Transcript
John Saunders: Good morning. Welcome to the Business of Machining episode number 460. My name is John Saunders.
johngrimsmo: My name is John Grimsmo.
John Saunders: And John and I just freaking love machining. Need we say more? All right.
johngrimsmo: Yep, end of story hang up now.
John Saunders: Yeah, yeah, end of story. that sorry i I know we were off last week. I had a last minute chance to to attend a World Cup game with a buddy whose spouse couldn't go, and I said, yeah, why not? And the reason I bring that up is that I...
John Saunders: So I'm by the Computer History Museum, so I'm going to talk about that. quick but good home assistant stuff. Hand tools, measuring, um slide rules and physical computers.
John Saunders: And then lastly, but certainly not least, we have two new machines in the shop. Yeah, that's my list.
johngrimsmo: Nice. Nice. My list has foam, bloom update, toilet bowl float, hard milling, and a NAS.
John Saunders: Hmm. The artist, recording artist.
johngrimsmo: No.
John Saunders: Oh, okay.
johngrimsmo: good
John Saunders: Dude, go for it. What you got?
johngrimsmo: Well, I'll start with the NAS. So I bought a network attached storage, like a NAS hardware or a hard drive type thing.
John Saunders: Okay.
johngrimsmo: I've never had one. I've never done. I've never touched a raid. I've only heard about it. This is my first foray into that. So I got a a four bay NAS with two hard drives, two 16 terabyte hard drives um in the raid one platter.
John Saunders: Ooh. Platter or SSD? Okay.
johngrimsmo: It's set up in raid one. so
John Saunders: Which is striped or mirrored or...
johngrimsmo: Mirrored.
John Saunders: Okay, so backup.
johngrimsmo: Yeah, backup.
John Saunders: Yeah, right.
johngrimsmo: So I have i have two 16 gigabyte hard drives giving me 32, but I only have 16 workable because they're mirrored, whatever. And then if you want RAID 5 or 6, you need four hard drives. And I was like, I'm not going to spend that much money right now.
John Saunders: Right.
johngrimsmo: It was already over $2,500 just to get like the NAS and the two hard drives.
John Saunders: Ooh.
John Saunders: Synology.
johngrimsmo: I went with Ugreen.
John Saunders: OK.
johngrimsmo: Ugreen case and then Seagate, Iron, Fist, whatever, hard drives. Yeah, it came in yesterday. i had Ryan install it. He said it was easy. And then he said he sent me the password, but he didn't. And then I couldn't log in.
John Saunders: funny.
johngrimsmo: But so today we'll like fix it up. and But yeah, so the goal is to store all of our media footage, all of our videos. I still have one terabyte of footage that I filmed in California that I haven't edited yet.
johngrimsmo: So this will give us a place to put there.
John Saunders: Right.
johngrimsmo: And then we're also going to experiment with external video editors for some of the bulk stuff.
John Saunders: Yeah, right. Sure.
johngrimsmo: And this, while we won't give external people access to our NAS, this at least lets us like store and share internally like all of those files. Because right now they're hidden on like Ryan's Mac you know editing PC, and he turns it off at night.
John Saunders: Yep.
johngrimsmo: And I just have no access to any of this, and it's driving me crazy. I can't just pull a clip from Jay Shapiro Watch Company or from Lainey Machine Tech when I was there and like slap it on Instagram because it's hiding.
John Saunders: Yep. Mm-hmm.
johngrimsmo: um So yeah, i'm I'm really excited for that. It's gonna, you know, I thought about the plan I want to do to increase our video output.
John Saunders: who
johngrimsmo: And getting an ass was at the first step to that. So i' I'm really happy with how this is turning out so far. And The next step that I was researching last night was file structure and folder structure for like how you want to start this like clean slate from the start.
johngrimsmo: You know, how do we make this be the greatest thing ever?
John Saunders: Right.
johngrimsmo: Not like annoying and hard to find stuff. Um, so that's been really interesting to, uh, go back and forth with chat and and other research and try to figure out how I want this to look. Cause I get a clean slate now and I have already have close to 20 videos that need to be edited. So I need to like organize them and have it be good, you know?
John Saunders: Yeah, that's exciting.
johngrimsmo: Yeah.
John Saunders: It's fun because it feels so like like and it like they're not new, but it's also like, oh, this is really nice.
johngrimsmo: Yeah. Yeah. I'm looking forward to that.
John Saunders: And that's it's located in your facility.
johngrimsmo: Yes.
John Saunders: Okay. And is your outbound internet like bussin or just good?
johngrimsmo: Our upload speed, you mean?
John Saunders: Yeah.
johngrimsmo: I think it's really good. It's bussing bussing, as kids would say.
John Saunders: Yeah.
johngrimsmo: um I forget. I think we get... tell you right now. We'll keep talking.
John Saunders: Yeah. Right.
johngrimsmo: Mm-hmm.
John Saunders: I think while, while you're talking, while you're looking up, I want to say, and and don't quote me, I want to say we're like 400 down, but only 50 or something. it's It's not, it's not, it's not a problem, but it's not like, oh, I'm on fiber where you're just like, oh my God, like uploading a gig file is like, just, it's unbelievable.
johngrimsmo: I'm on Wi-Fi on this computer and I'm 200 down 75 up, which is not bad for Wi-Fi.
John Saunders: Yeah.
johngrimsmo: I think we get like 400 down wired.
John Saunders: somewhat comically you're, you just pixelated a ton when you were doing that speed test, which which is like, okay, so it isn't all being made fabric made up right now.
johngrimsmo: Really? Yeah, that's funny.
John Saunders: like
johngrimsmo: yeah
John Saunders: Like, yeah.
johngrimsmo: that's funny okay
John Saunders: I was just curious, cause we went through this and I ended up, I ended up pushing more away. We have two small cheap NSs and short end, I don't, not worth going to, I don't love them.
John Saunders: I don't like them at all actually.
johngrimsmo: What about like what makes you not like them?
John Saunders: Well, some of this is my own fault. Some of this that they're probably a little bit older. They are Synology DS220s. um And I find they they turn themselves off or like, don't re like this is none of this is a good example. So don't like judge your but by this. um Basically, I don't pay attention to them. And then when i actually want them, I look around like it's it's been turned off. Like what, why?
John Saunders: why
johngrimsmo: That's interesting, to like things to look out for, you know, to avoid.
John Saunders: Yeah.
johngrimsmo: Yeah.
John Saunders: Now, like I'm sure a battery backup could help, but it's also just like frustrating. And then but most importantly, I hate the web interface for them, the search and and that forth. Now, Amish, I believe, has the same one and has a much better experience. So it tells me I may be doing stuff wrong. But um for me, in NAS, I want to be able to quickly find what I want and then probably copied back to a local computer for whatever purposes and punchline is i just moved to google drive which is plenty good for most things i also once i did google drive which i think we talked about i then went through the time google offers a like takeout option where they will take your whole google drive which for me is three terabytes or something and it will actually send it to you in zip files at whatever size increment you want
johngrimsmo: Really?
John Saunders: Um, so I did it and don't even know 50 gig chunks or something. And then I did take the time to copy all of those to the NAS, which is just basically a point in time archive that I can then put a fireproof safe and you know, it's a starting point, stuck but you know what mean?
johngrimsmo: yeah. you Cool.
John Saunders: Well, you for video footage, this isn't video footage.
johngrimsmo: Yeah, it's good to know.
John Saunders: like I get what you're doing.
johngrimsmo: Yeah, and I mean, I've been using Google Drive forever, but it's not great for video footage because it's like big bulk transfers of files, right?
John Saunders: Yeah. Right. Totally.
johngrimsmo: So. sir
John Saunders: one Well, that's a great segue to um nerding out on the Computer History Museum. If anybody's ever in the San Francisco Bay area or little bit south near Sunnyvale or wherever the computer techy area is, Silicon Valley, it is super cool. like Nerding out, actually, to be critical, poor layout, like very, very hard to understand the flow of the museum.
johngrimsmo: When I was there a few months ago, i was looking, i had a few hours and i was looking for things to do. And I think that came up within walking distance, but I didn't, it was closed or something.
John Saunders: Oh, really?
johngrimsmo: I didn't get a chance to like, to go.
John Saunders: Yeah.
johngrimsmo: Hmm.
John Saunders: So what I really enjoyed was they started off talking about the basics of calculation of things, and it kind of makes me wanted to spend a little bit more time um understanding slide rules.
johngrimsmo: Hmm.
johngrimsmo: thank
John Saunders: um I've always had a soft spot for like the KERDA calculator, looking at the differential engines and people that built physical computers and Babbage and his mechanical device. I'm by no means an expert or even a novice, but like I've been very fascinated by that concept of yeah how do you mechanically carry the one and and calculate stuff.
John Saunders: And then I don't think I fully appreciate how powerful slide rules are. and I know anybody...
johngrimsmo: Yeah, I don't either, really.
John Saunders: Yeah, if we have an audience members of certain age, I'm sure you're probably yelling at me right now, but like, it's like unapologetically, that's a bygone era, like, but also it's cool and I enjoy it. So, um,
John Saunders: I don't want to ramble. It was really cool to see that. It was really cool to see the progression of of transistors and computing technology and servers and digital equipment and IBM and so forth. But then the thing I thought was really cool was they have the original Cray-1 supercomputer that was sold to Lawrence Livermore, probably to build bombs, but it's there. If you don't know what it looks like, it's like a hexag, either six or eight-sided,
John Saunders: object. It's about the size of a big desk, but towers up. Um, and it as, um, the guy was very eccentric. So like the design was both purposeful to maintain equal distance of the wires that the signals arrived at the same time, but it was also a lot of showmanship.
johngrimsmo: Love it.
John Saunders: So all the power supplies were in the ground that was effectively a, uh, it was an actual bench. Like you could actually order the computer with the cushions of your color choice. Um, It was $42 million dollars in today's money for this computer. And I did the obligatory, you know, LLM inquiry. Um, my current Mac book pro that I spent probably two grand on is approximately a thousand times more powerful.
John Saunders: And I know we all know that, but like the the way that Chad answered that was quite funny. It was like, if Lawrence Livermore, if you could teleport your MacBook back to Lawrence Livermore, not only would they throw the cray away, but your computer would be the single most precious object in all of our national defense system. You're just like, it's like unbelievable how far computing has come.
johngrimsmo: Yep.
John Saunders: And then seeing like the nostalgia stuff, like they copies of SimCity from the nineties that, you know, the stuff like that we grew up on.
johngrimsmo: Yep.
John Saunders: Yeah, it was fun.
johngrimsmo: Very cool.
John Saunders: Sorry, unrelated to machining, although arguably not with all the...
johngrimsmo: Arguably, oh, it's technology.
John Saunders: Sure. In hearing the most recent episode, I'm only part way through it, of the precision microcast.
johngrimsmo: Mm.
John Saunders: to those guys for getting an episode out. But hearing them talk about building Silo, sorry, Cyrus is his name. I always say Silo's Garage.
johngrimsmo: Yeah.
John Saunders: um You know, working on building mechanical devices or or machine tools at Atomic Semiconductor. i think that's the name. Is that right?
johngrimsmo: Yeah.
John Saunders: Atomic. um You know that's very much in the world of quote, unquote CNC.
johngrimsmo: Yeah. Looking forward to that. So I've I've caught up to episode 19. on my my drives to work.
John Saunders: Oh, OK.
johngrimsmo: So I haven't even seen Cyrus in the podcast yet. I think he's coming up soon.
John Saunders: Oh.
johngrimsmo: So that's really good.
John Saunders: Yeah.
johngrimsmo: But yeah, Cyrus gave me a tour of Atomic Semiconductor and I was not allowed to film at all.
John Saunders: Yes. Well, I think he's since to moved to Texas.
johngrimsmo: Couldn't even, we took one photo. Yeah, they started a branch in Texas and they moved him there.
John Saunders: OK. Yeah.
johngrimsmo: And he's like, i mean, good and bad because he said I will effectively make more money there because San Francisco is really expensive.
John Saunders: Yeah, yeah, sure. Yeah.
johngrimsmo: but But yeah, he's young and looking for every opportunity he can get. And such a smart guy.
John Saunders: But it was a cool tour you said, right?
johngrimsmo: So cool. Yeah.
John Saunders: Yeah.
johngrimsmo: Yeah, it was really, really cool.
John Saunders: I think if there's one video I would tell anybody who's listening to this podcast that they they have to watch, it's not optional. It would be the ASML video. And then was that Smarter Every Day or Mark Rober?
johngrimsmo: Veritasium, I think.
John Saunders: Very fascinating. Thank you. Forgive me.
johngrimsmo: Yeah.
John Saunders: Yes. It is probably 40 minutes long. So it's a little bit on the long side, but just not optional.
johngrimsmo: Yeah.
johngrimsmo: Yeah. Must. Yep. Really cool. How they, what is it? They shoot 50,000 streams of tin per second and shoot with a laser three times each to pancake it.
John Saunders: three times per yes a different intensities to crew yes
johngrimsmo: But yeah. to To pancake it, to atomize it and to whatever, just to make light. of which 90% of it gets absorbed by the mirrors along the path.
John Saunders: yes yes
johngrimsmo: And then like almost nothing ends up at the thing. And that makes our semiconductors today.
John Saunders: Well, and also from kind of a like, not to get all hokey with like a business lesson, but, um, they effectively failed at this for what was it like nine years and no one probably would have been able to argue with the decision to pull the plug from a capital outlay and management standpoint.
johngrimsmo: Right.
John Saunders: Like they just hit a technical roadblock that they couldn't solve. And then they eventually did crack it.
johngrimsmo: Yeah.
John Saunders: Yeah.
johngrimsmo: And it's like changed everything, you know?
John Saunders: Yep. It's fun.
johngrimsmo: I love it.
johngrimsmo: Yeah. And it makes you like hearing, it's interesting listening back to the precision microcast because they had an episode from like 2020 about the ASML machine.
John Saunders: Yep.
John Saunders: Oh, I should go instead of again.
johngrimsmo: Yeah. So it was like an older, you know, four years before that very tossing video you're talking about, but it's the same stuff they're talking about. And, and Adam and Josh just had like a really cool perspective on, that and technology and how asml doesn't manufacture anything themselves but they've teamed up with partners like zeiss for all the lenses and all these different partners to basically corner the market because zeiss will only make for them kind of thing and uh that's why they're the best is because they've created the best partnerships and you can't just start up a factory with any amount of money and do what asml has done because they're teaming with the best partners in the world that nobody can beat
John Saunders: Oh.
John Saunders: Yes.
johngrimsmo: Like, it's really, really smart.
John Saunders: Yeah. Yeah.
John Saunders: Okay. That's a segue to my trivia that I wanted to throw out today. And I'll, I'll ask the question that maybe like if you listeners can pause if they want to take a stab at this, but can you name commercially available CNC machine manufacturers whose machines exist because they originally just wanted better machines for their own product creation and internal factory use.
johngrimsmo: I can think of two.
John Saunders: So pause it for listeners now, and then John and I are going to, and I cheated. I'll tell you what I came up with, and then I'll i'll read the list that Jack came up with, but go for it.
johngrimsmo: Okay. Yeah.
John Saunders: Who you
johngrimsmo: I mean, off the top my head, there are two that I've said many times. um Brother for their, I guess, sewing machines, but also whatever else Brother Printers used to make. And then Citizen for their watches.
John Saunders: Yeah.
johngrimsmo: Those are the two.
John Saunders: Yep.
johngrimsmo: I can't think of any others.
John Saunders: You can, you know, you can't.
johngrimsmo: Probably. i mean, are going to say Kern? No. Kind of?
John Saunders: Yeah.
johngrimsmo: Yeah, probably. Yeah.
John Saunders: No, hundred percent.
johngrimsmo: 100%.
johngrimsmo: Yeah, because they had their their job shop forever, decades, and then they made a better machine.
John Saunders: Yeah.
johngrimsmo: You're right. Absolutely.
John Saunders: Cause they made up a better machine for that customer's products. And then that customer is a very tight relationship. And then I think the customer was ultimately like, would you be willing to let us also run them here? And then that just kind of, i mean, a very short version of story, then that morphed into like, we'll just make them commercially available.
John Saunders: And then there's, that's why there's K1, K2, et cetera, et cetera.
johngrimsmo: Yeah. Exactly. And these are companies that did not start out to make machine tools.
John Saunders: Right.
johngrimsmo: um They started out for decades as doing other stuff.
John Saunders: Right.
John Saunders: Yeah. So the others that I did not get, um Star is apparently similar to Citizen where it um was doing machines for watchmaking.
johngrimsmo: Okay.
John Saunders: This one I had never heard of, so this could be hallucination, but Datron was electronics manufacturing company and required their own machines to do that sort of stuff, which makes sense, right?
johngrimsmo: I'd buy that.
John Saunders: um
johngrimsmo: Yep.
John Saunders: It gave me list of some others that I didn't really believe, so I'll ignore them. And then the only other one I wanted to throw out, which is which is actually not, it's It's a cool fact for folks that are in the firearms world, but I don't think it qualifies to the actual question, but as H&K or Heckler & Koch, I am told, and I saw pictures at one point, that they, I think, filled a good portion of their factory with their own vertical machining centers that they designed and they made, but they never sold them.
johngrimsmo: Mm-hmm.
John Saunders: um So I understand that it's kind of the spirit, what I mean, and that it's different if you just build a couple of customized one machine here and there, but like to build a whole like factory of your own stuff is pretty interesting.
johngrimsmo: Yeah, it's ah it's a big detour.
John Saunders: Yeah.
johngrimsmo: And actually listening to the prison microcast, they talked about Inolite, the diamond lathe manufacturer.
John Saunders: Yeah.
johngrimsmo: they I'm pretty sure I heard Christian say that they um started off doing diamond turning and needed better machines.
John Saunders: Right.
johngrimsmo: And this was only late in the past 10 years.
John Saunders: Right.
johngrimsmo: like This is not ancient technology.
John Saunders: Yeah.
johngrimsmo: And in the past 10 years, they kind of looked at the other markets on the machine, the other diamond turning lights. And then they're like, I mean, i think we could do better. And as far as I know, they kind of have like, it's pretty cool.
John Saunders: Yeah, right? Yeah.
johngrimsmo: And, and that kind of concept of like a young scrappy team that just says, you know what, we're going to, we're going shake it up a little bit. And it obviously is not cheap to do this kind of investment in technology, but, uh,
johngrimsmo: you know Hopefully it pays off over time. But it's really cool just to be able to kind of what Cyrus is doing, actually. He's he's building machines for Atomic Semiconductor, um using everything he's learned and everything he is learning and continues to learn and just gets to like, I think I can do it better.
John Saunders: Oh, for a topic.
johngrimsmo: And it's so cool.
John Saunders: Are you allowed to, for what you learned, are you allowed to comment on how much they are actually buying raw material in-house versus the Kern or ASL model of just relying subconvents?
johngrimsmo: um uh they definitely contract a lot of parts they had their own small machine shop that i got to see just with a few hosses and things and things like that which gives them you know ideally same day turnaround time but they've also got suppliers all around the country that give them you know less than a week turnaround time probably friends of ours actually definitely friends of ours because cyrus is the one contacting them and uh
John Saunders: Yeah.
John Saunders: Sure. Right, we right, right.
johngrimsmo: So definitely contracting out. but And then Cyrus just basic basically gets to say, okay, I'm going to buy spindles from precision, what's it called?
johngrimsmo: Precision instruments, like airbrang spindles from there because he used to work there.
John Saunders: OK.
johngrimsmo: And like he's he's so knowledgeable and has his fingers everywhere that he can just, that's the answer for that.
John Saunders: Right.
John Saunders: Mm-hmm.
johngrimsmo: And that's the, end I'm not going to reinvent that wheel because they've already figured it out. And my job is to make this new machine, not every component of that machine. Yeah. And it's a very noble like way to do it. It's really cool.
John Saunders: All right, thanks for indulging my side quest.
johngrimsmo: Love a good side quest.
John Saunders: What's next on your
johngrimsmo: I live off of side quests.
John Saunders: list? Yes.
johngrimsmo: Alright. Toilet bowl float.
John Saunders: yes
johngrimsmo: So I will make a video on this because I think it'll be good good content. Do you remember a couple of years ago, i ended up wrapping on the Kern, the coolant level floats in the transfer tank. I wrapped a yogurt tub around one of them.
John Saunders: I remember lots of frustrations about that sensor. I don't know if I remember the word yogurt.
johngrimsmo: Okay. Yeah, I did end up making a YouTube video and there's literally like a white yogurt tub wrapped around a a float sensor because what happens is the floaty ball, it's a small stainless steel ball that slides up and down on a shaft, chips would get around the interface between the two and it would false read and like e-stop the machine or worst case, not read the coolant properly and flood.
John Saunders: Okay.
John Saunders: Yeah.
John Saunders: Mm-hmm.
johngrimsmo: um
John Saunders: Yes. Yes.
johngrimsmo: And i'm pretty sure Kern has since stopped using that sensor and used a different sensor, which they sent me that different sensor um chips wrap around that still cause false readings. And then i upgraded to, I called the sensor manufacturer and I was like, here's my problem. and they're like, oh, you need this sensor. So I bought that sensor and then still continue to have problems with small chips,
johngrimsmo: wrapping around the sensor where they shouldn't be and either causing a flood or causing a high e-stop alarm, like high level e-stop.
John Saunders: Yeah.
johngrimsmo: So I'm going nuclear on this while also simplifying it completely um and solving it for good using a McMaster parts, but a toilet bowl float.
John Saunders: hundred percent 100%, right?
johngrimsmo: So this gives you a rubber ball that's floating in the coolant.
John Saunders: Like...
johngrimsmo: I don't care how many chips wrap around it. It might get slightly heavier and alter the buoyancy, but it's it's not going to have a critical mass that's going to be any more than just like some chips, you know?
John Saunders: Can chips, I mean, they won't be hot by that point. So how could they even... maybe it's
johngrimsmo: It's the small little floaty ones that just kind of stick to whatever's around.
John Saunders: Yeah.
johngrimsmo: Yeah, so the weight's not critical, but those chips will alter sensors that are near the coolant.
John Saunders: Got it.
johngrimsmo: So I'm like, sensors shouldn't be anywhere near the coolant. They should be above the coolant. So it's basically a floaty ball with a vertical rod. Like in a toilet bowl, there's the arc, the pivot point, whatever. So I'm just going to do it vertical.
johngrimsmo: So basically a 12 inch long quarter inch diameter rod that's just going to float up and down with the coolant.
John Saunders: Yep.
johngrimsmo: And then I'm going to have four inductive proximity sensors measuring a ring that's just going to do it.
John Saunders: Mm-hmm.
johngrimsmo: um
John Saunders: Yep.
johngrimsmo: And because the Kern needs latching logic, like when it hits the lower switch, the pump turns on and it stays on until it hits the upper switch and then it latches off. So I ended up buying a click PLC based on something you said a few months ago and it's stuck in my brain.
johngrimsmo: And i was like, oh, click PLC, let me look at that. Automation direct, 150 bucks, done. So I played played with that over the weekend at home.
John Saunders: Yep.
johngrimsmo: I've got it wired up. I got it to blink LEDs on sensor inputs. And it's actually really, really freaking sweet.
John Saunders: Ooh.
johngrimsmo: like
John Saunders: Right.
johngrimsmo: I finally understand PLCs and it it only happened once I have one in my hand and I get to play with it. And that's when everything starts to click. And that's a common theme I'm seeing with my life is no matter how much I research, no matter how much I read about it, look about it, talk to chat, do whatever, until I get my hands on it, i don't really, really get it.
John Saunders: You just got some buy-in.
johngrimsmo: Exactly. And now this weekend, I got that with the PLC. And I mean, it's basically like I've used Raspberry Pis, I've used Arduinos a little bit. It's like all of that, but so much simpler. And just, I understand what ladder logic is now because I've seen it. I've written my own.
johngrimsmo: Like, I didn't really know until Sunday. like
John Saunders: Yep, yep.
johngrimsmo: So it's cool. And it's just basically, for those who don't know, it's basically just like a programmable logic controller that just says, if this, then that, but but over and over and over repeats thousands of times per second until the end of time.
John Saunders: that Yep, yep.
johngrimsmo: And one thing that Chad explained really cool was, you know how does it differ from a Raspberry Pi was my question. And they said, well, wait in a Raspberry Pi, you have, I'm making up numbers, 90% of your CPU dedicated to running the Raspberry Pi.
johngrimsmo: And maybe you get...
John Saunders: and ah operating system Yeah.
johngrimsmo: Yeah, maybe you get like 10% for the logic stuff. Whereas with a PLC, you get 100% of the CPU in the operating just running the ladder logic. And it's like simpler and more reliable, more robust, and just do the same thing over and over again.
johngrimsmo: And it was saying for a small short program, it might run like 2,000 to 5,000 times per second where it's cycling through the ladder.
John Saunders: yeah
johngrimsmo: And then as the program gets bigger, I assume that slows down a little bit and you get less per second. But like, oh my gosh, that's that's a lot.
John Saunders: Awesome.
johngrimsmo: So it's cool. I'm pretty happy with that.
John Saunders: it or Did you hook up the float sensor in the current yet or is that still a...
johngrimsmo: but That's next.
John Saunders: Yeah, sweet. That's awesome.
johngrimsmo: Yeah.
John Saunders: um I am also, ah you might, I wrote down a note there to figure out if, if Cliff PLC has an integration with Home Assistant. no idea if they do, but, um, I am like kind of one of those personal notes between me and you, but like one of my things I always worry about is I get super, super excited about a project and then I don't necessarily follow through.
John Saunders: i don't, you know, do enough.
johngrimsmo: This is a safe space, private conversation.
John Saunders: Yeah, exactly. he's Exactly. And I'm really proud and happy to say, actually, hold on a second. There's a person at the front door that doesn't know. need to message. Sorry guys. Um, scrap pickup is here.
John Saunders: Is here. Guy is walking in parking lot. Sorry about that. um The, I'm very proud and happy to say that I have like made leaps and bounds of progress on home assistant, both at work and at home. um Some highlights are home assistant has very,
John Saunders: ah Nice integrations. i don't need know that either one is technically official, but integrations Bamboo and Tesla. So I now get an email alert anytime my Bamboo um needs attention. So like, because we just got our seventh printer.
John Saunders: So like if it's not a filament or there's an error, we just get notified now. So we can react instead of like, oh, 30 minutes ago, something happened. And then I get an email every night at 9 p.m. if my Tesla battery is below 45% and it's not plugged in.
John Saunders: Like, love it. Like, that's just such a fun life hack around that.
johngrimsmo: yeah.
John Saunders: um
johngrimsmo: It takes the thinking out of your brain. Cause I have to, when I jump out of my Tesla, I have to like, make sure to check the battery before I go inside. Otherwise I'm going to come into, you know, at a less than 20% battery. And the one time I have to go on a trip, that's the time.
John Saunders: Right.
John Saunders: Exactly. Um, my home office, uh, floating led lights and some shelves that i built did it all myself i'm proud of built the shelves myself out of wood.
johngrimsmo: Right.
John Saunders: So it was really unfortunate. i had to use, do some woodworking, but I, it's like just a step above lathe work.
johngrimsmo: It is. Yeah. Sorry. Sorry to hear that.
John Saunders: Um, and the home assist, the LEDs are hooked into home assistant. So when I unlock my computer, the lights turn on when I lock it, they turn off.
johngrimsmo: Okay.
John Saunders: Um, The bathroom flood sensors now, I replaced the just buzzer sensors. They're now home assistant sensors, so we can get email and text alerts. And then finally, the Brother has, ah sorry, the Brother has three Reolink cameras.
John Saunders: The Willman has one. The three Reolink cameras on the Brother, one is aimed at the robot, one's aimed at the infeed, one's aimed at the machine tool itself, like Spindle. We probably won't need that many long-term, but it's been wonderfully helpful to have it plus Frigate to basically say, hey, when something goes wrong, we now just have immediate replay to look through.
John Saunders: Okay, it's a it was some scale on the material caused one of the two gripper pads to not work, or we just had a little problem with the robo wrench and now we know.
John Saunders: And to have those, oh yeah, yeah.
johngrimsmo: And you can see you can see that kind of detail on cameras. Sick.
John Saunders: So it's been, it feels great to have that done.
johngrimsmo: Nice. Yeah, and these kind of things take time. Like, how long have you been playing with a home assistant?
John Saunders: Oh, probably over a year, but really it was the first five months of the year was a lot of stuff going here at the shop with like new machines, horizontal, all that.
johngrimsmo: Yeah. No, but that's how it happens.
John Saunders: Yeah.
johngrimsmo: Yeah.
John Saunders: And it just didn't feel like i had the time to like, I didn't feel, I didn't feel like I was doing the team, the right service. If while all that's going on, it's like, Hey, I got some home assistant stuff working.
johngrimsmo: Yeah. And they're like neat, but I'd really like help with this.
John Saunders: Now is the time to be doing it.
johngrimsmo: i fall into that trap too. But some of these things take, take, you know months and take a year to like really, really get past those, you know, initial jitters and like bumps in the road. And you got you gotta to play with it for a bit and wrap your head around it And then you got to put it away and like like ignore it for three months and then think about it in the background. And then you know when you come back to it, you're all excited again and you can have another like huge breakthrough. and I find that's how these projects just go. It's frustrating sometimes, but it's like, yeah, um I'm coming to terms with it. Yeah.
John Saunders: I'm also doing a good job at saying no. Like there's so many cool plugins, like Home Assistant even scanned your network.
johngrimsmo: Yeah.
John Saunders: and It's like, oh, printer plugin and this and that. I'm like, nope, let's just stay focused. What do I need this to do? Not bells and whistles that create lots of noise.
johngrimsmo: nice
johngrimsmo: um think i talked to you two weeks about the bloom laser being funky with uh air quality going in and things like that so i have all the parts now the bloom new dock just came in so that's the uh they call it the smart dock it's what mounts underneath the laser basically air and power goes into the smart dock and there's little pneumatic valves inside that like say, send air to the tool blast, send air to the shutters to open the laser, things like that.
John Saunders: Yes.
John Saunders: OK.
johngrimsmo: That's gummy on mine. And Bloom bloomm confirmed the theory that I had, or they think I'm right. um So I bought a new smart dock and then just came in today. So we'll be installing that.
johngrimsmo: And then i and I bought a five position air filter system that I'm putting just for the laser um type.
John Saunders: Oh, wow. What you five position?
johngrimsmo: There's a five micron pre-filter and then there's a condensation somewhere to forget.
John Saunders: Coalescer?
johngrimsmo: Yeah, yeah, probably coalescer filter and then something else and then two ah carbon filters.
John Saunders: Oh, wow.
johngrimsmo: So like total overkill, but yeah, yeah, like five cans.
John Saunders: Do you really mean five stage? Okay, yeah. Got it.
johngrimsmo: Um, all from McMaster, all like not expensive, 60 to $80 each. So for like 400 bucks, I'm, I have the best air, like laser quality air. I'm like, sweet.
John Saunders: Yeah. right, right, right.
johngrimsmo: And so part of me did it just to like prove to myself, okay, now I know how to feed clean air to a laser.
John Saunders: Yeah.
johngrimsmo: So if in the future I ever need to like build an optics lab or something like that, like now I know how to do that. And I have a physical example in my shop of like, I've done it. but also to give us long-term reliability on this laser, because we're literally trying to hit 10th on hard milling on the speedio.
John Saunders: rare
johngrimsmo: And I just, I just need it to be perfect. I just need it like for 400 bucks done. Um, plus replacing the smart doc as opposed to trying to rebuild it myself or like whatever.
John Saunders: Yeah, yeah.
johngrimsmo: Like once I take the old one off, I am definitely taking it apart just to see what it looks like. But, uh,
John Saunders: Yeah, kind of annoying. i don't feel like that machine is that old. You know what i mean?
johngrimsmo: four years old, but, but it's, I guess it got dirty air, my fault, I guess, sort of. um And I just needed it to be right. So happy to fix it properly. And i'm I'm happy to even put like maintenance schedule on these filters and be more proactive than reactive kind of stuff. So it's it's teaching me a lot of lessons, which I'm really appreciating. in And hopefully and in these conversations and what I have with my team, we can all learn from some of these things where,
johngrimsmo: you know Some things are worth and making them matter and going the extra mile to like add add total process reliability.
John Saunders: Yes.
John Saunders: Yes. I've had a, um, I've separated out some of my, like,
John Saunders: you know, year end stuff, regular notes, things to revisit. Like have all these, and I'm assuming other people do too, like notes, things that like impactful to you, blah, blah, blah. And I separated some of them out. So I like some stuff I revisit once a month, sometimes it's once a year.
John Saunders: And that's been good. Cause I don't, it's something that's once a year. I don't want, if I look at it every month, I get annoyed by like, stop.
johngrimsmo: Yeah.
John Saunders: But one of the things I had written down was, you know bootstrapping is okay. Sorry, bootstrap is not okay if you have the money and it isn't happening. And then and to your point here, like like, yeah, clean air is, like, we can debate it all you want, but the reality is you're just better off for it. and And to know that you can spend hundreds of dollars and get exceptionally good reliability, totally good.
johngrimsmo: and And to know that that's not required shop-wide. like Like you said, there may be benefit in putting it first, um but it's also not like totally required for every station to have that level of cleanliness right by the device.
John Saunders: Yeah.
John Saunders: Mm-hmm.
johngrimsmo: So it's it's learning. But even for our anodizing station, I think we have a probably not that great filter, but we do have a filter. ah When the guys are anodizing and they need to blow off the parts,
John Saunders: Oh, sure.
johngrimsmo: And I want that air to be clean because if they're spitting water dirt or whatever, like that's not good for anodizing.
John Saunders: Yep.
johngrimsmo: Now I want to revisit what they've got called on there.
John Saunders: Yeah, right.
johngrimsmo: But yeah.
John Saunders: It's funny. ah Speaking at lasers, I have Keon coming in tomorrow to
johngrimsmo: Oh, you give them a fake email.
John Saunders: um No, they actually, we had the guy come in to demo that like measuring machine that they make six months ago.
johngrimsmo: Okay.
John Saunders: the Um, the ICL, ICX, I kind don't know. Um, but we want to look at a non-contact laser. i actually brought this up on the podcast.
John Saunders: I have some notes.
johngrimsmo: Yep.
John Saunders: I think Danny Rudolph reached out to me, which I appreciate. it I need to go dust those off. This all happened like literally this morning, but, um, um Grant, to his credit, we have this, sum is it Precision Metrology is a company. Justin Gray also knows them and has been, they sent him some stuff that he's put up on social.
John Saunders: Okay, the guy sent us a real nice letter and included ah a free indicator. which is like, we don't get that stuff anymore, which is, so was kind of like fun and it seems like a good dude. And then what's really cool is that Grant wanted the indicator to work a little bit differently as a digital indicator. And they like rewrote the firmware and reflashed it. um But it's still the wrong tool for what we're trying to do, which is, in short, just think like manufacturing a half-inch gauge pin type product.
johngrimsmo: Okay.
johngrimsmo: Yeah.
John Saunders: Like, we're take it off the the mill um and have the robot present it to a measuring device and then either you know log it or potentially recheck it if it's out or put it in a wrong bin.
John Saunders: That's sort of obvious PLC logic-y type of problems.
johngrimsmo: Yep. Yep. Yep.
John Saunders: and what grant built is really cool it's a miniature high quality linear rail that has v-block the robot sits on there the robot slides it in we've done this but you still have multiple sources of potential error as you're starting to measure products in an automated environment with chips and coolant over time where it's just not the right way to do it.
johngrimsmo: Yeah.
John Saunders: And it's it's a good example of like the bootstrapping, like you you need to get a non-contact measuring system. And so the one we, I know a little bit about the Mitsutoyu laser scanning micrometer. I'm really hoping there's an option that's a little bit less expensive than those.
John Saunders: but now you're talking about the ability to measure tenths or sub-tenths every time, non-contact.
johngrimsmo: Will that work on oily parts? Are you assuming the parts clean or?
John Saunders: Let me get back to you on that.
johngrimsmo: Yeah.
John Saunders: I don't know the answer to that. um We will probably, the parts, we want we want the part relatively clean anyways. Yeah, let me get back to you.
johngrimsmo: I actually seem to remember um a tour of Mitutoyo Japan. I think it was when Titans guys went there and they showed some of their in-process inspection of their Mitutoyo, like measuring tools when they're making them.
johngrimsmo: And I seem to remember there was a lot of really interesting in-process stuff that they had going on, which doesn't surprise me, but it's like cool to see. Yeah.
John Saunders: Yes, 100%. I've kept some pictures over the years of factory tours. The other thing I'm trying to make right now is, ah when I say make, like a jig that supports commercial off-the-shelf hand tools, we're not I'm not trying to make the metrology aspect. But um when we make half-inch parts that we make, square prismatic parts, not round parts, um If want to check the flatness in three points, I want to have gate instead of using a hand mic, we're going to get, I think they call them snap gauges, but like they're Mar makes them with Mar, like the millimus, which are super great.
johngrimsmo: Yeah.
John Saunders: Or you can just build a taco fixture, if you will. And you just drop the part in there and it measures the thickness at three points pinching it.
johngrimsmo: Okay. Yeah.
John Saunders: So that way,
johngrimsmo: Yep.
John Saunders: We also care about the flatness, but I don't, I want to isolate out the flatness aspects. so I don't want to have it on a, just a granite surface and just check the top measurement, if you will. And so designing some of those, um, has been top of mind and in quite fun.
johngrimsmo: Mm hmm.
johngrimsmo: Yeah, cool. Yeah, I've thought about that for our blades too, because we do a lot of, we do our own heat treat, we do our own lapping and surface grinding and the flatness of the blade does absolutely matter.
johngrimsmo: And we check it visually right now with a light gap on a flat gauge block and we look at the light underneath it and we can tell if it's cupped or or
John Saunders: Yes.
johngrimsmo: and then will either run it differently or bend it accordingly. But I've definitely thought about having some sort of flat plane with you know three sensors underneath. and You just slide the part over it, and it hits three little indicators or whatever, and it displays something.
johngrimsmo: But I haven't gotten very far on that theory.
John Saunders: Dude, use your software, take it, get a camera lined up in your current light box set up and just have your software process the image of viewing the light below the light gap.
johngrimsmo: That's interesting.
John Saunders: It's actually like not hard these days.
johngrimsmo: That's and a neat idea. Yeah.
John Saunders: Right?
johngrimsmo: Yeah.
John Saunders: Hmm.
johngrimsmo: That's interesting. I don't know if it would give me what I need, but that is, i hadn't thought about that yet. That's kind of cool.
johngrimsmo: Um, let's see what else.
johngrimsmo: to Last two more things on my list, uh, foam. So I might've mentioned about a month or two ago, Leif and I went up to our foam company and, um, looked at switching type of foam that we use for our packaging.
John Saunders: Yes.
johngrimsmo: And, you know, months ago, I talked about injecting my own foam and like playing with that whole smooth on kind of thing.
John Saunders: Mm-hmm. Huh.
johngrimsmo: um That's been shelved and the guys actually boxed up and taped a box and put it in the corner and said, John's foam stuff. and i was like, okay, fine. I guess it doesn't need to be out. I haven't touched it in six months.
johngrimsmo: but But yeah, so we got sheets of this new foam, different brand, different consistency. They make it differently. It's actually got nitrogen bubbles inside of it, which is kind of interesting.
johngrimsmo: But ah so we were able to cut a few test pieces the other day and it looks amazing. There's some differences, but um the big hurdles from the previous foam was that it would gum up on end mills.
johngrimsmo: And then the end mills would get like sticky and then they would smear and then they would make bad parts.
John Saunders: Right.
johngrimsmo: And then you have like, like go in there with a piece of wood and chip off the end mill every, you know, 10 parts or something, which is stupid.
John Saunders: Weren't you running the end mill against wood every 10 parts as part of a cycle?
johngrimsmo: Yeah, and then we we stopped doing that a couple of years ago.
John Saunders: Okay.
johngrimsmo: I'm not sure why, but it probably wasn't worth the squeeze kind of thing.
John Saunders: Got it.
johngrimsmo: um And were just consuming wood, like, annoyingly.
John Saunders: Yeah.
johngrimsmo: Which is fine, because wood deserves it. But... um But yeah, so we tried this new stuff and it's great. No loading of the tools whatsoever and great finishes.
John Saunders: Oh.
johngrimsmo: It actually looks really cool. And then now what we have to figure out, I'm calling it spaghetti, the web in between parts that tends to become these long pieces of spaghetti and then wrap around the tool.
John Saunders: God.
johngrimsmo: And then it's not welding itself to the tool, but it is wrapping around the tool.
John Saunders: yeah. yeah
johngrimsmo: like like a bad drill chip kind of thing, and it's taking out the next piece of foam or like causing a problem. So one theory that actually heard Josh talk about on the Precision Microcast with the wire EDM, he was complaining about slugs.
johngrimsmo: When EDMs a hole, slug falling out and like causing a problem.
John Saunders: Yes. sure
johngrimsmo: And he said, even though it takes longer, I just erode the entire slug away now. And he's like, process reliability went from 10% to 100%.
John Saunders: Yeah. Yeah.
johngrimsmo: He's like, I don't care.
John Saunders: You can't find the wire when it's been evaporated.
johngrimsmo: Exactly. And i was like, well, maybe that's what I need to do here with the foam is just turn the entire web into chips instead of assuming it will fly away accurately in a way.
John Saunders: here
johngrimsmo: But foam is also compressible and flexible and kind of moves out of the way sometimes. And you gotta be a little strategic with how to do that. So that's where i with that.
John Saunders: Could you have the end mill and the machine to go over to a a U with a hole in it, like a slot in it and just have the end mill, almost like a tool changer, but like slides the tool and then lifts straight up because it's foam.
johngrimsmo: o
johngrimsmo: Right.
John Saunders: It's gonna succumb.
johngrimsmo: Yeah. Even if it was off, like turn spindle off, move to there, have little nylon grippers or Delrin or whatever, and then flick it off.
John Saunders: Exactly. There you go.
johngrimsmo: Yeah. But If it's repeatable, like if if the spaghetti always hits on the same spot, then I can take it off next. But if it's unconsistent, it happens randomly, then you can't.
John Saunders: Yeah.
johngrimsmo: Because basically, if it wraps around, then the next five seconds of machining are bad.
John Saunders: Oh, I thought you meant it only happening like between parts. You're saying it could be happening at any point time.
johngrimsmo: Yeah, yeah, yeah.
John Saunders: Yeah, fair enough. I hear you.
johngrimsmo: But it's a neat neat concept for sure.
John Saunders: Hmm. Reminds me back when, boy, this is now going way back with Fusion and like the 2D adaptives where sometimes you'd be left with a thin strip in the middle and then it would go down, but then that thin strip could actually cause tools to chip and all that.
johngrimsmo: Okay.
John Saunders: And they they came out with a solution. i don't remember what it was, but it just sort of made maintained the minimum amount of material that it would would it would make sure that instead of it being a whisper thin foil in the middle, it would actually have it be 30,000 thick.
John Saunders: That way it was a it would actually fly away and not wrap around it.
johngrimsmo: Mm-hmm.
johngrimsmo: Yeah, and those are the kind of fine things about machining anything that you don't think about it until they start biting you and you're like, oh man, how do i solve that?
John Saunders: Yeah.
johngrimsmo: And then like everybody deals with it if they really care.
John Saunders: Is this one of those um like high polish single flute router type things?
johngrimsmo: Yeah, high polish, Datron, triple flute, very small helix, ah very lazy helix, whatever you call it.
John Saunders: Yeah.
John Saunders: Okay.
johngrimsmo: um So not a lot of spirals and and very sharp, very sharp.
John Saunders: Yes.
John Saunders: Is it up cutting?
johngrimsmo: It is regular. So yeah, up cutting, not down cutting.
John Saunders: It's right here. Where do they make it down
johngrimsmo: They do. um And it's used a lot in the like wood industry and then some foam end mills and stuff. but But then in some way you're pushing the chip down, which can cause vacuum issues when it, if it gets under the park.
John Saunders: Yeah.
johngrimsmo: So I don't know, I haven't really played with down cutting end mills and a lot of people have suggested them. Maybe I'm ignoring them, but I think I have a spaghetti management problem more than I have a up cut down cut problem.
John Saunders: Wouldn't the the question though, wouldn't the down cut minimize the chance that the chip is pulled up into the flute?
johngrimsmo: I don't know. It's a good question. Might be right.
John Saunders: Yeah. Yeah. Cool.
johngrimsmo: Cool.
John Saunders: I sort of promised or teased an idea about how we would handle our compressors all about automation last episode.
johngrimsmo: Yeah, I actually thought about that on Saturday. Leif and I came in and ah i I was like, oh yeah, John asked me about that.
John Saunders: Okay.
johngrimsmo: And you know we come into a quiet shop and the compressor's running. And I was like, yeah, we really do leave it on all the time, don't we?
John Saunders: Yeah.
John Saunders: Do you have any thoughts or else I'll spill my...
johngrimsmo: Um, no, I didn't really, but what do you got?
John Saunders: So, well, so look, and the point here is not to be, is the goal is not to actually maximize the amount of time that compressor's turned off. It's just to like take a reasonable stab at taking a chunk out of, you know especially your quintessential three-day weekend.
John Saunders: If the automation stops Friday at 6 p.m., compressors have like 70 hours where they're just not muted.
johngrimsmo: true.
John Saunders: That's kind of what we're trying to tackle without um having to get manual intervention. yeah.
John Saunders: The system is, I think, going to be relatively simple. I'm going to have it sort of structured in two different buckets from a risk standpoint. This is over-explaining. But basically, Monday through Friday, 6 till 7 p.m., it's almost impossible to turn the compressors off. They just are going to be on.
John Saunders: But then you can have a couple of different sensors that can interrogate the situation and decide when to turn them off. But most importantly, the key is is we have one employee who does come in now, some on the weekends and do work.
John Saunders: When that happens, I have zero appetite for there ever being a situation where a machine stops because the compressors tried to get greedy and shut shut off again, like to stay there.
johngrimsmo: Yeah.
John Saunders: And so the simple answer is, how do you know if somebody's here and the machine's running? Well, machines running is easy, except what if they're you know idle for half an hour because it's a weekend and they went to grab sandwich and they just don't like, i don't want to just rely on that. um Same thing with like the lights being on or off.
John Saunders: That's a weird thing.
johngrimsmo: Yeah.
John Saunders: thing It can work, but it doesn't. But the cool one that I'm excited to to implement would be if an employee's phone is attached to to Wi-Fi. It ends up that that's a easy thing to interrogate in Home Assistant.
johngrimsmo: Whoa. Really?
John Saunders: And I asked everybody to like make sure, like, cause it's kind of one of those things like if people don't want to be on our wifi for personal reasons, totally get it. But everyone's like yeah, we don't care. So, um, I need to think this through a little bit more.
johngrimsmo: Yeah.
John Saunders: I meant to do that before this episode to, to present kind of a more robust solution to everybody. I'll try to do that for next week. But, um, you come up with a sack of criteria where basically like, okay, the compressors are default to be off on Saturday and Sunday, but if anything happens, like if an employee's phone connects to wifi, if, uh, The machine turns and starts running.
johngrimsmo: Yeah.
John Saunders: The compressors open up and then for them to shut off again, they will generate a email or an alert or something. There's a different few different things we can do. We can do do an audio message that's actually projected in the shop with a 30 minute countdown.
John Saunders: So basically to make sure, okay, and in no circumstance am I going to get caught my pants down here.
johngrimsmo: And then how do you cancel the 30 minute timer? If you're in the shop and you hear it.
John Saunders: what The simplest way would be to log into Home Assistant and and override it. ah I'll probably do something where in that email you can hit a nope, don't do it response.
johngrimsmo: Okay.
John Saunders: We might even do a physical button. um
johngrimsmo: And would it email everybody on the team? Yeah,
John Saunders: That's what I need to think through. I don't want to create a bunch of noise, which is one reason why I kind of like the idea of...
johngrimsmo: yeah. yeah
John Saunders: um of a um of a light, like a blinking light in the shop and an audio message, kind of like the, what's what's the old World War II cox horn sirens, like, hey, you're turning it off in 30 minutes.
johngrimsmo: Yeah.
John Saunders: Yeah.
johngrimsmo: Yeah.
John Saunders: Yeah.
johngrimsmo: You need like a microphone that, you know, when there's the audio message, that goes out that hears, no.
John Saunders: Yeah. yeah And the downside, which I still want to think through, the downside is I come by the shop a lot on the weekends. That will turn the air on when I come by. So I probably will think about ways to isolate that. And maybe I'll even just do that for me.
johngrimsmo: But this is that not a problem? Like,
John Saunders: It's not a problem. And again, I'm...
johngrimsmo: I kind of like that idea is like, you know, shop's quiet. And the second you open the door, the compressor starts running or you hear it running in because of your phone connected to the wifi, like unless you have wifi problems.
John Saunders: Yeah. Yeah.
John Saunders: Yeah, but I come by a lot for other reasons and that's probably only me.
johngrimsmo: Sure.
John Saunders: So maybe y I'll do something for myself where for me, they turn on or they will turn on within five minutes, unless I click a button that says no, ignore me. I'm just stopping by. Or I could, for me, I could say, i have to be in the shop for more than five minutes.
johngrimsmo: Sure.
John Saunders: Cause a lot of times i just come by and like grab a print, you know what mean?
johngrimsmo: Yeah. yeah And then the only thing I caution you on is creating too many rules and too many specific situations. You're trying to you know find all the variables and all the things that just creates a very complex system to ultimately you know hinder something that is so inherent in manufacturing, which is error.
John Saunders: Yeah. Yes.
johngrimsmo: Um, and like, I'm trying to caution myself with a lot of the projects with GURP and with things that I'm building that are like pretty complex, even the microscope projects, like this is becoming too complex to give to anybody else and like have them run it. Um, so you're just creating like a weird, like super solution that is very, it could be very difficult uh, for anybody else to come fix debug.
johngrimsmo: Um, keep in mind.
John Saunders: No, it's a great, it's great advice. And i that's one reason why i wish I had done a better job preparing for this episode to share that because that but it won't be that it'll be home assistant automations, which are very like WYSIWYG readable.
johngrimsmo: Yep, yeah.
John Saunders: Um, the, if you can, you can convert from the WYSIWYG into YAML, the YAML code can be edited by chat. And then, um, to your point, keep it simple, but then also I've done some good things and, and this is where, um,
John Saunders: LLMs and even Homes hasn't done a good job of like, it's like, okay, I want to notify my phone. Well, i noticed that when I created the YAML code to notify my phone, it was using my phone's specific name in ah in a single script.
John Saunders: And like, I'll get a new phone at some point. Everybody will. And so I realized, oh no, you need to create a separate notification script
johngrimsmo: Yeah.
John Saunders: that can be referred to that way. When you get a new phone, there's one spot where everybody can update this thing like that.
johngrimsmo: yeahp Yeah.
John Saunders: Instead of it's like, well, this script didn't notify me. Like to your point, I didn't get notified because I got a new phone, which meant I didn't realize the bathroom was flooded.
johngrimsmo: Yeah. yeah But once you update your phone number in in your system in Home Assistant, like employee page or whatever, then then you're in the new system.
John Saunders: Bingo. Yeah,
johngrimsmo: Exactly.
John Saunders: that's good advice.
johngrimsmo: Yeah.
johngrimsmo: Sweet.
John Saunders: Well, I'll save the two machines for next week.
johngrimsmo: Perfect.
John Saunders: just
johngrimsmo: They'll be even more installed by then. so Cool, man. Yeah. Tell us all about it next week.
John Saunders: What else is on your plate today?
johngrimsmo: um i've got a lot of like ongoing projects that i love and i'm trying to um kind of go nuclear on a bunch of things to get them like like good get them in a good stable spot so uh it's nice i have a probably five or six projects kind of concurrently happening and i can dive into one until i get to a point where i've ordered all the things and now i wait and then i dive into the next to the point where i'm waiting on the next person and like I'm jumping around, but I'm feeling very productive on all of it.
John Saunders: Yeah. yeah
John Saunders: Yeah.
johngrimsmo: and And a lot of them are coming to a close like in the next week or two, which is fantastic. So i feel like I'm being productive and I'm doing the right things and I'm doing good things.
John Saunders: Good.
John Saunders: Is this like the interprom?
johngrimsmo: A lot lot of which we've talked about, like and more like, like the toilet bowl float and like, like fixing actual problems that are constantly halting production or, um,
John Saunders: Yep.
John Saunders: OK.
johngrimsmo: so that hopefully that becomes a lot more stable and we have a lot more process reliability and I can step away more from like the day to day. And then I can, from there, the goal is obviously to add more, more products, more variations, more, more, you know, profit kind of thing, like more stuff to do.
John Saunders: Yeah, sure.
johngrimsmo: so I need to make everything current kind of as smooth as possible so that then I can stress the system by adding more and then we grow accordingly.
John Saunders: Sure.
johngrimsmo: So that's where I'm at.
John Saunders: Yeah.
johngrimsmo: You.
John Saunders: Ironically, the same topic, which I'll save for next week about just like, I'm really happy that we're finding all these like hiccups and flaws in automation systems. Like Grant walked eight minutes before and was like, this isn't working.
johngrimsmo: It's kind of fun.
John Saunders: And I'm only happy because i got frustrated when I would talk to people and they would be like, nope, these systems just always work. the The robot always works. I'm like, the hell it does. Sorry, I'm sorry.
johngrimsmo: Yeah.
John Saunders: So I'll share more about those in a good way though, frankly. um
johngrimsmo: Yeah. yeah
John Saunders: And, but it's really good. Like the brother that we have running is, it's probably now been a full month of where we're, we're just figuring out how to manage it. Like, how do we handle, do we do it by day?
John Saunders: Do we do it by part quantity? When do we handle switchovers? But like, it's just, it's really, really nice.
johngrimsmo: Oh, that's awesome.
John Saunders: Yeah.
johngrimsmo: Yep. I've, I've learned that I love debugging systems and like, like finding the issues and making it ideal and make, like, I love that whole process.
johngrimsmo: So.
John Saunders: Yeah.
johngrimsmo: All right, man.
John Saunders: Yeah. Good. let's I'll see you next week.
johngrimsmo: See you next week.
John Saunders: Goodbye.
johngrimsmo: Bye.
