are. And sure enough, I thought I was a 10 and a half. When Rob brought in his rings, all of his 10 and a halves were different sizes by the way.
Benjamin Moses:Of course.
Stephen LaMarca:Because I took the Mitutoyo Calipers to them. And we're talking within 0.05 millimeters of each other. It made a difference though.
Benjamin Moses:Oh yeah. Well something's contact your skin, every little bit matters.
Stephen LaMarca:But anyway, the company sends all these quarter sizes around the size. Sure enough, I thought I was a 10.5, 10.75 is the most comfortable so far. So I'm wearing my training wheels right now. This is just aluminum, but it's got the comfort bevel on the inside. They milled it all nicely or turned it all nicely.
Benjamin Moses:That's cool.
Stephen LaMarca:It's really comfortable. It just looks like garbage because it's not the real thing.
Benjamin Moses:I'm excited to see that. I do like oxidized titanium even though it just looks-
Stephen LaMarca:Yeah, it's pretty. [inaudible 00:05:33] motorcycle exhausts.
Benjamin Moses:Motorcycle exhausts. Pagani has an amazing exhaust. I just love their process and how they keep their name imprinted on it. It's cool. Speaking of wedding bands, I was doing some yard work and did some stuff outdoors over the weekend and I don't wear any accessories when I'm doing that type of stuff, but for some reason I just came back from Amelia's swim practice. I was like, I'm going to get some wood for the fire pit later from. We got some woods in the back yard. I'm just going to go drag some out, cut it with a circle saw, get it prepped for the fire pit. I put my gloves on, just working gloves because it wasn't super cold, did the stuff. I came back inside to washed my hand and I noticed right away. I always noticed my wedding ring on. It's never comfortable for me. As soon as I get home, I take it off.
Benjamin Moses:Both my wife and I are like, "We don't need this because we don't need to tell people we're married because we're at home." So we just take-
Stephen LaMarca:Yeah, there's no appearance to keep up.
Benjamin Moses:Yeah. First of all, no one's going to approach me anyway. But when I get home I take off the wedding because, to your point, this is a heavier metal and I do feel it on. It's a good cut, but I always feel it on. So when I came back in, went to wash my hand and I looked down. That's strange. I thought I had a wedding band on.
Stephen LaMarca:Uh oh.
Benjamin Moses:I went back to my bedroom where I always put my wedding band, wasn't there. So I was panicking because I was in the woods, dragging up pieces of wood. I thought I wore my gloves the whole time, but one of our bird feeders fell and I was dragging that back out and I wasn't wearing my gloves. "Did I drop my ring in the woods?" Which is literally the worst place. If I dropped it there, there's no point in getting it back.
Benjamin Moses:Luckily I looked around, rummaging around, did a two-minute search in the woods like, "There's no way I'm going to find this." So I gave up. I thought, what if it's in my glove? So I checked the glove, wait a second, let me dig around. Luckily it was in the glove. So when they took the glove off.
Stephen LaMarca:No way, it came off with the glove.
Benjamin Moses:It came off with the glove.
Benjamin Moses:Thank God I didn't lose my wedding ring.
Stephen LaMarca:I was thinking it was in your pocket or something.
Benjamin Moses:I checked all my ... That's what I thought. I thought I took it off when I washed my hands and left in the pocket. It was like, nope.
Stephen LaMarca:That's terrifying.
Benjamin Moses:It's very terrifying.
Stephen LaMarca:I'm going to have to make sure I check my motorcycle gloves every time I ride with the wedding band on.
Benjamin Moses:Yeah. It's actually nice that you have that aluminum test ring. So when I go to the gun range or when I do any type of-
Stephen LaMarca:You bring a beater.
Benjamin Moses:You could wear your beater. Because I don't wear my wedding band when I go to the range. When I used to go on the shop floor and do work on the floor, I used to take all my watch and jewelry off to prevent it from getting scratched.
Stephen LaMarca:I don't think beater and weddings should be used in the same conversation.
Benjamin Moses:Unless you're drinking.
Stephen LaMarca:Oh my God.
Benjamin Moses:Can you tell us about our sponsor today?
Stephen LaMarca:Our sponsor is the Modern Machine Shop Made in the USA Podcast. Tune in for Modern Machine Shop's Made in the USA Podcast to explore manufacturing issues faced by companies making an intentional choice to manufacture in the US. Featuring commentary from OEM leaders, Made in the USA blends its nearly century long expertise with a unique audio storytelling experience to shine a spotlight on the past, present, future of American manufacturing. Find Made in the USA on Apple Podcasts, Spotify, and all major podcast platforms. Follow Modern Machine Shop on Twitter, Facebook and LinkedIn.
Benjamin Moses:Thanks Steve. I want to get into some articles today.
Stephen LaMarca:Yes sir.
Benjamin Moses:We got a bunch. The first I want to talk about automation. Last time we talked about automating programming for robotic arms.
Benjamin Moses:Getting to streamline process. Here we've gotten an article from Automation World talking about what I would call redeployment of equipment. You could also talk about the initial deployment, but if you're in a smaller factory or a factory where they're doing high number of part numbers but low volume, you have a constant changeover of parts. So you're constantly changing from a square to a round or things like that. So to be able to redeploy your robotic arm to either new situations or new parts, it's fairly important, especially when we look at job shops or contract manufacturing.
Benjamin Moses:So there's an interesting article covering a technology from OnRobot. I'll read you a quick summary. So users only have to scan the QR code on their module to securely connect to OnRobot's cloud system, which can then detect all robotic components in the work cell. The robot as well, it's end of arm tool line. And then other components in the work cell are added in the software environment through intuitive software interface.
Benjamin Moses:So they're saying, one, using their end of arm tooling and being able to connect through their cloud platform, they know what the end arm tooling is. So getting to a simulation or a digital representation of the end of arm tooling, that's super fast.
Stephen LaMarca:So this is an on robot product. Is it an end effector that they just plug in and use it as you'd normally use your end of effector for or end of arm tooling for.
Stephen LaMarca:But the end of arm tooling also communicates with your computer system and tells you everything that's connected to it?
Benjamin Moses:Then they also allow the ability to define the robotic tool, the tool path. What they're doing is also, instead of issuing poses and specific positions, they're saying don't do that. Tell it what the objects that you want to pick up and the place and the path will automatically define itself. Similar to the conversation we had the past couple of episodes of, there probably isn't the need to say this pose needs to occur as in all the joints need to go through this orientation. From the end user perspective, they're saying, no, I would like you to just tell me where you want to be and I know roughly what else is in the cell. Now some things in play here, like what is in the cell?
Stephen LaMarca:Right. What was it called last time, last episode? It was called Voxel reservation.
Benjamin Moses:Voxel reservation.
Benjamin Moses:So very similar. Also, when you're designing the cell or the digital environment, you're defining what's in the cell. So that's a very good starting point of course. Then you have the real time data that's going along with that and all the different scenarios. Not just this, but all the different scenarios. So the big takeaway is, and we've been talking about this for quite a while, is the need for automation and the ease of implementation for automation.
Benjamin Moses:So not only are we seeing the cost come down, but the ability to say we don't need a roboticist to get a robotic arm on the floor anymore. Now there are the significant value in integrators and subject matter experts in robots as in help us define what the correct arm we need, maybe help us define what end of arm tooling we need or capability. But in terms of I've got a robotic arm that's been working here and now I need to change it to either another cell or move it or change the part that is picking up, that can occur faster and more frequently. So the return on investment is skyrocketing now because of ease of implementation.
Stephen LaMarca:Yeah. Man, the accessibility of automation, not only is it really impressive where it is now, but it's almost hard to believe how much better it's getting.
Stephen LaMarca:But that being said, OnRobot's doing a great job with all these innovations and they are becoming a staple, a household brand name for end of arm tooling, which probably works against their accessibility because if they're becoming the name brand for end of arm tooling, you know they're charging a lot.
Benjamin Moses:Sure. Maybe.
Stephen LaMarca:We'll see. Yeah, I'm still going to fanboy over them.
Benjamin Moses:Sure. You got an article on additive in the fashion industry?
Benjamin Moses:Tell me more.
Stephen LaMarca:Dior. So we're used to the Adidas working with carbon to make their soles of some of their shoes. Well Dior was like, let's print as much of the shoe as we can.
Stephen LaMarca:So Dior released two images. Well two, I would say concept shoes, because I don't think you can buy them yet. They released a Darby and a boot that I originally thought was a Chelsea because of their lattice structure that the shoes are printed in. At first when I saw it, I was like, oh wow. They just made the entire thing out of elastic.
Stephen LaMarca:Which a Chelsea boot is typically a leather boot that has elastic on the side so you can slip it on and off. But with this, because they're trying to print as much of the boot is possible, the entire thing is a lattice structure.
Stephen LaMarca:So it actually looks more like an Ugg boot, but it's still a totally novel design. The other shoe was a Darby and I did notice there were certain components of the shoes that were not printed. They can only print so much. It was really cool because they included pictures of the raw near net shape part coming out of the printer. Oh, I'm pretty sure. I think it's nylon. It's definitely polymer.
Our team here that knows additive was able to do some detective work and determine that the rich color of black that these boots are printed in. They said the only material brand that can make that rich color of black in either polyurethane or nylon, I think, which is of common material. It had to be nylon because polyurethane would make sense for a sole, but they're not printing the sole.
Stephen LaMarca:So if it's nylon, there's only one company that can make the material with that rich color of black. This is a big one, SLS, selective laser centering for polymer.
Benjamin Moses:That's what they used. Yeah.
Stephen LaMarca:So we're used to laser powder bed fusion for SLS, which is inappropriately marked ... well not inappropriately. Still follows the standard that Jason Jones sent us with the diagram and everything. If you want to be more specific and in detail, SLS is typically polymers and there's very few SLS machines and brands, namely just one, that can use this material, and there's only one brand of material supplier that makes this color black. So it was really cool watching them do their detective work to determine who was making this.
Stephen LaMarca:Who Dior was using to make this, which was cool. But the other stuff, yeah, the sole is not 3D printed. That's clearly injection molded. Then there was other stuff that's leather, patent leather like the tongue, the eyelets for the Darby and the tongue of course, and the welt around the soles of the shoes. That was all leather stuff that was probably stitched on after the fact. But with additive there's always a lot of post-processing. But it's just cool seeing this.
Benjamin Moses:I do appreciate ... so additive and clothing and fashion's been going on for a little bit and I do like the mixed material. So they kind of understand the boundaries of where additive can push them, but also you mentioned the soles, they're probably manufacturing mass assembly.
Benjamin Moses:So is there-
Stephen LaMarca:They do print in the mid sole,
Benjamin Moses:The mid sole, right.
Stephen LaMarca:The mid sole has a lattice, just the Adidas carbon 3D shoe.
Benjamin Moses:Right, but the contact with the actual ground itself, that's a different-
Stephen LaMarca:The actual bottom sole.
Benjamin Moses:Yeah. So that's one thing I do like to see in additive is not just the random exploration, but companies and end users understand you probably don't need the whole thing printed. There's limitations at this point particularly. Get to a state where you can do a portion of it and then attach the other things as needed. To your point, additive is entire process. It's not just 3D printing. It's you prepping the material, growing it, post processing, and then final assembly.
Stephen LaMarca:Yeah. So a lot of people, I was one of them, thought that additive has this huge advantage over subtractive or traditional manufacturing, whatever you want to call it. Material removal in that, with material mover, with CNC machining or turning, all of the setup has to be done in advance. Then it makes the part and the parts done.
Stephen LaMarca:Additive, nope. You're off to the races right away. Click print. Tim hates it when I say that, but just click print and it goes. Then when it says it's done, then you open the door to a big dusty chamber and now you got to move it. You got to brush it off and maybe bake it or something like that.
Benjamin Moses:It's a whole process.
Stephen LaMarca:To what you said, I'm glad you said ... I forgot what you said.
Benjamin Moses:Limitations, boundaries, process.
Stephen LaMarca:I totally lost my train of thought.
Benjamin Moses:All right, let's move on then.
Stephen LaMarca:We can have the cut back.
Benjamin Moses:We got a couple more on a combination of computational capability, fluid design and sustainability. So Airbus, massive airspace company, Airbus or Boeing, not too many options there. Airbus selects all terra SIM solid and zero E sustainability aircraft initiative. So Airbus from the OEM side is driving more sustainable designs of the aircraft, both on the fuselage engines, everything. So what they're looking at is improving their computational fluid design, CFD, to help predict more sustainable designs.
So one of the limitations we've seen, and this is where the benefit of additive is reaping the wards of this is fluid design is very, very difficult. It's very computational intense. So there's been limitations for the past bunch of years on computer hardware. Basically you're simulating particles or every molecule of gas as you push it through something. So if you can imagine tons and tons of elements that you've got to represent and determine pressure drop, determine temperature gradients. A lot of variables involved in that process.
As we've seen, as we talked about the GPU markets and their effect, as new GPUs have been released, new CPUs and the ability to how we use those technologies for computational design has changed radically in the past couple years. Being able to push ... When I was a wee little design guy, everything had to be done fairly local. So my capability was whatever RAM I had on my computer, whatever CPO I had on my computer. But now we're seeing the ability to design semi local, but then push your design back to the server farm to scale up to, I can run over a thousand CPUs and terabytes RAM if I need to. So that ability allows for incredibly complex CFD models that will allow us to come up with some very, very radical designs. Some of the stuff you'll see at the fairly short term are on drones. Drones and propellers for boats, there's a lot of turbulence that comes off of those type of blades and there's a very notable audio signature.
Benjamin Moses:For especially drones.
Stephen LaMarca:Yeah, especially drones.
Benjamin Moses:So what you'll see is, instead of the blade designs, you'll go to a [inaudible 00:21:25] design. So there's connected and it's fairly interesting to actually see that and see if that actually works. But a lot of the turbulence comes from the tip of the blades. So in [inaudible 00:21:36] there's no tip because it's all connected. So it's very fascinating to see the process they got to that iteration and then for them to verify that it works and it actually produces enough lift for the aircraft. So that's the type of iteration we're probably going to see is on the consumer level, fairly radical designs and what we see for blades and moving air, and probably fans. I'm sure Dyson's got a interesting thing coming out soon too.
Benjamin Moses:But that'll transition to aircraft in the next iteration.
Stephen LaMarca:Listen, Dyson is a liar. They say they have that bladeless fan. It was probably a month in before the jig was up. It's not a bladeless fan.
Benjamin Moses:They moved the blade.
Stephen LaMarca:They made a suppressor for a fan.
Benjamin Moses:Big fan of suppressors.
Stephen LaMarca:Big fan suppressors, don't get me wrong. It was like-
Benjamin Moses:Speaking of people who use suppressors, Steve, you've got someone to pull the wool over AI's eyes.
Stephen LaMarca:Yeah. Okay, so last week's tech report, I put in a video at the very end. Well, I put in two videos. One was an almost 45-minute video of a Corvette factory tour. But the second video, I think some people may have glossed over because I put it at the very end, out of a fear that somebody was like ... may give me some negative feedback like, how is this manufacturing related? But hear me out. AI's a big topic in the news as of late, especially with Chat GPT and all the fear and all the hubbub that Chat GPT is causing. I love open AI, for one. But anyway, we can get into that later. There's a lot of AI in the news and long before Chat GPT was a major topic in the news, manufacturing industry has been utilizing, or at least trying to utilize AI for quite some time now. Just talk to Autodesk. They'd be like, we've been using it for the past 20 years or something like that.
Benjamin Moses:Basic vision systems are using a lot of machine learning to object recognition and things like that.
Stephen LaMarca:Absolutely, and they're only getting better. Open AI would probably say, yeah, we've been implementing this. These programs were written 20 years ago. It's just taken them this long to mature because it's an artificial intelligence. It needs to grow up and mature and learn how to do things.
Stephen LaMarca:But anyway, DARPA, there's a book coming out this month that DARPA was working on an AI powered, an AI driven AMR to be employed by the military. These little four wheeled autonomous mobile robots would follow soldiers into a combat zone and help force multiply the fire teams situational awareness, help scan the environment for them and notify them if there's anything out of the ordinary that a human might miss, that the human soldiers might miss. Naturally, before you send something expensive and that required a lot of state-of-the-art technology behind into a war zone, even though that's a great way to field test something, naturally, you want to test it in a controlled environment first.
Stephen LaMarca:Because you also don't want such technology to be captured. So you test it in a friendly area. DARPA decided to test this AI powered, this AI recognition system with an AMR that's driven by the AI as well. Wanted to test it up against a squad of Marines.
Benjamin Moses:Sure, and the goal of it is AMR would detect object or detect humans, right?
Stephen LaMarca:Yes, the Marines were told that you're going to start a thousand feet away from the robot.
Benjamin Moses:A thousand feet or a thousand yards?
Stephen LaMarca:A thousand yards, sorry.
Benjamin Moses:So fairly far away.
Stephen LaMarca:A thousand yards away from ... literally three times what I said. Thousand yards away from the robot. And you have to sneak up to the robot without it detecting you and physically touch the robot.
Benjamin Moses:Right. So a game similar to clear and present danger when Chavez had to sneak up on the trainer.
Stephen LaMarca:The robot has been programmed and learned to recognize people moving, people hiding, like crouched or prone and look for snipers, look for any sort of human movement. The Marines were told this, and clearly with what happens next, they clearly thought, okay, we just don't have to act like normal humans.
Benjamin Moses:Or be a Marine.
Stephen LaMarca:So the first two or three Marines charged the robot right away, but by doing cartwheels, a thousand yards of cartwheels all the way up to a robot.
Benjamin Moses:That's a lot.
Stephen LaMarca:If there's anybody that can do it, forget the gymnasts that go to the Olympics. Marines can do it. They cartwheel all the way up to the robot, touch the robot. Never even had anything. I guess it thought they were tumbleweeds or something. I don't know.
Benjamin Moses:Who knows.
Stephen LaMarca:But it's helping the AI learn and develop. The second, the next marine that went out, went into some nearby woods, ripped all of the limbs and foliage off of a Fir tree and put it on his uniform. And I quote, "proceeded to walk towards the robot like a Fir tree."
Benjamin Moses:Like a Fir tree.
Stephen LaMarca:Apparently Fir trees walk. I haven't seen it before, but the Marines know how they walk. Totally undetected, touched their robot.
Stephen LaMarca:Another failure.
Stephen LaMarca:The last two or three Marines found a large empty cardboard box nearby, put the cardboard box over themselves and snuck up to the robot and touched it. The engineers and observers noted, wrote down in their notes and their observations, that they could audibly hear the Marines giggling as they were sneaking up like solid snake and metal gear solid to this robot and touch it. The AI system and the robot was a total failure. The engineers have to go back to the drawing board because the Marines got them.
Benjamin Moses:They got them.
Stephen LaMarca:So if anybody's worried about Chat GPT, just send the Marines.
Benjamin Moses:That is very entertaining. It's cool to see that ... you should always look at DARPA's challenges. They're amazing. The type of stuff that they put in place and the failures through some of these challenges are fantastic.
Stephen LaMarca:You have to make sure you give DARPA credit, because they fail a lot, but they learn from their failures and they fail because they're the first ones doing it.
Benjamin Moses:Yeah. Correct. And this also brings up-
Stephen LaMarca:You're never going to do anything successfully the first time. Rarely are you going to succeed at something the first time.
Benjamin Moses:It's interesting you brought up the kind of concepts that are in the training model for that type of AI or machine learning application. All you have to do is do something outside of the training set. So obviously they're very creative. Marines are very creative and said these are the boundaries which it's set up and go around those boundaries. I thought that's a very interesting example.
Stephen LaMarca:I'm about to get really nerdy because my mom was a Treky. I remember when she showed me some of the episodes of the Borg, you'll always find a way to defeat the Borg once.
Benjamin Moses:Once, right.
Stephen LaMarca:And you'll never be able to use that way to defeat it again, because they're AI.
Benjamin Moses:Yeah. That's good example's. The way to bring it back to Star Trek.
Stephen LaMarca:Star Trek saw all of this first.
Benjamin Moses:Where can they find more info about us?
Stephen LaMarca:AMT online.org/resources, hit all the buttons.
Benjamin Moses:Thanks Steve.
Stephen LaMarca:Bye everybody.