DIY Automation That Actually Works: How One 80-Person Shop Cracked High-Mix Lights-Out Manufacturing, Ep #39

[0:00] You’re three for three right now on a killing, like a common industry misconceptions on automation, right? That it’s not for small shops that it can’t be used on like non-standard materials and that it doesn’t work in high mix and you’ve made it work in all those scenarios. Nick. What’s up, Eric? How you doing? I’m hanging in there. Really? We’re good. Yeah. You know, I know that this won’t come out, uh, in a timely fashion, but, uh, we both survived snowmageddon. Yeah. That’s part of the struggles right now. It’s like, it was beautiful last week, 75 degrees. We had our European colleagues here. And, you know, Rick, our German managing director, he’s like, oh, it’s so, it’s so nice this time of year. And I was like, yeah, it’s going to be a blizzard in a few days. And he looked at me like, what? Yeah. That’s crazy. The weather around, uh, the Midwest in March. Yep. Uh, you know, it’s funny. A lot of teams are in the Midwest are starting their baseball seasons. So it’s like, man. All right. Well, uh, always good to see you, my friend. Uh, so you remember the episode we did on some of the cautionary tales of DIY automation and cobots. Yeah. It was a theme of our story too, with, uh, the automation trap with Sven and, or what was it? The Klaus and Gary.

[1:33] Klaus and Gary. Yes. Iron clunk machining. Very good memory. Yeah. So we’ve touched on sort of like the cautionary tale around cobots and the cobot dream and sort of how it came to fruition, but kind of not like, wasn’t all the hype that the industry was told it was going to be well, I’ve got some great news for you, Nick, today’s guests aren’t all bad. They’re not. And today’s guest is going to tell a great story and walk us through what happens when you get cobots right. I’m excited to have her introduced. So tell me about our guest, Eric. Yeah. So Katie Murphy works for a company called Archimed in Indiana. And I first met her during my time with fifth axis. Uh, and she’s, she’s got an amazing automation brain. I don’t want to steal all her introduction thunder, but all I can tell you is she led a team, brought in a cobot, brought in their own automation hardware, built their own sort of dedicated cell, and they’re having a ton of success with it. And they do monitoring and they do all these great advanced manufacturing techniques. I can’t wait for her to tell her story. So without further ado, Katie, welcome to the lights out podcast.

[2:48] Thank you guys for having me. I’ve been looking forward to it. This is a, an awesome opportunity. Yeah. We’re super excited to have you on. And, uh, yeah, so start, maybe tell us a little something about yourself, how you got into the industry. Uh, and, uh, we always start with the origin story. Yeah, exactly. We want to know what, what about someone’s life and background caused them to go into this crazy metalworking world that we live in. Yeah. I happened here by chance and I’m really grateful for it. I studied, um, biomedical engineering and computer science, which, um, ended up being a pretty good combination to start to explore deploying automation robots for a medical, uh, company. And so I kind of happened into manufacturing and fell in love with the hands-on aspects of it. And then automation really tied in my love for computer science and logic-based processing, but also that creativity of design and really seeing a project all the way through from start to finish, I like what she said. Logic-based processing. And my, my creative brain was just like, well, I don’t know what that is. And then the next thing she says is, yeah, but the creativity of design and bringing a project from start to finish. And I was like, oh, I’m back.

[4:04] Absolutely. So you seem to be able to blend the right and the left brain pretty well, Katie. Yeah, when you’re doing the DIY projects, you kind of have to have a little bit of everything. You’re wearing a lot of hats and you have to be able to combine that real logic engineering brain with also the creative, how the heck are we going to achieve this brain? And so, um, I’m just as likely to be drawing a napkin sketch of an automation cell as I am 3d modeling it in, um, like a CAD software. Yeah. I feel like people like you are kind of unicorns, um, to be totally like balanced and specialized in both areas. I don’t know, Eric, maybe you and I are more on the kind of dreaming up stuff side of things, or maybe that’s just me, but then it’s like, all right, now I got to figure this out and build it with my hands. Right. Right. I’ve always, you know, dream it up into, into reality and then hand it off to somebody who’s actually qualified to make it work, right? That’s awesome. So Katie sounds like you can do both. But when I hear the name Archimed, um, medical, I’m just assuming from the med in the, in the word. Yeah. We’re a surgical case and tray manufacturer. So when a surgeon has to perform an operation, they need specific hardware and tools and instruments in order to do that.

[5:26] And they need a means of storing all those and transferring them from hospital storage to the operating room. And so we put them in these nice little boxes called surgical cases and trays. And so we designed and manufacture those here at Archimed. We’re a smaller company, about 80 people. And we have two big sides of shop. We have our fabrication side, which looks at the sheet metal cutting and forming of our cases and trays. And then we also have a machining side, which is using a CNC mills to make our surgical caddies out of poly plastic. And these caddies are holding individual screws and bone plates and things like that. But, um, basically just a means of transporting all these important instruments, um, that medical OEMs are going to need for their products. I’m excited for our listeners, because when you say 80 employees, we’re really telling that story that like automation, right, uh, advanced manufacturing is attainable, right? Anybody has access to it. Right. So, um, saying you’re a small, you don’t have to be GM, right. Um, and, and, and what we’ll talk about later is the plastic manufacturing. That’s what I love too. Cause now we’re taking a pretty non-standard material that can be difficult to machine.

[6:43] It can, it can, uh, deform, you know, if the right work holding isn’t used. So the fact that you’ve automated that is even cooler, you know? Yeah. Yeah. It reminds me a bit of our business where we have Henig, which is more on the fabricated, like kind of sheet metal type, uh, manufacturing and then advanced machine and engineering, where it’s all CNC machines and lots of milling and turning and things like that. So it seems like you’ve got a blend there, Katie, but are you machining metals too, or is it just machining plastics and then doing more of the fabricated metals? Yeah, it’s more rare that we machine metals. I mean, occasionally we’ll get involved with aluminum or 17-4, but truthfully most metal processing is done on our fabrication side with stainless steel and things like that, but, um, we’ve been able to automate both sides of shop. And so that’s been really exciting for us. Yeah. I can’t, I can’t wait to get into that, but you know, this is just a crazy topic and your background is crazy for me because we’ve got this long time family friend and, you know, he’s like, I know the Goldner’s, they’re manufacturers and his dad’s a podiatrist and he sells medical equipment and he’s always like, why don’t you guys make this stuff? You guys make stuff out of metal. And he’s sending me pictures of trays like all week, this week and last week, Eric, you know, Ryan Bush, he’s like, you guys could do this. And I’m like, yeah, there’s probably a lot to it. There’s probably lots of regulations.

[8:04] He’s sending me these pictures of these like screws too, that you were mentioning. And so, uh, I’m like, we, just cause we can, doesn’t mean we should. And it doesn’t mean it’s getting easy. There’s a lot to this stuff, even if it looks like a tray or something relatively simple. So I’m excited to get into that and hear a bit about your production and why ultimately you decided to add automation to it. Yeah. I think that’s a great segue into the next question, right? Is why did automation become important to Archimed, Katie? Like what, what was the challenge, the immediate challenge that you guys were trying to solve? Well, being an 80 person company, a lot of people often hear a lot of hats. And then we’re, we are the worst case scenario for automation. We are extremely high mix and low volume. So our quantities are anywhere from five parts to 20 with our biggest jobs being no more than maybe a hundred. And with that, we’re constantly utilizing operators to set up machines, to do first article inspections, to support creating fixtures for these custom pieces. And so as a result, our operators were stuck standing at machines, waiting to cycle them and either they could let the machine sit and do their other responsibilities or they could stand and let those other responsibilities fall to the wayside.

[9:21] And so we immediately saw a need for a relatively low skill of loading and unloading a machine. And we needed a means to automate that while still accommodating the high amount of changeover and low quantities that we’re facing. I’m glad that that’s what we’re going to talk about because that’s the, I mean, we, we talk about it all the time, Eric, like I don’t have the production volume, I think you even put it in your stupid Jeopardy show as one of the questions, like, did that, where it’s like, Oh, I don’t know the show you speak of, I know a very intelligent, witty Jeopardy show that we did recently. Eric Trebek. So I want to get into this automation project, but I think maybe to help our listeners get their bearings, like what was your role in this? So, uh, I’ve been in your shop. I know you guys use some Haas 5-axis machine tools. Um, like you described, you were having, you know, a lot of setup issues, high mix environment, which I think I love, like Nick said, talking about that. Cause that’s another topic that just, you’ve hit your three for three right now on, uh, killing like, uh, common industry misconceptions on automation, right? That it’s not for small shops, that it can’t be used on like non-standard materials and that it doesn’t work in high mix and you’ve made it work in all those scenarios. So what was your role in this project? And then I really want to get in a little deeper into the problems that triggered the need for automation.

[10:53] Absolutely. So I’m part of a department and there’s only one other person in the department where the continuous improvement team, and we bring in that advanced automation. And so I can’t take all the credit. My, my boss, Aaron Cain has certainly been heavily involved and is oftentimes the brainchild of a lot of these automation identifications and executions. So, um, want to give credit where credit is due. Shout out to Aaron. I hope he’s doing well. Yep. I love that the continuous improvement team is also the automation team. I think that’s really cool. Oh, we wear a lot of hats on any given day. I’ll be, you know, programming an advanced robot. I’ll be making a sticker for a 5s shadow board. I’ll be writing some AI scripting for, to support our finance team. We wear a lot of hats, but that’s, that’s part of the fun. Um, so with that being said, we are involved in every single aspect and I was involved in every single aspect of these deployments, so I have the poorly drawn sketches in my notebook all the way to the 3d models, the failed 3d prints that prototype that I have the PVC pipe. I made the cells out of in the cardboard. I taped to it. And I was also the one doing the manual machining. Aaron helped me with a lot of the welding things. Um, we’re assembling the stuff out of extrusion or whatever components, and then down to drilling the holes in the concrete to mount the robot to programming it.

[12:21] We really pride ourselves in not having to utilize any outside integrators because it gives us control over the cost, the process, and also our ability to expand the process or modify it as needed. So we really do it all, which is probably unbelievable, but also hopefully a, a good story for anyone hoping to kind of step into this. If they’re a smaller or midsize manufacturer. It’s an amazing story. Yeah, no, absolutely. That is so cool. So you talked about high mix production, set up time. Were there any other challenges that kind of triggered, you know, what was the main problem, operational challenges that you guys were trying to solve with, um, this automation cell? It’s what every shop wants. It’s more capacity and utilization on their machines. And so we’re running skeleton shifts in the evenings. We don’t have a shift that runs on the weekends, but we had parts to run. And so it’s, how can we, without adding costs of labor, how can we possibly run more parts, especially with our lower quantities? We knew that we weren’t going to get necessarily bigger jobs, but how could we make the time with our existing jobs more efficient or run more jobs through the shop in the least amount of time? And so that really brought about our need for automation.

[13:42] And the other thing that was really nice was right around the time in 2020, when we started to tackle this journey, um, the Indiana governor, uh, Eric Holcomb was, uh, really pushing for manufacturers to explore advanced technology and gave us the manufacturing readiness grant that helped us acquire our first two collaborative robots. And that really helped kind of set us on a route of success because we knew that we’re fully committed. Now, if we have two robots sitting in the shop, we’ve got to do something with them and really utilize them for good. Okay. I’ve got, I’ve got two questions while you’re talking. Um, so first let’s just talk about that grant for a minute. Um, I think that people need to really understand, especially with automation, but really any form of, uh, manufacturing equipment or, or any initiative, there’s probably a grant that can help you pay for it. Something Eric and I talk about a lot. And, you know, Mike Payne, one of my co-hosts on Making Chips and the host of By the Numbers, he’s always talking about take advantage of the grants. Who within your organization was the one who went and captured that grant and applied it to your business? I think it was a shared responsibility between our CEO, John Asalvo, who is a huge proponent for advanced technology, pushing the envelope, constantly streamlining our processes. And then also, like I mentioned, my boss, Aaron Kane, I think those two together really found the grant, secured the grant, and then also Connexus Indiana, which is a, um, nonprofit here also assisted with the grant process. Is that the MEP in Indiana? It’s not.

[15:19] Purdue MEP is our, is our Indiana representative, but Connexus is heavily involved with advanced manufacturing and logistics here and helps support all those smaller and midsize companies to networking and getting any resources they could possibly need to be successful. Katie, you’re five for five now, just so you know. Now we’ve got grants and we’ve got CEO top-down buy-in. I mean, this is just like, it’s no wonder that this whole automation thing you and Aaron kind of took charge of was such a smashing success, you know, it’s just, everything’s lined up. Hey guys, it’s Nick for Making Chips. And I just wanted to tell you, if you’re looking to push your shop forward this year, you need to be at the DN Solutions Manufacturing Without Limits event. That’s where I’ll be at the DN Solutions Chicago Tech Center, May 5th and 6th. And I’ll be standing next to a lights out monster, a DVF5000 with a Hennig workflow feeding the beast. And they’re running more than 15 other machines live on the floor. Not just demos, live cutting too. Plus you’ll have expert led seminars, automation insights, and the chance to talk directly with engineers and other manufacturers solving the exact same challenges you have. And they’re making it easy.

[16:34] Breakfast and lunch are included both days. So you just show up and focus on learning. If you want to see what’s possible for your shop, this is the place. Register now at the link in the show notes or visit dn-solutions.com. That’s dn-solutions.com. So yeah, back to like the challenge that you were solving as a continuous improvement person, you have your current state of, I think you were mentioning utilization was kind of the target and getting more out of the labor you have there. And then getting more from that when they’re not there and then no one’s tending the machine. And then you set a target. I would imagine that’s how all the continuous improvement people I have around my business. It’s always like, here’s where we’re at. Here’s the target. And how do we close that gap? Can you kind of tell us that aspect of the story? You guys want, you had some sort of a initiative for automating. It became your responsibility. I imagine you probably had some metrics attached to that or some sort of a scoreboard that you were keeping track exactly how successful you were. And if you knew if you were winning or losing, kind of give us some insight into that and how much improvement you were able to create with automation. Yeah, absolutely. So we really measured on how much time a machine can run unattended.

[17:55] So previously with our cycle times of being anywhere from 20 minutes to about two hours, a machine could go 20 minutes or two hours without being interfered with by an operator. And so we really were like, okay, let’s start small. Let’s start, can we go 10 hours without someone having to interact with it? And then once we got there, we continued to iteratively increment and we’re now up to 58 hours where we can run without a single person having to interact with it. So what that means for us is, all right, let’s set up a long running job on a Friday afternoon, let’s send in someone on a Sunday afternoon to maybe put in a couple more parts and then come in Monday morning, the bad boy is still running. You just got to empty out the chips. So, um, that’s, we saw about a 40% increase in our production capacity. And I think there’s a really common misconception with automation. I think people often challenge, well, your robot’s not moving as fast as a human would do it. And you’re absolutely right. We have a press brake robot that takes about 20 seconds for every single, um, single bend it introduces on a bracket. And you’re absolutely right. The operator that’s been here for 30 years can certainly press a pedal and get moving and grooving. But where the gains are is on those off shifts, on the weekends, during lunch breaks, on Easter, on Christmas. And so we started to get all those gains, um, on those billable machining hours when no one had to be there and there was no labor to offset those costs. And so that’s where we really saw those biggest, biggest gains.

[19:30] You’d be amazed. Like, you know, when we, when, when we talk automation systems, it’s one of the first critiques someone will have of like a pallet system. Like, ah, changeover is too slow. And I’m like, it’s happening on Sunday night, dude. You know, that’s, that was my first thought. So Katie, I don’t know if you’re familiar. We were bringing a new product to market at Hennig called the workflow. And what it does is it, it’s a pallet automation system. It automatically loads and unloads like a five axis mill. The way ours does it is it comes in and it swaps the pallets in the machine and comes out. So instead of the robot having to go in, out, in, out, it just comes in, does the swap and you’re off and running. And we think that’s cool. A lot of people think that’s cool. Uh, cause it does that in the machine. It saves, I don’t know what, Eric, 30 seconds, a minute total. Sure. Something like that. And that, but we have some customers and prospects who are like, I only want it if it, if the machine has the, uh, the space inside for it to do that swap in the machine, because I don’t want to lose that, those 30 seconds. And sometimes we look at each other, we’re like, you only run one shift. The biggest problem you have is that you don’t run for 16 hours a day. Not like the 30 second extra time of having to go in, out and in and out.

[20:45] And so it’s just kind of to your point and the way you were thinking, what’s the amount of time that this machine can be making parts with nobody there. And you got to think first in maybe lunch break increments, an hour or two hours or something like that. Then you’ve got to think in, uh, overnight increments, and then you’ve got to think in weekend increments. And it’s so cool that you guys, uh, yeah, you got up to like 58 hours where no one needs to be there loading and unloading the machine. I think that’s incredible. I think something that our listeners really would like to understand a common problem that we see when we automate like something like a tending operation, right, is any potential downstream or upstream bottlenecks, right? So a common example would be, um, you know, I have a part, it’s got two operations, so I’ll run the first op at a different cycle time than the second op, and I’m creating like this stockpile of parts in between the two processes. So how did you and your team go around, go about like identifying any potential bottlenecks created by automating the tending op? Yeah, we very much didn’t want to create just local optimizations. We wanted to create it through the entire process. And I think we are able to do that just naturally through freeing up our operators.

[22:03] Because we don’t have a dedicated operator for every single cell, those operators are now freed up to do other value add tasks that work up and down the flow of our parts. So the best example I’ll draw is we had an 18 year old kid, his name is Jerry. And he sat in front of one of our CNC mills and every single day he would cycle apart, hit go, wait 20 minutes, cycle apart, hit go, wait 20 minutes. And about a year into his job, we put a robot on it. And from there, it was then able to run for about 20 hours unattended. And what that did for Jerry is it freed him up to do other value add tasks that ultimately made the entire process more efficient. So from that, we actually upskilled Jerry into a CNC programmer for that same machine. So now we have value added in the front of the process. And while that machine is running, he’s able to go and now laser etch multiple parts on a different side of shop because he doesn’t have to stand watch. And so those kind of automation things, we’ve turned our resources into more cross-functional, which ultimately brings about value throughout the entire process while not adding any waiver overhead. I love what you said about the, you don’t want to just have local optimization, make one machine better. You have to understand the entire process and ultimately the whole throughput equation.

[23:22] And we say all the time, like if you automate something, you probably just can have a bigger bottleneck before and after it. And that doesn’t necessarily help you get more quality parts out the door every shift or every day. And so the point you made about that operator being able to do more than 10 to the machine and maybe move to the next area where that bottleneck was created and relieve some of that and solve that and kick the bottleneck to some other area, then fix it again. And constantly as you’re doing that, you’re getting more throughput. I think that’s a really good point. Unlocking the human capacity to do that by making sure they’re not just sitting there, staring at a machine and opening and closing a door. And you use, you use the phrase to describe that, that I don’t want it to be lost on our listeners, local optimization. I think that that’s such a cool, such a cool framework. And I want people to remember that when they start considering automating their own manufacturing processes is look beyond local optimization. Wow. Good stuff. There’s plenty of forms of waste. And oftentimes we forget that under utilizing a skilled employee or an employee who has a capacity to gain those skills is one of the biggest forms of waste. I mean, that kid was fully capable of becoming a full fledged three axis CNC programmer and we would have had him just stand in there pushing a button for a couple of years, had we not been able to deploy our automation efforts. She’s like the prototype of the show.

[24:45] Yeah. She’s like the automation superstar. Cause these are all the points we make. It feels like in every episode and, and people don’t connect it to like, it’s automating is caring for your employees. It’s empowering them to be the best and to have the greatest career. And it just totally dispels this idea that like the robots are going to take people’s jobs. Cause look at Jerry now, right? He’s so much more capable. Yeah. Jerry just bought a house. Look at this. Yeah. We see about 15 to 20% wage increases of all of our operators because they’ve been upskilled by the low skill tasks the robots have taken from them. And instead of taking it from them, it actually drives them closer to job satisfaction and more advanced problem solving that is ultimately worth more to the company and the employees. Yeah. And not, not even just compensation and value that they’re creating for the business, like nobody wants to do those jobs anyway, they make you go crazy. If that’s what you have to go do for eight hours a day, just some boring, repetitive, monotonous tasks. Well, it’s a joke. Yeah. Like creating the next serial killer. Right. It’s not healthy. It’s not mentally healthy. No, but you know, everything you said about like, uh, like wages going up and like, I mean, this is just such powerful stuff, Katie, I’m so glad that we have you on, uh, and, uh, I, I mean, I’ll say it now. I think you’re seven for seven, but who’s counting.

[26:07] I think the next thing I really want to touch on is there’s a lot of options out there, especially in the world of just machine tool tending, right? You’ve got cobots on a cart and cobots attached to decks out in front of the machine, and you’ve got pallet systems and all these options, but Archimed decides to tackle it, do it yourself. So I’m curious what the decision was behind that. And, uh, yeah, I’m curious what the decision was behind that. So many people just go apply like an integrator to it and say, we want to automate. What would you do? Um, you guys, I mean, most people don’t have a Katie and they buy, they buy something for a hundred grand and they plug it in and the cobots running off the pendant and they’re just, you know, they’ll look 37 raw pieces of stock on a, on a cart or on a deck and go to town. Yeah, absolutely. I mean, you can do that. And we, I, our best example is that press break robot. I mentioned earlier, uh, you can buy a turnkey system for that for about 500,000 and, uh, excluding the robot, we spent about 20 K on the rest of the system. So the robot, you know, was another couple thousand, but, um, keeps the price down was a decision to do it ourselves, but I think the bigger influence was actually no matter what turnkey system we were going to get. We are not the perfect case for automation. And so we were going to have to customize it and tailor it to our own needs, which was going to add additional costs, labor and resources.

[27:39] So we saw the systems and yes, they’re very clever and they’re very advanced, but we thought to ourselves, we have the same fundamental and foundational skills we can build up to do the same thing. If you know how to manual machine, if you know how to model, if you know how to do basic circuitry and understand digital inputs and outputs, a lot of times that’s enough to just get you started and get you smart on that process. And we had that internal confidence that no, we could pull this off. Like, yeah, there’s lots of bells and whistles available, but for what we’re doing, honestly, like if we just had a bookshelf with some parts on it, it might be okay and we can kind of finesse it to work. And of course, in due time we get better and we do purchase some bells and whistles to help us, but ultimately the decision to make such a custom solution for our very unique type of work suited us better than purchasing a generalized one that we would then have to tailor or sacrifice different parts that we could potentially run on it. So Eric, I’m trying to visualize like what she built and you were there. I think you were, you played a part in it. That’s why Katie’s here. You know her from your past. Tiny, tiny part. Yeah. Yeah. Obviously Katie’s got a lot of talent and her team has a lot of talent and I want, I want you guys to describe like what the solution was, because it sounds like they didn’t just go buy a product, like you said, pay a hundred grand and plug it in and it works. They had to kind of construct it and find all the components of it and put it together and figure it out. So Katie, can you paint a picture of like what the end solution became and how you got there? Can maybe walk us through that journey of how you put those pieces together? Yeah, absolutely.

[29:16] So we started off with kind of a hokey bookshelf, like I mentioned in the previous section. And so it was made out of extrusion and it held exactly five parts. So it ran for 10 hours total and we were sourcing all the individual electronic and pneumatic components. So things like solenoids, relays and then the big things that really kind of tiered us towards the ability to fully do it was a pneumatic chuck and vice system that allowed us to remotely grab onto the part and secure it and then cycle the machine. But after that initial deployment of that hokey bookshelf, we were so early in the automation process that a lot of times we’re moving parallel to other vendors who had some of those products that were going to assist us available. And so around the same time we were looking to step up that system, we discovered a fifth axis shelf. And with that, it had the capacity to hold 28 different parts. And it was a lot more secure and structurally sound than our hokey bookshelf. And so as a result, we optimized that process, integrated that in and were able to really utilize a vendor to step up our capacity from that 10 hours to that 56, 58 range.

[30:36] That’s such a beautiful jump in productivity. I just whenever whenever you say that number, I just I get a smile ear to ear because it’s just I mean, what a dream outcome. Right.

And so, yeah, that’s how that’s how we met, Katie, when you opted for those fifth axis automation components. And I think another interesting thing to touch on is, you know, you did opt for just vice loading, you know, integrated fixture exchange is something we talk about for for an automation option on this podcast, often most commonly in the form of like a pallet system. And you guys opted for loading the vices directly.

[31:17] Can you go into any detail on what went into that decision? Absolutely. We have an incredible amount of variance between our parts. So we have parts that are as small as something like an engagement ring box and then as big as like two shoe boxes put together.

[31:32] And so with that, we needed a common feature to be able to run the robot. Our goal is always just one robot program that accommodates all of our wide breadth of parts. And the common feature we found was actually the vice that the part was able to sit in.

And so instead of having to worry about and dynamically adjust for all these different parts of all these different shapes and sizes, we instead just grab on to the vice itself. We load that into the machine, have some fail proof mechanisms in there. We spiral into a spot so that we guarantee that it’s going to hit the entrance holes that it needs to.

And really, it created a common feature amongst a lot of uncommon parts that allowed us to automate it. And that’s a reoccurring theme in all of our automation projects is identifying a feature on a part or that a part can be made it up to that. Ultimately, you can automate from the other example.

[32:31] And we talked about a little bit offline was our press brake robot. We have all these single band brackets that are wildly different widths, lengths, shapes. But the thing they have in common is they all get a single flange that is all a half inch.

And with that one common feature, we’re able to cleverly present the parts, grab the parts and then use that as our source of truth in order to make a consistent and repeatable process, despite the inconsistency from part to part. Oh, it’s so good. Yeah, there is some elements of standardization that’s always required with automation.

[33:08] Can’t be different every time. You know, I don’t know about some of these robots that are coming up. Maybe they’ll be able to figure it out. But what we have today, you have to find some common elements, some common standard, like you said, to just kind of consistently present the part to the machine or in the fixture or however you’re you’re automating. And I think that’s so cool how you guys did that across a wide range of parts. Yeah, in manufacturing, growth isn’t accidental. It’s engineered. But a lot of suppliers are still building their pipeline on outdated lists, inconsistent follow up and hope based marketing, hoping their website will somehow bring in leads. Gabe Draper, the founder of Factor, knows that struggle. Years ago, he ran a shop that went bankrupt, not because of quality, but because he didn’t have a steady way to generate new opportunities. That experience is what led him to build Factor. Factor helps suppliers create consistent sales opportunities by combining targeted outreach with marketing that puts you in front of the right buyers.

[34:11] Their team finds the right companies, starts real conversations and sets qualified meetings. So if your pipeline swings with referrals or seasonality, Factor brings a system that puts you back in control. And for listeners of this show, Factor is offering a free custom report that highlights the best opportunities in your industry, plus a list of real companies that match what you do.

Request yours at factormfg.com slash chips. That’s F-A-C-T-U-R, M-F-G dot com slash chips. So now you’ve got you got the CEO buy in, you’ve decided we’re going to do it ourselves, right? You source all the hardware, it all shows up.

[34:53] And now it’s time for the team to go to work and implement. I’m sure that process was just smooth. There were no challenges.

[35:01] You guys ran into no headwinds, anything like that. Right. Can you walk us through some of those? Yeah.

Did it work the way Eric described or it was like, hey, let’s we think this is how we’ll automate. Let’s buy all the components and then put them together. Or were you were you deciding on components in the process of figuring it out and maybe making some different choices? Oh, 100 percent.

It’s an iterative design. Where we start is never where we quite end. There are always components that you’re going to need for almost every project.

[35:32] You’re probably going to need some solenoids. You’re going to need some relays in order to, you know, control error and look at those digital inputs and outputs. But really, it’s an iterative process. So as we would implement something, we’d see a gap we need to identify. So, for example, OK, great. This robot’s able to put it almost in the machine, but it reaches a singularity. So we need to find a way to get it to the last two inches or we need to find a way to make sure it has actually properly secured. OK, we need to buy an additional whatever laser component to make sure or a pressure check to make sure that everything is seated properly. So it very much is an iterative design process, which not every company is a fan of. But our leadership team really subscribed to and they recognize that off the bat, you’re probably not going to get it perfect. But if you’re getting at least 80 percent progress and you’re able to encompass 80 percent of the parts, the use cases and the value that we need from it, then that is plenty insufficient for us because it’s still moving the needle. And there’s opportunities to expand that in the future. I mean, that’s why we are continuous improvement. I love how you said that. And my brother, he’s our president, but he came from continuous improvement and he always says continuous improvement needs continuous improvement like that, that group and those people need to be continuously improving.

[36:58] And as you’re talking, I’m realizing, you know, you and your team, you have more than a base level understanding of automation, you know, your solenoids and what component to use and this and that. How did you get that? Did you is that something you got from school? Is that something you got from experience on the job? Was that something one of your team members was really strong in? Yeah. Is that on the job learning? Yeah.

Yeah, it was definitely on the job learning. The credit goes to, again, my boss, Aaron. He was really had a lot of those fundamental elements kind of mastered.

[37:32] He had worked with solenoids and relays and things. And oftentimes we were just kind of learning as we went to like we had a lot of baseline understanding. I have a programming background that was really helpful. He had he’s really mechanically inclined, which was really helpful. And so we had these fundamental elements, but we’re just really good problem solvers. We didn’t know what the difference between NPN and PNP was for a relay like, but we know what to Google and we know how to read documentation. And truly, if you have that confidence to say, OK, like the humility to say, I don’t know this probably, but I’m confident I can figure it out as I go. We did a lot of trial by fire, but now we do know those things and it makes us that much more effective when we do a new project. So lots and lots of learning on the go. Lots of mistakes we made at the start that we wouldn’t do by hollow extrusion. And you realize you need the strength of non-hollow extrusion. There’s so many nuances that you won’t know until you just step into it. And you can make almost anything work, but there are certainly more optimal ways. And it all relies on those foundational elements, those basics of circuitry, the basics of pneumatics, of mechanical truths, of knowing that gravity is always free and things like that. So, yeah, and I think what I’m inspired by when I listen to your company’s story is they created the capacity for this experimentation.

[38:55] You know, so many shops, Eric, we see it all the time. Like every person has to run a machine. They’re all probably wearing two or three hats and nobody has enough time, resource, capacity to iterate through to a solution like this. So you’re always just like barely breathing. And I think they obviously created an environment where you and your team could figure this out. I guess the alternative to that is you spend the half million dollars.

[39:21] You buy like the full turnkey project that you bolt up to the machine or you hire an integrator to come and teach you and show you how to do it and set it up for you and then just say, OK, push this button, then push that button. But I think the lesson is don’t rob the business, don’t rob the team of the ability to iterate through that and learn everything along the way, because now you have that resource employed. Right.

You know, and what’s the next implementation look like? Yeah, I mean, I’m not saying it’s all one way or the other. Sometimes you want to bring in an expert and say, hey, we don’t have anyone here who knows circuitry or knows the first thing about any of this. So come on in and get us set up for us and teach us.

[40:01] But I think your DIY approach obviously taught you a lot. You’re like Eric said, on the next one, you’re going to be even stronger than you were on the first one. And that’s just going to continue to build and build and build. And now your business has a bit of a superpower. Is that how you see it? Yeah, I think so. I think a lot of people think these systems are wildly complex, but in reality, it’s just some fundamental learnings that have compounded over time and really brought about the complexity of the systems. But it’s not unfeasible to dedicate resources to learn little aspects or tackling a simple automation project or a small thing that is a relatively robust process with a really specific part and kind of starting that learning process, keeping those resources internal. Because you know what? If a robot misloads tomorrow, you’re going to have to call that integrator, pay them to come in. And that cost is adding on top of what you’ve already invested in that turnkey solution, whereas I’m already here working, might as well stop by and identify what’s wrong or try to solve it or optimize it and things like that. So having that resource internally really is continued cost savings over time. And it’s a wise investment for our company. Yeah, I think what’s remarkable in that bit of a story is that, you know, we always talk about the big wins of automation and they’re usually related to productivity and utilization and manpower and all that.

[41:30] But I mean, really, what’s cool about the Archimedes story is with their automation journey, they also get like this whole team of automation experts, you know, I mean, I think that’s brilliant. And just add that on top of the pile of all the benefits of automating a process, whether it’s, you know, the simple tending of a machine tool or something more advanced like what you’re doing on your press break. And I think that that’s a phenomenal, you know, segue into the next questions, which would be the impact on the business, the impact on Archimedes, because I know that you guys monitor your spindles, right? And so I know what the values are. Nick doesn’t, but I think this would be another cool opportunity for you to flex a little bit, Katie. I think that your spindle utilization numbers before and after were, I remember them being pretty remarkable. Would you be OK sharing those? Yeah, I think so. I’ll have to speak in approximations, but I think we were, you know, let’s say for a full 24 hours rather than monitoring on a shift, we were probably around 50 ish percent because of the amount of setup time. And then when you think about downtime in night shifts or weekends and now where we approach 80 and still that you’re never going to eradicate the setup times and things like that. But truly, we’re getting, you know, close to 18 to 24 more hours of runtime unmanned that we wouldn’t have otherwise.

[43:06] And so that’s really where we’re sourcing a lot of that value. That’s awesome. Oh, my gosh.

So like a couple of things you said there that I love and that we say, if you if you have a 10 hour shift at your shop and you run eight hours, you are not at 80 percent utilization. There’s 24 hours in the day. And so I love that you said, hey, in a in a 24 hour day, we got to 80 percent.

[43:29] It’s a lot more than 10 hours. And so I think that’s the point we always try to make is like these machines, they run all the time. They don’t need to go to bed, you know, they don’t need to go home and make dinner for the kids like you do. So hopefully while you’re doing those things, your machines are still working as much as possible. And it’s really cool that the first thing you said is, yeah, let’s talk about a 24 hour day. Not like so we see some shops like, yeah, we’re 80 percent utilization for how many hours are you thinking about the whole day? So I really, I really love that. All right. Shop owners, listen up here. I think all of us know we are in one of the most in capital intensive businesses there is and margins are tight and we’re always fighting issues. And then when we make money, what do we do with it? Stock market’s up and down all over the place. Well, I get asked a lot about where can I invest my money? I’m a firm believer as business owners, you’ve got to invest in yourself first. You need to reinvest capital into yourself.

[44:26] So the number one thing that I’ve done in this industry investing was ProShop ERP. We’ve been using ProShop since 2018 and my payback literally was within weeks with their per user based pricing. I knew when I made the decision, I only needed to save my workers roughly one hour per month.

We saw within the first week we were saving multiple hours per day. That type of ROI just doesn’t come along very often. Outside of that, the real value has been able to quickly implement best practices across my entire shop from the time we get an order to the time we ship it, scaling my business with confidence, being able to go out and know that I can deliver on my promise to deliver quality parts to my customer, and then gaining insights into every job, employee, machine, and every process in our business to really drive efficiency and look for opportunities for improvement.

[45:23] That’s where the real value in your business drives. So visit ProShopERP.com, book a demo and start realizing ROI today. That’s ProShopERP.com. So I kind of want to get into what the future of automation looks like at Archimed, but I kind of, I don’t want to, to miss some of the amazing things that have already kind of come your way from automating, right? I think Jerry’s story is remarkable. I think even your journey and Aaron’s journey is just an outstanding story. Uh, and you know, the fact that the, that your, uh, your wages have gone up and that your, your, your people are doing way more value added tasks for the business and people are buying houses and like, that is an amazing manufacturing story. So what does the future of automation look like for Archimed? Uh, any, any cool, exciting plans coming down the pipeline? We, the boring answer is, oh, we want to duplicate some of our existing systems. Uh, we have a couple of unautomated CNC mills and a couple of press breaks, but, um, the two more fun answers are even right now today, right before this, this podcast, I was working on expanding our press break. So like I alluded to earlier, our press break, um, automated, so only does single bend brackets. And at the time that was perfect, that has given us so much utilization and capacity, but we’re continuously improving.

[46:54] And so we want to start to encapsulate multi-bend brackets. So lately I have been working on the process for how to introduce multiple bends into some of our stock parts that we make 50,000 of a year. And so, um, using specialized dyes and things like that to create multi-bends on the same exact automated cell with modifying as little as possible so that we can start to get that around the clock value from that press break as well. Um, and then if you’re looking at long future state, um, vision, we are right on the cusp, the technological cusp of being able to utilize vision and artificial intelligence to dynamically, you know, update robot positioning and actions, but we’re not quite there yet. And so we’ve been closely monitoring our ability to integrate those types of things with our existing systems or new systems. We, um, we’d love for the ability to do 3d bin picking where you don’t have to masterfully orient a part. You can just pull it out of a bin, no matter how it’s sitting in there. We want to be able to dynamically see if a part is seated correctly for safety and quality checks and things like that. So we’re right on the cusp of that being ready commercially.

[48:09] Um, but we’re not quite there yet. So we’re, we’re biding our time, but we’re, we’re certainly at a lot of trade shows evaluating what’s new on the market and how we’re going to bring those into our shop. That’s so cool. So my, my big takeaway for the whole episode is like, you’re like, Oh, the boring thing is we just want to replicate what we already did. But the exciting stuff is this vision plus AI and self-correcting robots and things like that. Well, the boring stuff isn’t boring to the CEO who now it’s already figured out. The replication is going to be so much easier than the first time. And because it’s already figured out on more of the simple, simple application of automation, it, it, you’re onto the next thing. You’re onto the AI vision stuff that is just absolutely mind blowing. So you have to do the boring stuff. And, and the first time you do the boring stuff, it’s exciting. And then what becomes exciting is so far beyond that the boring stuff becomes table stakes. So yeah, we already know how to automate that. We’ll just replicate that and then we’ll go out and tackle the next thing. And what a cool story. And it really speaks a lot to the business that you have and, and to you personally and to your team. It’s really impressive. And thank you so much for inspiring us for this hour of the Lights Out podcast.

[49:25] Thank you. You’re not getting off that easy, Katie. Our listeners at the end of every episode, we like to do like this three tactics segment, right? Where, uh, but I want to do it a little bit differently, right? I think for our listeners, what advice would you give them if they’re considering DIY Cobot automation? I think the biggest thing you can do, and I talked about a little bit earlier was the 80-20 rule. As engineers and as people looking to automate in many factories, you want it to be right a hundred percent of the time you are always striving for perfection and sometimes an automation that can actually result in paralysis or a Cobot sitting. And so I would challenge these shops to explore the 80-20 rule where you’re encapsulating 80% of the use cases, 80% of the parts, the odds that you have a perfect process where you make 10 million of the same part in the same way on the same machine every time. It’s pretty slim. And so really, if you are able to capture 80% of those use cases and yeah, you’re going to have oddball parts that are a little bit too small or a little bit too big. That’s okay. Those can still be done with the manual processes, but think about that 80% you’re offloading. So I think that would be my biggest advice. And secondly, it would be to make sure your process is robust. It’s pretty common for us to want to tackle a process and then realize that, well, our operators, you know, they kind of shift the part around just a little bit before it goes into the machine.

[50:57] So sometimes you’re going to have to modify or reinforce those bulletproof processes to make sure that a robot can consistently do what it needs to do without that human interference. And we’ve had to learn that lesson time and time again of we really need to define this process. We need to put in some fail-safes or some guardrails in order to make sure there’s no human innate compensation that of course is going to happen in a process. But how do you prevent against that or at least build it in with that expectation when you’re automating? I love that. We say it all the time, you cannot automate an unreliable process. And you don’t even realize it’s unreliable. Like you said, innate human. Yeah. We just adjust it a little bit. Well, the robot does exactly what it’s told. It doesn’t just adjust it a little bit. So you have to eliminate all of that play in the process and you just got to, you know, make sure it’s totally dialed in and stabilized. And that’s, that comes from somebody who’s been there and done that and seeing what can happen when, yeah, we just adjust a little bit or you just set it in there kind of like that. No, you can’t do that with automation. You got to use just the right length cheater bar to tighten that vice properly. So IMTS here. I mean, you already talked about some cool things you guys are getting into. I’m assuming you’re going to go to IMTS.

[52:16] What are you most excited to see for advanced manufacturing techniques? I mean, anything that is getting, getting your blood pumping, getting you excited. I think the obvious answer is the artificial intelligence and integrating that into different things and always evaluating what tools we can use to bolster our robot system. So smart grippers, vision systems, cameras, in-feed systems that are more advanced than our homemade ones. And so we bring a whole team with us to IMTS. We usually take 20 guys on a bus and it’s an absolute blast. And we basically just push them all to Chicago and say, come back with some great ideas. And it bolsters that culture of improvement beyond just the continuous improvement department. It makes an entire company that’s poised to want to bring about change and process improvement. So we’re, we always welcome the opportunity to be at IMTS to be on the cutting edge and the forefront of all technology, whether it’s machining, fabrication, or automation. I mean, every part of it, we’re, we’re just big nerds and we, we love it all. Nick, don’t you want to just go work at Archimed? I kind of do. And like I say nerds affectionately about all of the smart people at our company.

[53:35] I’m like, Oh, we got to get that nerd over here. You know what I mean? It like he’s smart. So I love that Katie. And yeah, thank you so much for, for joining and for our audience. I always say the three Rs, please rate this, please review it and please refer it to somebody. I mean, Katie is such an inspiring story and there’s so many takeaways from this episode. If you’re like me and you’re listening to everything she’s saying and wanting to share it with somebody, then just do that, go ahead and share the episode. And, uh, we appreciate that. That’s how we get more listeners to the Lights Out podcast. And we tell more stories like Katie’s. So thank you so much for joining us. Yeah. Katie, thank you so much. Knew you’d come on here and hit a home run and, uh, what a great episode. And, uh, I mean, almost gonna, gonna have to have you back here in a little bit, just to, just to hear more about how your journey is going. This AI vision system she’s got going on. Outstanding. Well, thank you so much, Katie. And with that, we’ll turn the lights off on this episode. We’ll see you, Katie.