Become a Creator today!Start creating today - Share your story with the world!
Start for free
00:00:00
00:00:01
#357 Override Process Sheets and Process Bins
436 Plays1 year ago
TOPICS:
- override process sheets and process bins
- Kids and work
- Days Off In The Shop is a failure
- Gridfinity!
- Grimsmo's erowa and Speedio update
- Automation and pallet changing
- Hardmilling
Recommended
Transcript
Introduction and Podcast Overview
00:00:00
Speaker
Good morning and welcome to the business of machining episode 357. My name is John Grimsmo. My name is John Sodders. And this is the weekly podcast where John and John talk about their manufacturing journeys and kind of the ups and downs and ins and outs and, you know, day to day, um, it challenges that we face with our ever growing businesses. Agreed. How are you doing?
00:00:28
Speaker
I'm doing good. I guess I'll go first. Um, basically busy, lots of projects on the go, lots of, um, steady.
Production Challenges and Customer Demands
00:00:38
Speaker
I'm trying to make everything predictable and steady so that we can grow from there. Um, it's kind of a big, it's always been a focus of mine. It's like consistency and reliability and, uh, you know, trust the machines, trust the process coming into the current still running in the morning. Um, we've spent a lot of effort into those kinds of things so that
00:00:58
Speaker
we can just produce. In a way, it's kind of bit us in the butt because we're producing relatively the same thing over and over again and the customers want variety. Got it. Right. So it's like from an internal perspective, it's like, sweet, we're crushing, we're doing good, we're making lots of parts. And then the customer is like, I want different, I want something else, I want new, I want new product, new things like that. So we have to find that balance between
00:01:22
Speaker
really good consistent predictability and like add new variables often enough. So I'm really trying to spread myself to other people to accomplish those kind of goals. So I'm not just, you know, all the new is on my plate kind of thing, which is fun. And I want it all to be on my plate, but it's not sustainable.
Gen 3 Mod Vice Updates and Process Improvements
00:01:44
Speaker
Yeah. Amen to that.
00:01:47
Speaker
We've learned this, we're better off where we're at in terms of people and equipment, but it's also, we're victims of our own progress in terms of releasing the Gen 3 mod vice because we now know what we want it to do and we know how we want fixtures to be run and the process bins and the dimensional drawings and that
00:02:09
Speaker
It's not that that stuff slows you down. It just doesn't speed you up because it makes you, you know, we are, it's not just John in his garage and I'm going to sell two mod vices next month and doesn't matter. Yeah. Like I can just kind of make one at a time, tweak it. Like, and, uh, I'm sure folks know the story. It's just when you live it, it's different. So we are.
00:02:28
Speaker
makes me really appreciate it's kind of that like catchy or kitschy saying of like, you know, we're not in the business of making parts. We're in the business of being in the business of making parts. I'm not, you know, I have to think about all of that. So it's actually been a great past week for us because we kind of just finally crossed that for us the threshold on Gen 3, which is
00:02:52
Speaker
We filmed the instructional video, we put all the digital assets up like the PDFs, instructions, drawings, sub-assembly stuff, both internal and public facing stuff. And we have all the process bins and inventory bins, not maybe all of them, like most of them are there, plenty enough for us to start running with it. And so now it's just, actually, coincidentally, this morning more foam showed up for our new boxes.
00:03:16
Speaker
And so now it's probably just another few weeks of just building up some inventory so that we've got our buffer. And then I guess we'll flip a switch. I didn't really talk about that yet, but like, you know, flip a switch on having them available.
00:03:35
Speaker
I can't emphasize it enough. It's just so much different with a team of people and the horizontal runs and makes these things overnight and then somebody else is QC them and then all that. Yep. Yep. Just slowly back away and you don't really realize it. And then you're like, actually, I'm not responsible for that anymore or that anymore. And I mean, I felt bad. I had to use the phrase.
00:03:56
Speaker
A couple of days ago to Skye, I was like, sorry, not my department. I felt rude saying it, but however true it might have been, I don't know the answer to that question. You're asking the wrong person.
Team Management and Preventing Burnout
00:04:10
Speaker
What was the question? It was, are soft blades going to run tonight so that I can know if I'm going to heat treat them in the morning? Sure. I'm like, I'm working on hard blades right now. I have no idea. I'm the wrong person to ask. Yeah.
00:04:24
Speaker
Not to like brush you off or anything, but Angela's your man there. Well, so we had a good one of those things that kind of felt right breakthrough moment. As silly as it is, this is the stuff that I get excited about. So I mean, I can't imagine anybody listening to this podcast doesn't know what a process bin is, but it has the part label on it. It's gray, green tape around it. And then in it, we found that there's kind of three main things. Number one would be a laminated dimensional drawing. Number two is for products, by the way, not like for the air compressors.
00:04:54
Speaker
limited dimensional drawing override sheets, which we just recently started with that. So
00:05:02
Speaker
Take, for example, the half inch dimension on a product that we consider the most critical. And one of those probably could be ground, but we're able to hold it based on some ways that we run the Akuma. First off, it's a really good machine. The thermal, how warm up is it, how we run a custom touch off program that finds the high insert. We look at the wear of the inserts, blah, blah, blah. So we can hold.
00:05:25
Speaker
I don't even know what the outside my head.
Maintaining Manufacturing Precision
00:05:28
Speaker
I think the trash like scrap range is plus or minus three or four tenths. But the target range is plus minus a few tenths. So the question is, how do you cop that?
00:05:41
Speaker
There's three different ways. We can argue about this, but there's three different ways. You can, so this obviously could be for a length feature or a width feature, meaning axial or radial, but there's three different main ways. Number one would be you lie to the machine by adjusting the offset, you know, the G64. Number two would be you lie to the machine about the tool length. Length you can always adjust in the controller.
00:06:06
Speaker
The diameter of the tool doesn't really do anything unless you're using wear or cut or comp, which generally speaking we don't. Or the third is fusions, axial or radial stock to leave.
00:06:16
Speaker
And this is really important because number one, I view any three of those as quite significant and sacred changes. So we do have some tools that share that are made, like a face mill that's really expensive that is used across a few different products. So you can't change it and not realize it has consequences. And then similarly, some fixtures are actually be able to share across different, similar but different part families.
00:06:42
Speaker
So, you need to know if you're going to bump it by a few tenths. Number one, why are we bumping it? If it's carbide, coating wear, or something, fine. That I understand. But if the fixture moved, hold on. That's not good. Because generally speaking, I want to be able to reprobe a fixture if we want to and not have a whole, oh my gosh, we're chasing our tail on getting back to it.
00:07:02
Speaker
So we've done that, it works great, but the two sort of problems with it were number one, I realized it was stressful to find the process bin. I'll come back to that. And number two, once you found it, there was four or five sheets of paper in it and I wanted this one.
00:07:20
Speaker
We now, we bought yellow printer paper off Amazon and now those override sheets are printed on bright yellow paper. Nice. And makes it super easy and you can now see them. And then the process bins, instead of having a wire rack with all of the mod vise process bins on it, which John Saunders thought that this was a great idea four months ago, like, oh, that'll be right there. Anytime you need a mod vise process bin, you just go over to that rack. Well, it's a pain in the butt because you've got to now go to that rack and you've got to
00:07:48
Speaker
find the one on it, even though you would think that that's easy because there's only 15 bins on it and you've got the part number. It's just not. So we realized, wait a minute here. The new wire racks now have one shelf per part and the first bin is the process bin and the next three are our inventory bins, which we're working on a pull system to where, I don't know exactly what this is going to look like, but basically
Organizational Enhancements and Efficiency
00:08:13
Speaker
The horizontal, which is usually Garrett, will make up to three bins. Serena, whose operations and fulfillment and so forth will grab one, swap one out, which means there's an empty one. It's that pull system of making sure we don't underproduce but also don't overproduce. I like that. So you need more wire racks basically to hold all of the different products.
00:08:38
Speaker
Yeah, more floor space, et cetera. I like that. Not really. A whole row is the process bin and then production and inventory and things like that. Instead of having three products next to each other.
00:08:53
Speaker
with production bins and everywhere else and all different places. It's cool. Simple. Yeah. You can't not find the process bin because when you're putting the parts away or whatever, they're right there. Yeah, I like that too.
00:09:09
Speaker
It's like one of those easy moments when you realize the bootstrapper needs to get put to bed and you need to bring out the next generation leader. Go buy these wire racks. They're 100 bucks from Sam's Club. And then they come with five or six shelves. Well, I want 10 of them maybe. So you end up buying two wire racks and throwing away half of one because you want the shelf density on what? That's fine. Don't worry about it. It gets you what you want done.
Balancing Work, Family, and Personal Projects
00:09:40
Speaker
cool. Yeah. Anyway, that was a long-winded answer to my, it feels good to cross that threshold of like, you know, bringing a product to market now where you have the same thing, right? You're buying material and grinding and Angelo and Sky and... Yeah, exactly. I like to see you excited. We can nerd out about this stuff. This is a safe space, you know.
00:10:09
Speaker
Okay, speaking about nerding, I don't know if I haven't noticed this before, like those listening, John and I are on private video with each other. There's a forklift picture in the background. Is that new? It's new as of probably Thanksgiving.
00:10:23
Speaker
So it's like two, three months that I haven't noticed it. Is there a story there you want to share or no? I don't know. Yeah, absolutely. So fun fact, I love my kids. I really do. I'm now 10 year old son William and a seven year old daughter Jane. And for no explanation or reason, Jane and I have always joked about forklifts.
00:10:42
Speaker
When I would read to her in bed, I would switch out. Sally would then get on the school bus to go to school. I'd be like, Sally would get on the forklift to go to school. Nice. She's actually growing out of a teasy joke phase like that, but orcas have always been our thing. When she's at the shop, she loves to climb up on it. Now she can ride with me on the weekend when no one's around and we're just doing something.
00:11:06
Speaker
So I come home one day and she has drawn me a shockingly accurate toilet or forklift and signed it. And I was like, Jane, I'm going to frame this. And so I framed it, put it on the wall in the shop. And then she was in the shop a week or two later and she like couldn't believe that I actually framed it. That's awesome. That's so cool. I love that. Yeah.
00:11:25
Speaker
Your kids come to the shop much these days? Not much. Sometimes Claire will join me after hours if I need to pop in. We're hanging out anyway. And I'm like, ah, Kern's down. I got to run to the shop for 10 minutes. And she's like, so an hour is what you mean, right? She knows my tricks. But she just either hangs out with me or hangs out in the lunchroom and just eats our snacks and hangs out. Lafe will come during the day sometimes if he can.
00:11:53
Speaker
He's shy, but he does like to hang out with the boys. He likes the teasing kind of thing, but he also doesn't like it. One of our guys decided to call him a different name every time he came in. He's like, hey Billy, what's up? He didn't like that, but he likes to come out and hang with the boys.
00:12:13
Speaker
We got him trying our gauntlet of hot sauces here in the shop and now he likes hot sauce. That's funny.
Family Influence and Career Reflections
00:12:20
Speaker
Yeah. Because in the fridge, the guys downstairs, they probably got 15 different hot sauces in there. It's a thing. We got a little pizza or something and put a little cup of dabs of some crazy hot sauce on it and this was going to burn your face off and life really like that. That was fun.
00:12:36
Speaker
Have you seen that YouTube series? I think it's become pretty famous. Hot ones? Yes. Yeah. Yeah. So half of those sauces are here in our fridge. Hilarious. And even they've seen episodes of those. So that was fun little things.
00:12:50
Speaker
Yeah, neither one of them really like to come and help and do stuff as much as I would like otherwise, but that's okay. I'm not forcing it. I'm not pushing it. It's here if you want to kind of think. Actually, what I do want to do, I have it on my list for eventual talk about, is I want to have a corner of the shop for Leif's projects. I want to make a little engineering desk for him with, put a soldering iron there, a little voltmeter.
00:13:17
Speaker
He wants to learn how to weld eventually. I need to remember how to weld kind of thing. Little things like that because he's such a budding engineer and not because I'm forcing it, but like he plays a lot of Minecraft and the amount of things he can do in Minecraft.
00:13:34
Speaker
it matches any adults. He goes deep on redstone circuits and make these doors open with pressure plates. There's all these mod packs for Minecraft, like the Create Mod Pack that lets you build so much stuff. Him and I learned how a watch works using Minecraft.
00:13:56
Speaker
how gear trains work and multiply and add speed so that your seconds move in things. And because he was playing with gears in Minecraft and he's like, how do I get faster? I have a big gear and a small gear and as many as I want. And how do I, how do I multiply them and get faster? I was like, let's sit down and figure this out. It's like 6am in the morning. I just wake up. Um, but yeah, we figured out how to stage the gears so that it speeds up. And then he had basically infinite RPM at the end and he's,
00:14:25
Speaker
10, almost 11. Loves it. Anyway, I want to have a little corner of the shop where he can come and just fart around and feel like it's his space.
00:14:37
Speaker
vividly remember, I mean, my grandfather was quite impactful. When I was probably life-saged, he built me a green toolbox. When I say toolbox, it was like a workbox. It was probably four feet long, two feet by two feet. I could crawl in the thing at that age.
00:14:56
Speaker
He welded it up, painted my name on it. And then just one day it was like, John, here's your toolbox. And let me tell you what, I was a made man. That meant so much to me. And it was so cool. And I had a little spot in his shop, like for sure. Cool. Yeah. Yeah.
00:15:18
Speaker
I forget to internalize how important those kind of things are to kids, to our employees, to things like that. I'm not good at observing stuff like that, so it's good to hear you say that. No, I said it for the same reason. That's what stinks about parenthood is you don't know when one of those memories, as innocent or small as it is or as huge as it is, is one of those life-changing moments or whether it's just blasé, whatever, nice thing.
00:15:47
Speaker
Well, I realize even the past couple of years and your kids now too. So my kids are 10 and 13. Actually, Claire is turning 14 today. She's now 14. Wow. Yeah. Happy birthday. So anyway, 10 and 14 now. I remember 10 and 14 as a child. I don't remember seven. I don't remember five. I don't remember two, right? Like my kids are now at the age where they will start to remember almost everything. Yeah. And that changes the way I think about things.
00:16:16
Speaker
you know? Yeah. Because, you know, when they're three, they're like, you're not going to remember this when you're, I'll give you a good life, but you know, it's fine. You know what I mean? But the trip to Disneyland when you're three is not as important as when you remember it for your life kind of thing, or whatever the thing is. Anyway. Yeah. Yeah. So I do think about that.
00:16:38
Speaker
The skills you teach them, the experiences you have together as a family, it's going to stick good, bad, or otherwise. So make sure it sticks good. Yeah. Sometimes you think about, are you being a good parent? And at the end of the day, just spend time with the kids. Just hang out with them. There's that fun hashtag, I'm not here to be your friend, but spend time with them. We've tried to do a better job at,
00:17:05
Speaker
We're rotating between playing card games, playing board games, or building puzzles. It's not like an every night thing, but definitely that's better than letting everybody pick up their smartphone and go into the corner.
Project Prioritization and Filming Setup
00:17:22
Speaker
For sake of accountability, I kind of got to fess up that
00:17:27
Speaker
do it right now is a total failure. And I'm okay with it because what happened is I realized I need to kind of just 100% mentally and time management focus on Gen 3.
00:17:40
Speaker
It was all these little things. It was filming that video and filming that instructional video in and of itself wasn't particularly difficult, except we wanted to be decent. That meant nailing down Torx specs, making sure those match the PDF, documenting stuff. We found one minor change that we wanted to make. We made that to the actual physical product.
00:18:00
Speaker
setting up the, we don't have a free easy clean machine these days. Like they're all used. I didn't like that for the video. So we ended up setting up a, uh, Anno blend fixture plate. That's totally fine. Uh, on a old welding table that's rigid enough where you can like clamp on it and it doesn't look under camera. It doesn't look like it's moving around like a
00:18:19
Speaker
Reiki Cartwood. And then and then set up kind of got all that rigged up together. But it basically, I realized I need to just be focused on this and it was great. And it's kind of like, okay, rather than just tell myself I'm doing this and then keep failing or not say well, I'm not gonna do it today because this this this I'm like, okay, let's get through this and then we'll come back to it. All you're saying is that work got in the way of your idealistic two hours a week, twice a week kind of
00:18:46
Speaker
Yeah, of course this is fine. I don't owe it to anybody except it is evidence that...
00:18:56
Speaker
even having some fair amount of discipline and so forth, like you, I right now just was like, it's not worth the, I just couldn't do it. I will. In the days I've done it are awesome. And I've done a few, what was there two? Well, so frankly, the setting up the camera rig ended up being its own form of a do it's because we have a lot more instructional and marketing videos we want to do. And so rather than create a one off. Yeah, you're creating a base.
00:19:26
Speaker
what I realized, ironically, you know that wire rack I just told you we pilfered? I kept I haven't built this yet. But there's one shelf left on the top. So it's just four poles without all the shelves. And then that's going to be our semi permanent filming rig. We're gonna hang the DSLRs and the lights that we already own but don't use anymore. Yeah, off of that it wheels right over the fixture plate. It's like a top down rig kind of thing but on wheels.
00:19:53
Speaker
Yes. Now, anytime me, Alex, maybe even Yvonne want to snap a new photo, product photo, demo use case, we have the light and the camera set up.
00:20:05
Speaker
its own form of like a process that I love. That's great. Yeah, that's something our team has talked about is having a mobile filming rig that's always ready to go. It hasn't really worked out. We haven't wanted it bad enough to get it done kind of thing. And it is kind of weird and awkward for a lot of things. But I think for your case, perfect. And you've got the accessories available.
00:20:26
Speaker
Perfect. Yeah, I liken that to setting a goal for yourself, like, I'm going to work out every single day.
Dealing with Setbacks and Tooling Efficiency
00:20:33
Speaker
Well, today I slept in and I got to take the kids to school. So I've lost the ability to work out before school kind of thing. So even though the perfect plan, oh, I failed. I fell off the wagon. No, it's like life gets in the way sometimes.
00:20:50
Speaker
It's not over. You will hit another do it and that's fine. The whole goal for any of that is progress. It's just forward momentum. As you've seen those one to two hours every now and then that you get to spend turning your shop into the dream shop that you want to put your name on, that's worth it. Maybe you can't do it all the time, but you'll get it done.
00:21:19
Speaker
Oh, yeah, honestly, I think some of the excuses last week included, hey, we're selling some of those machines, I was getting those decommissioned, which that's some I'm in ended up being involved in. Yeah, that's done actually. So or some most that's done. Anyway, the one thing I did do as a do if this is in the spirit of a do it's but it was just a one off was we bought a new dedicated one inch face for the horizontal.
00:21:46
Speaker
I don't like sharing some of those tools. It's so much nicer to not share tools when you can avoid it. Share between different machines? Share between different products on the horizontal. Yeah. Okay. Yeah. Yeah. Dedicated for product. I ended up buying a Haas tooling. We have used a bunch of their face smells and we liked them. I thought, I always forget about them. I literally told my buddy at Haas, like, hey, I have to laugh
00:22:10
Speaker
because when I think of Facebook, I either go to like Sandvik or YG1 or Mitsubishi. I always forget that you guys sell tooling and it's all, I think some of it's actually just rebranded YG1, et cetera, but the prices are stupid good. So, bought a one inch face mill, couple packs of inserts, it came and it came as well with two extra insert screws, which is great because these screws are stupid tiny. And I was like, I'm tired of doing this the wrong way.
00:22:36
Speaker
And so we grabbed a Gridfinity base. I grabbed my 2x3 Gridfinity Fusion file already. And I went ahead and 3D printed it with the tool number and part like tool 69 castle grips in the actual Gridfinity label bin. I don't have to label it, it's already there permanently.
00:22:58
Speaker
And this is the best thing in terms of me trying to brag or flex on this. I measured the insert pack. We had about 0.7 inches on the right side of this 2x3 Gridfinity bin. And it went diffusion, extruded a wall, which created a small pocket on the right side of that bin. So the inserts fit on the left portion. And in the right, we put the little mini-zip lock bag with the extra insert strips. Yeah.
00:23:26
Speaker
Now you open this drawer, tool 69, castle grips, inserts, and oh, we've put the Lex ID in there as well so we know how to reorder it. Then the insert screws are right there. Freaking love it. Oh my gosh. Exactly. It took you a little bit of work to set that up, but not only are curating the process for the future, like yeah, everything needs to be this clean and this tidy, everybody in the shop.
00:23:48
Speaker
Exactly. Speaking of Gridfinity, I started printing a cabinet for Gridfinity's desk. It's just the design I found. I think it's on whatever the bamboo online platform is, Maker World or whatever.
00:24:04
Speaker
So you print gridfinity pods, containers, whatever. I already have a bunch of them. So this file comes with a drawer and with a housing that all clips together and is printable and dovetail together. And it's like one foot by one foot by one foot cubed, basically, with however many drawers you want. And the gridfinity bins just go into it.
00:24:30
Speaker
Once I've got it all together, it's almost done. I'll post a bunch of pictures, but it's great. I didn't modify anything. I just picked the drawer thickness that I want and then they have skinny ones and fat ones and all the grid finities I've printed so far are the same, whatever, two inches tall. So I'm standardizing to that. But all the clutter and junk on my desk and all the knife handles and pen parts and stuff that I'm actively working on, but it's really just clutter now has a place to live.
00:24:57
Speaker
And it's gonna be sick. And I was gonna print this double wide one that's almost two feet wide. And then I printed the first layer and I'm like, where am I gonna put this? This is way too big for my desk. Yeah. So I'm gonna have two little cabinets on either side of my computer. And it's gonna be great. Super excited.
00:25:14
Speaker
I just pulled up that maker world in search gridfinity. Are you talking about the five by five gridfinity box by pred? I don't know. Okay. I'll poke around or if you, it's got like, it's got like eight drawers. Um, Oh no, that's not this. And there's a multi and a double. Um, you'll find it poke around, but oh, modular good for any drawers. Maybe interesting. Okay. Yeah. It looks like a Kennedy box, but small.
00:25:39
Speaker
Okay. Holy cow. There's a bunch of things out there. It's awesome. It takes quite a bit of filament as any bigger project does, but whatever. It's cheaper than buying a Kennedy or a Husky box or something and it's purpose suited and has all these little bins inside. When I showed Garrett, I got that reaction. He was like, oh my God, yes. That's so nice because we have
00:26:04
Speaker
In the past, we'll try to rubber band inserts together. You can write the tool number on the insert pack. That doesn't work over time. We have a box, a drawer with boxes in it just for the replacement screws. That doesn't work. Just put the screws with the inserts. I love it. I love it. For end mills, I've got, they're not Gridfinity, but they are very similar sized little containers. I need to make them Gridfinity. But they have little squares for end mill containers. OK.
00:26:33
Speaker
So we buy, let's say, a quarter inch ball mill. Maybe we have 10 of them. And maybe 10 is our max number. I ended up not printing. I was going to print. But just taking a white paint marker and writing on a little insert piece, 10. And you put that in with the thing so that when there's three and you're ordering from that vendor again, we just top up to that number. Because whenever we order from Du Bois Tour or Lakeshore or whatever, Harvey,
00:27:01
Speaker
It's like, okay, I'm making a big order to Lakeshore. Shipping to Canada is kind of expensive, so I want to order like $2,000 worth of tools. So I walk around the whole shop. It's not a lot of places. And I'm like, I'm topping up on Lakeshore, you know? So I need to know how many to order. And I never really had a maximum before, but now I do. And now it's like right at the source. I'm like 25 of those. Okay. I need to order three because we already have a bunch, but whatever. I'll top up. That's good.
00:27:29
Speaker
Okay, big one is, so the speedio slash a row install is done comma. No, it's good.
Automation and Custom Solutions
00:27:39
Speaker
It works. I got it working. And I had to write a I wanted to write a custom macro to basically call the palettes that I want.
00:27:52
Speaker
And I thought of a lot of different ways to do this. And there are a couple of different ways to do this. But basically, I want to use the 600 series macro variables, which are not used for anything else. 600 to 699. And basically, that's my order schedule. So 601 is going to run first. 602 is going to run second. 603 is going to run third. So that's the order you want. And then in that, if I type 69.1001, it's going to call palette 69. And it's going to run program 1001.
00:28:22
Speaker
And having thought about it a lot, and once I actually typed it into the macro page, I'm like, yeah, that looks really good. OK, I like that. It's obvious. It's simple. The math behind it is kind of complicated to strip out the 69 and to call program.1001.
00:28:43
Speaker
But it's working out super good. And it allows me to call multiple palettes in a row in whatever order I want, whichever one's first. And so I've been working on that and testing it out. And it worked awesome. Can you give me the layman's example or explanation? Like you're explaining it to your aunt. You have a Kern. You have a brother. They share an aroa.
00:29:06
Speaker
Who's in charge? That's a good point. A lot of people will get the Aroa job palette manager, whatever it's called, so the Aroa is in charge. You literally post your programs to the Aroa and it sends them to the Kern and sends them to this video. I didn't want to do that for one because it's a $5,000 upgrade and I'm cheap, but also I just didn't like it and it's not how we run the Kern.
00:29:30
Speaker
So right now the current is in charge and it says, give me Palette 12. And the aero goes, okay, here's Palette 12. And then the current is in charge of running whatever program it's scheduled to run on Palette 12. Okay. And I want the same thing on this video. So this video is in charge of its own life. And I wanted to tell the aero
00:29:47
Speaker
Give me palette 72 and it goes, okay, here's palette 72. Then what program are you going to tie to that? That's what I'm using the macros for. 72.1001 is going to call palette 72 and run program 1001.
00:30:02
Speaker
So you don't have to do that on the Kern because the Kern has a little bit tighter integration, I guess, with the Aroa. It already knows. It's similar. I'm doing actually quite a similar thing on the Kern. I'm asking it to call a certain palette in whatever order I want first and run this program tied to that.
00:30:20
Speaker
scheduled order and run in Palette. So there are all these variables that you have to tell the Palette changer what you want. And we find that we like to reschedule the order. If we make a 12, 18-hour schedule, sometimes we're like, actually, we need this first. So we'll jump in there and we'll schedule. So the order is pretty important for us. So I'm using the 600s as, like, 601 is going to be first. Whatever you type in that is going to call first.
00:30:46
Speaker
and then 6.02 and then 6.03 and then 6.04. And if you're at 6.04 and you need something now, you put another job at 6.01 above it so that it runs first. And there is some, I did some wow do logic in Acro to make that. I actually needed two levels of wow loops to be able to pull that off. So there were a few hours of a late night, you know, screwing around to get this logic to make sense and not just error out all the time.
00:31:16
Speaker
But once I got it, I was like, jeez, this is exactly what I wanted. This is sick. And because we want to be able to call, you know, in an overnight call 10 to 20 different palettes in all different orders and orientations. You know, Dennis was saying he uses his similar setup.
00:31:32
Speaker
to basically start a palette one and then go sequentially and all run the same job. Okay. So if he's got, you know, 20 vices with 20 pieces of aluminum in them, it's going to run them all the same. And we don't have that. We don't want that. I want to be able to run this and then run that and then run this. And actually I want this instead, uh, which adds a lot of complexity and logic to the system, but
00:31:53
Speaker
Dude, it works awesome. It works so good. That's great. Yeah. You're not overnight, you're not doing a part first on the current and then that same night unattended, it would get past the speedio? I'm not yet, but I want to and I think that's the next level of complexity that I want to add. The speedio is going to
00:32:14
Speaker
say make a soft blade and do everything except for the engraving, which requires kind of a five-axis tilt. So I want to machine it all on the speedio, park it in the aroa. I need to export or de-print something from the speedio to something like a Raspberry Pi that the current is going to read and then it's going to write to the current and say, okay, speedio finished. Here's proof.
00:32:39
Speaker
now schedule this, run this, and then the current is going to finish that and write it as proof. There needs to be some communication back and forth, which gets a little tricky. I don't know exactly how to do it, but I know conceptually how to start that.
00:32:55
Speaker
It's going to be sick.
00:33:11
Speaker
down on it. Like, there's just such a lack of basic, like, I don't even know. I'm not trying to plug this in. It's not easy. It's like an AI thing. But like, there's so many, you know, our Akuma is great. Well, it's not great. There's things I really don't like about the automation side of it. And they're never going to get fixed. They're never going to get fixed because there's all sorts of quirks about like, yeah, everybody else that has those machines and big companies that blah, blah, blah, blah.
00:33:37
Speaker
you know, why is it that if one tombstone break, at least as far as I certainly don't know everything about everything, but like, if one tombstone three detects a tool break, why can't it more easily or at all look at the next tombstone that's scheduled and say, well, so long as you don't use that tool, then I keep running you or scenarios like that. And you did a great job of getting pretty custom with it, but that's not,
00:34:04
Speaker
That's not within the purview of most people that want automation. That's pretty tribally stuff. I think if you're nerdy enough, it's within Realm, but it's even difficult. Of course, it's difficult for me too. I'm just willing to bash my head against the control to make it work.
00:34:26
Speaker
The problem I find is when the solutions provider, whether it's Okuma or Kern or whatever, they sell this package, they've only forethought so many variables, so many problems and solutions and scenarios. And then machinist guy is like, man, I still wish it did this.
00:34:43
Speaker
I wish it did differently. You can't plan for everything. You have to pick a direction and say, I'm doing it this way. For us, I'm not saying my way is best, but I'm saying it works extremely well for us and it would fail at other people's shops for whatever reason. That's not how they run things.
00:34:59
Speaker
So that's where automation gets weird. And sometimes roll your own custom is attractive, you know? Yeah. No, I hear you. Like we can't, you know, let's say Tombstone 3 has seven programs on it. If I don't want to run three of them one night because we didn't reload those parts, or we're not happy with something which is unbelievable, I'm going to manually delete those out of the thing. And then you've got to retype them in later. Like there's no reason you should be fat fingering these long string names.
00:35:29
Speaker
Similarly, there's no practical way to flag different programs as different materials. We're still trying to work on this long-term chip management thing. It's like I'd rather run all the steel at night and then program a dwell while the conveyor runs, clears and stuff like that, and then go run all the aluminum. These are very simple logical things that systems just have been built for. Yeah. Which has only come up as you've needed them. Because I'm running it like this, man, I really wish it did that.
00:36:01
Speaker
But other people, there's no way all of these things are unique to us. We do the same thing where we wait priority on overnight runs. We always run one tombstone at the end because it's the longest. It makes the most parts, meaning if we do
00:36:19
Speaker
Goof, it's easier to catch back up because it makes so many. It also tends to be, has more slotting, so it sometimes is more likely to break a tool. And I don't want that to stop the rest of the night run. Simple answer, but like... Yep, same. Yeah. What was it? Oh, so the other thing is...
00:36:42
Speaker
because now we have an aroa and now we have a kern and a speedio that are both tied to it and we kind of thought that this would happen but it hit us the other day.
00:36:52
Speaker
Monday morning, we come in, did a good weekend run, good 30 hours after Friday. That's awesome. Unattended, except for at the very last pallet, it probably ran 20 pallets, like 20 different programs. At the very last program, it failed for whatever reason. I can log in from home and I'll be like, yeah, it had a pallet change here or whatever. It's the last one. I don't really care. A 30-minute cycle.
00:37:14
Speaker
didn't miss much kind of thing. Angelo can deal with it in the morning. So Angelo comes in and him and Steven deal with it. But now because the two machines are tied together and the E-stop circuit got a lot longer because now it has to go through this video. No. It got weird. So basically what happened was in the Aroa rack,
00:37:37
Speaker
say it was calling for palette 75. Well, palette 75 wasn't there, but it was still scheduled to run for whatever reason. The guys forgot to load it or they typed the wrong number into the call palette spot. So the row grabs air and it goes to put it in the chuck and when it loads it into the Kern table, that's when it notices that there's nothing there.
00:37:59
Speaker
The grippers themselves don't have sensors that say, I didn't grab air. So it grabs nothing. It puts it in the chuck. And then it goes, wait a minute. There's no feedback loop. There's no resistance noticing that there's a pallet here. So then it stopped, whatever. In that state with the aroa arm out into a machine, the service doors are open. It's a quote unquote dangerous situation.
00:38:22
Speaker
you can't, shouldn't open the access door and take off pallets and like lozier pallets, because now you're opening another door. And the act of opening that other door makes it extremely hard to recover from, to get everything back in. If we left that door closed, we've done this before once. If we left that door closed, it would be way easier to recover. You just home the robot and close the doors and it's fine. But it got the current into this very stuck point. So Monday morning for five hours, it was stuck.
00:38:51
Speaker
And I'm texting Tina at current. I'm like, we're stuck. Do you have a moment? Can I get on the phone with you, please? I know you can help me figure this out. And it's like any automation system, you need to learn error recovery. Because what's going to happen if somebody smashes e-stop when it's way twisted up? You got to know.
00:39:14
Speaker
And this video added some weird, like it has to be on for this to work. Yeah. Kind of weird things. Do you know? Yeah. What if you stop one machine? Does it stop the Kern or does it just stop the pallet changer?
00:39:28
Speaker
and it basically only stops its own machine and the pallet changer unless the pallet changer is also changing into the other machine. Anyway, the secret was there's this little button on the current control panel on hide and hide that looks like if you put your hands together like the heart symbol.
00:39:46
Speaker
It looks like that. It's like a heart healthy operating machine. It's called the permissive button. That means your hands are where they shouldn't be. That means you have a service door open and you can turn the spindle on or something like that. You have to hold that button through a full reboot and you have to let it go when it tells you you can't push the button and you have to push it again and let it finish rebooting and that's the secret to close the doors and to do it all weird but I didn't know that.
Automation Costs and Integration Challenges
00:40:15
Speaker
Now you do. Now I do. So anyway, got that going. Chris has been running great otherwise.
00:40:21
Speaker
I wholeheartedly agree when we knock on what hasn't happened lately, but like if we're, the Okuma is indexing tombstones and we accidentally switch it from automatic mode to manual mode. Yeah, you've mentioned this. Oh my God. Everybody, me and Garrett really have like, we are complete rule followers at that point. Pencils down, do nothing and then go back to the exact way it wants to be. Like there's no getting cute or creative.
00:40:47
Speaker
Otherwise, you're on the phone with service for a while. It's just a weird sequence of pushing buttons. And that's what makes automation so hard, is it's the sequence of events. You know, PLC logic and e-stop circuits. And it's logical if you're really smart and know exactly how it all works. Service tech comes in, and they're like, oh, clearly, it's blah, blah, blah. This is not communicating properly. Well, I'm an idiot here. How do I run my machine?
00:41:17
Speaker
No, you're not. I mean, this, this stuff, I mean, it's the, I've heard about it with people that have used some of the turnkey, like the third party automation solutions you see on Instagram. Like I've heard some pretty not good stories. I've kind of had some running side bed jokes with peers in the industry. Like there's a reason you're not seeing a proliferation of robots, cobots. I don't care what you call it. Like certain people are using them in certain situations, but they are not
00:41:44
Speaker
You re-round at this point almost 20 years ago, 15 years ago, there was in terms of the non, how would you call it? Basically, if you want to get into machining. 20 years ago, it was like Little Machine Shop, Grizzly,
00:42:00
Speaker
tag. There were some options. CNC wasn't really a thing or certainly not like it is then. And then all of a sudden you have all these players that have come online to the latest point where you've got now multiple players in the sub $50,000 range and there's lots of people using them and sharing their successes. There is no equivalent of that that I've seen in the automation world.
00:42:22
Speaker
Both at the price point, but also just like success of using them. Exactly. And some of those automation systems are on one hand prohibitively expensive. Like you spend a hundred grand on a really nice VMC, adding an automation sell is another 150 on top of that. And that's like, what? Like weird.
00:42:42
Speaker
But if you think about it, if you really can push it to 24-7 operation, like, yeah, it might totally be worth it. But it's still like silly money to be spending. Silly money and there's all these soft things that just aren't great about them. And what I was trying to say was you just don't
00:42:59
Speaker
pick a Cobot that might be 20, 30 grand if those even exist now, probably more. You just don't see them loading up a Tormach or whatever machine because it ends up the automation device and process reliability and gins and chip management and how do you unload them and what happens if you make a bunch of it. But it can be great.
00:43:19
Speaker
That's where optimism comes in. But I do think a lot about buying a UR robot and putting it on the mori or something and setting up a rack of small pallets and just doing it myself and rolling my own. I feel like I know enough about the needs and the requirements of the industry to be able to pull that off no problem. Still a lot of money, even doing it myself. UR is 50 grand or whatever.
00:43:47
Speaker
somebody, not in my lifetime probably will, or not in my working career, maybe like we'll figure out, wait a minute here, some combination of like good vision stuff. And then maybe I hate using AI, but like, no, like put some reasonable probabilistic intelligence into why something is wrong and self solve it in a way that's not this dumb brute force PLC logic ladder where it just shuts down, has no ability to resolve.
00:44:12
Speaker
Yeah, it's happening. But I mean, coming from somebody who, and you too, who have palette changers that are freaking amazing and have changed our life. I'm not trying to dissuade anybody here. They are amazing. They're a lot of money. I like the Aroa is somewhere between a hundred and 150,000 depending on who you ask and where you buy it from. That's a lot of money, but it came with the current and I just did it and I'm super glad I did it.
00:44:39
Speaker
So yeah, in a sort of anti-climatic way, we're transitioning Saunders to where we will have, when I say anti-climatic, it's not, we already have four verticals that make fixture plates. The Okuma is automated, it runs, quote unquote, everything else. We have a VF2 YT that's kind of general purpose prototyping, one-off stuff. And then we are, we're replacing our Wilhelmin with a new Wilhelmin that will be barfed, lights out,
00:45:07
Speaker
automated, it actually will have a robot to unload bigger parts. Nice. And then the lathe will stay, the lathe, the ST-21 for now, but I think it will get replaced long-term with a Moltus that will just be a better version of a lathe that will also be barfit. So my punchline is that we have four verticals that do our significant one-off fixture plate items, which would never be automated for all intents and purposes. And then everything else is automated. Yeah, exactly.
00:45:36
Speaker
It's fantastic. It's a good plan. But I mean, you've earned that plan over the past 10 years, right? Like you've earned your way into automation. Yeah. I'm not proud of the way I got there. It took me to be a believer. It took me to own the Okuma to understand and appreciate it. And that's only been a year, whatever you've had it?
00:46:01
Speaker
No, it'll be two years in March. Still, I remember our conversations before that, where you just didn't know yet. All the fans, you're like, yeah, automation is cool and stuff, but now we're drinking the Kool-Aid. Right. I mean, the Okuma is a
00:46:24
Speaker
Solution it's a purpose built. It's not perfect, but it does the job, right? Even with the current in the aroa cell I had to do a lot of programming to get it to do exactly what I wanted, but I could which is cool When you start to put a pallet changer on a machine that wasn't designed for right? Yeah this video a speedio has been a year and a half I've been working on this stupid thing to get it to pallet change That sucks. Yeah
00:46:49
Speaker
The last thing I had on my list to share that we're also doing fits well into this whole story is we're finally now. We have what I would consider very mature fixturing on the horizontal. We know what works. We know it doesn't work. We know what needs change. To be honest, we've done pretty common workflow is just make aluminum fixtures. Sometimes we'll just make them aluminum, period. Sometimes we'll helicoil them from day one.
00:47:13
Speaker
If they're pure aluminum, a lot of times the threads will fail. When they fail, we'll just throw a heel coil in. Not great, but it works. Now we just bought a bunch of A2 and we're remaking these fixtures and we're going to do, I've actually never really done this, although I'm not nervous about it. I'm sure we'll learn, but basically we're going to machine A2, leave a few thou, send them out to be heat treated. And then we will either grind or hard mill.
00:47:40
Speaker
And then we'll have beautiful long lasting fixtures. These are for your internal in-process fixtures. Correct. Which is only mildly confusing because you make fixtures on the fixtures. Oh, sorry. Yeah, too shy.
00:47:55
Speaker
Cool. Would you have any advice or endorsements if we end up getting like 58 Rockwell A2 back? There's some features that I don't think are going to easily fit underneath the Okamoto wheel right now. I don't really want to buy a narrow wheel. You hard mill. I don't hard mill. I have. We actually soft mill all of our blades. We don't really hard mill anything. Oh, I thought you used to. Because we lap them hard.
00:48:20
Speaker
Can I, on a good machine, can I hard mill 58 Rockwell and not? All day. Okay. Yeah, totally. That's probably what most people use their Okumoto surface grinders before it's to hard mill. What do you mean? I don't know. I feel like most tool makers and stuff surface grind hard materials, not soft materials all day. Yeah, but I want to
00:48:46
Speaker
hard mill it in the vertical. Hard mill or surface grind? What are we talking about? Sorry. I will surface grind the big open faces of these fixtures. There's some smaller features that I would like to hard mill. Oh, easy. Okay. Yeah. I mean, you're going to trash your tools eventually, so you use dedicated tools. They're not going to be used for anything else, but I've hard milled a lot. We hard milled all of our blades, a lot of features. The coating kind of helps a little bit.
00:49:15
Speaker
Yeah, don't use an aluminum end mill because it's so sharp, right? You know that. But the hard milling is nice because it just go light, a lot less SFM than you're used to, light chip load, and it's beautiful. It's such a shiny surface. You don't like it. Okay. Well, because I remember, we have hard mill. We did a Wednesday wedge on it. We did a proven cut on it. But I think most of the time I did, I bought the Lakeshore hard mill hard mills. And you were like, yeah, you just use steel.
00:49:46
Speaker
designed for steel and mills. Right, but just an off-the-shelf, like the basic basic. Yeah. Like your carbide basic regular stubby if you can. Yeah. For flute coated.
00:49:57
Speaker
Whatever. Just run and roll. If you're really trying to chase like a bore diameter or something like that and have it be like two tenths, then it's a little harder to play around. You don't want to do the bore tool path and go down and different stuff. But yeah, just play with it. Yeah. Hard to believe it's nice. And no, any concern on
00:50:21
Speaker
We'll cut, tap them, then have them set up for heat treat, but the taps, I don't know what class it is, but if the part grows by a thou, that's not going to mess up. It shouldn't move, yeah. Okay. Maybe do a test piece, make a test small fixture, set it up for heat treat, surface grain and play it, mill it.
00:50:39
Speaker
That way you're not investing the time and the effort and whatever in like a big, big thing. Good. But yeah, a soft milled tap, tapped hold, then heat treated. You kind of can't really chase it at that point because the thread is hard and you're going to break your tap kind of thing unless you tap it bigger and then chase it with a smaller tap or something. But I don't know. Don't worry about it. You'll get it. Yeah. Okay. I guess our informal time is up. I should update.
00:51:08
Speaker
everybody on what I've learned about heat reading for production. I get some good questions answered.
00:51:15
Speaker
demystify to me at least, which yeah, pay forward and share whatever. I'll say it next week. Sounds good. What do you, uh, what's on tap? On tap. Um, I got it. So the, like I said, pallet changing on the speedio worked and then it didn't work. And at that point it was one 30 AM and I was like, I'm going home right now. Um, so that's fine. Currently we just need manual pallet load, like just, just the one for right now. Um,
00:51:43
Speaker
So today I want to spend some time on this video. I want to test that. I want to figure out why it didn't work. It might have just been some weird current mismatch, signal error, restart everything and it might just work, but I got to debug that. And then I can't for the life of me get dprint to work on this video.
00:52:02
Speaker
where say you probe a feature and you want to log that result into a dprint file. I can't get it to work. It's driving me crazy. So I'm asking a bunch of people. You guys talked to Ken? I haven't talked to Ken yet. I've talked to some other guys at Yamazin and some other video owners.
00:52:18
Speaker
And some people have no problem. I wonder if I just fact figured a setting wrong, and it's just not working. So I'm sure whatever the error message it's giving me will lead somebody to the right answer. I'm just going, ah, it's got an error. Clear it. Try it again. Yeah. OK. So anyway, I want to figure that out, because that's going to help us with consistency and knowing what's really going on. OK. Interesting. I have a question for the audience.
00:52:47
Speaker
Going back to this whole chip thing, keep distilling it back to the basics. I just want to know what's coming out of the conveyor, full stop. I can make some other decisions downstream from there about how we were to separate it, magnetic or shoot lever that switches.
00:53:03
Speaker
I'm not always the best at simplifying things or not seeing the... Yeah, hear that. So we can bear a small-ish amount of chips. Now, not like you. We actually make a lot. We fill up the Gaylord every day or two. But like
00:53:19
Speaker
It's not like chips are coming out like a conveyor at a coal mine where each baffle is filled to the brim with the material. I want to know, is it currently running aluminum or steel? That should be possible because aluminum is not... I keep thinking of a Hall Effect sensor, but then I'm like, I don't feel like that's the right answer of putting a Hall Effect sensor or many of them and starting to look and see, am I getting enough hits on the Hall Effect or something to see if it's a ferrous material?
00:53:48
Speaker
And you want to debounce it because obviously basically you're saying like, Hey, as soon as steel hits, it's going to be sealed for quite a while. But that going back to my sort of like, whole thing on like automation and simplicity, like, this is not a complicated thing. Like any four year old could look at that and say whether it's aluminum, shiny, or steel, it's matte, or like there's so many differences. What how can I do this? Hmm.
00:54:12
Speaker
I mean, my brain goes to the source, the code, you could flag a variable, right? That's what I'm thinking. But then your program specific, you have to do it in fusion, you have to like do it from the beginning kind of thing. You want an end of line solution that just goes on care what's happening. This needs to figure out what's happening.
00:54:30
Speaker
There's a bunch of problems with that. One is the fact that there's almost a 15 minute lag sometimes because of how slow the conveyor is, or not many minutes. Interesting. It doesn't, it just doesn't, it's not. That's a good point. The elegant solution is tell me what's coming out of the conveyor because that's actually what I want to sort. And that's simple. And it will be fairly consistent. Once steel starts coming out, it'll be steel for an hour or two.
00:54:54
Speaker
Yeah. I don't know. You've got a good challenge on your hands. Yeah. I'm going to keep chewing on it. But if anybody has any, you know, the simplicity is the design here, the idea. You need like a little pneumatic robot arm that grabs a sample and dispenses it into a little scanner. Maybe vision is the answer. Like you can, I don't know. I got to think about it. Yeah. Okay. Cool. Sweet. Good luck. See you next week. All right, man. Have a good day. Bye.