Embedded - 370: This Is the Whey

Episode Date: April 22, 2021

Alvaro Prieto (@alvaroprieto) spoke with us about cheese, making, work, the reverse engineering podcast, weather, and motivation. Alvaro is a host of the Unnamed Reverse Engineering podcast. Some of his favorite episodes include #41 with Samy Kamkar, #14 with Joe Grand, and #23 with Major Malfunction. (Jen Costillo co-hosts the show and has been on Embedded several times.) Alvaro works at Sofar Ocean, making oceanic sensing platforms. He has a personal website linking to his other exploits. We talked about some Embedded episodes as well: #282 with Laughlin Barker about OpenROV #174 with Evan Shapiro about baby monitors and professional poker  Also, we’ve all really enjoyed the Disney’s Mandolorian.

Hello and welcome to Embedded. I am Alicia White here with Christopher White. This week our guest is Alvaro Prieto and we'll be talking about cheese, the weather, robots, who knows really. Hey Alvaro, How's it going? Hello. Hi. It's going well. I just have to say this is very weird because I usually listen to the podcast at like 2x speed and that intro was so slow.

We'll try to match 2x during lightning round. How's that? Yeah, that makes sense. We'll ease you into it and then we'll slow down. Sounds good. Yeah, that just caught me off guard because when we were speaking earlier, lightning rod how's that yeah that ease you into it and then we'll slow down sounds good yeah that just caught me off guard because when we were speaking earlier it was normal and then the intro is something i hear a lot and that was just wrong uh for people who haven't heard you be on the show in the past could you give us an introduction oh i'm alvaro. I am an electrical and firmware engineer, and I like cheese. Yeah, that about covers the basics. Okay, we'll get into more details. Probably not just cheese,

but maybe some other things. I like other things as well, yes. Are you ready for lightning round? Absolutely. Favorite obsolete technology? Obsolete? Oh, shoot. Okay, I'm not ready for lightning round. Take that back. Obsolete, but if it was obsolete, I wouldn't.

Phone calls. Favorite kind of cheese? Latour. Describe it. That, I wouldn't. Phone calls. Favorite kind of cheese? Latour. Describe it. That's not a kind. That's a specific cheese, but... What makes it good? It's three different kinds of milk.

So you got cow, sheep, and goats milk, and it's kind of a soft, spreadable cheese, and it's amazing. So it's Tres Leches cheese? Yeah. It's a good cake favorite kind of weather um cold and sunny favorite kind of robot a working robot for swiss cheese what happens to the hole when the cheese is gone whoa oh gosh what happens to the hole when the cheese is gone um it re-enters the the universe i don't know the cosmos it goes back to where it came from where did it come from would you

bring back the dinosaurs sure which ones all of them but um okay that, okay. That I would, if I were to bring all of them back, I would bring dinosaurs from specific periods in separate areas. I wouldn't want to mix them. Cause I know that not all the dinosaurs live together. So yeah. So we'll have different of the Jurassic park islands for each different period.

Okay. Do you have a tip? Everyone should know. Um, don't give up. I just, just try again later when yeah try again later 404 that's what i'm doing okay so you have been doing cheese and there's water robotics involved in jobs and there's some electrical things with a weather station but first you you not only are cheating on us with another podcast you started your own podcast i i don't cheat no um i've only

been in a few others but uh i think this was my first podcast that i was on uh 2015 so you know i'm just saying but but yeah i i do have uh the unnamed reverse engineering podcast with uh with jen who's also been on probably one of the first podcasts i guess right oh many of the first podcasts yeah what's the show about uh reverse engineering we uh talk to people that are much smarter than us Yeah. What's the show about? Reverse engineering. We talk to people that are much smarter than us about reverse engineering topics and just kind of taking things apart. And not just hardware, but a lot of it is hardware taken apart, like reverse engineering silicon, web protocols, like all sorts of stuff. The most recent ones, we talked to folks who make the tools for reverse engineering. So the open source ones or some commercial ones. But yeah, we kind of talk all things reverse engineering.

Okay, what's a recent one that you're super happy about? Oh, Sammy Campcar came on and did one about the the firewall uh what was what was it called uh but basically you go to a website and it can open up any port on your firewall from the outside which is pretty neat neat yeah slip streaming slip stream yes nat sleep slip streaming yeah he he came on to talk about it like a little bit after he published it, and since it's gotten a lot scarier.

Other people figured out ways to do more things with it. Yeah, so that was fun. I've also gotten to talk to some of my, I won't say childhood, but when I was younger, heroes of security, computer security, or like Joe Grand was a really fun one. Or I talked to this one person, his handle was MajorMileFunction, who I saw at DEF CON in Vegas back in 2006 when I was 18. And I saw a talk by this person, and it was really cool about hacking Mac stripes. So on train tickets, on credit cards, all that stuff. And I sent him an email

and he replied. And back then I was like, oh, this is super cool. This person actually is helping me out. And many years later at hardware.io, which was a hardware kind of security conference here in the Bay Area, he happened to be there and I didn't realize because he was going by his real name, but I got to interview him for the podcast, which was pretty cool. How far were you into the podcast before you realized you knew this person? Oh, no, no. I, I, I knew, I didn't know. Thankfully I didn't know the, uh, when we started the podcast, but I was in the conference already. And then somebody mentioned the name of like, wait a second, like, that's this person.

And then I went and asked him if he could be on the show. Cool. What have you learned from the show that, you know, you didn't really expect to learn? Whoa. I don't think about these things. I just like to talk to people and it's fun. But what have I learned? I've learned there's way too much fun things

and not enough time to do them. I learned about all these really cool technologies and techniques people use for kind of taking things apart or learning about things. And I always want to try them, but there's never enough time. And just hearing the stories of how people ended up and figuring things out is always a ton of fun. And I'm blanking out, but if I look at the list of episodes,

I could tell you some really cool things that I've learned. It's hard to do it on the spot. On the fly, right? Yeah. Okay, so what are you doing for work these days? I am doing firmware again, finally, after kind of doing all sorts of other things. But I'm doing firmware for Ocean remote sensing buoys. What else can you tell us about it?

It's a lot of fun. So it's a company called SoFar Ocean, and folks might know the name OpenROV, which was kind of a Kickstarter or something back in the day that they built like an open source underwater kind of rover or drone or whatever you want to call it. And then that company later on merged with another company called SpoonDrift Technologies, and then they formed SoFarOcean.

And right now I am working on the next generation of kind of remote sensing buoys. And these are buoys that you, well, we have a couple types, but you basically throw them in the ocean and they just float in the middle of nowhere and measure like sea surface temperatures and wave heights and other things

and then send the data back over satellite. But you can also buy them if you're an individual or company for research and deploy them kind of off the coast. Or there's a company that's deploying them to monitor coral reef water temperatures and stuff. And then you can get the data back over satellite. I mean, through our API or whatnot. But, yeah.

I was going to ask who the typical client was. Is it like NOAA or private companies? Yeah, there's different ones, right? So there's private companies. And so we sell the booties themselves that you can put somewhere, wherever you need to get the data. And you can moor them or put them in a static location. Then we also have our own that are just drifting across the ocean. That's why spoon drift back in the day. And those we can just sell the data. But another thing we're doing is using that data ourselves to

create a much more accurate weather models. So NOAA releases their own and then we can augment it because we have data points that are all over the place, and we use that to do kind of ship routing services. Oh, okay. Can you look for tidal waves? I think the science people could answer that. I don't know. Hopefully?

I don't know. Yeah, I'm not sure. But there's a lot of kind of grants with the, I don't know. Yeah, I'm not sure. But there's a lot of grants with the, I don't know who does them, but to do research of these kinds of things. So I'm trying to figure out what I can talk about and what I can't. But they are doing wave measurements, and I don't know if it's sensitive enough to just catch the one rogue wave kind of thing is that what you mean yeah i wonder i would have to ask and actually the the cto has been on the show before um back when he was working at a baby monitoring startup i think yeah evan shap. We talked about baby monitoring and I think using machine learning

to win at poker. Yeah. It was a weird joke. But it was pretty cool. Well, yeah, he's the CTO at Sofar now. Is there anything about engineering for the ocean that people don't necessarily realize

right off? I hope most people realize for the ocean that people don't necessarily realize right off i mean i think i hope most people realize that the ocean is incredibly harsh in all sorts of ways um like salt water will corrode everything but that's not the worst part then critters all sorts of things just start growing on every surface you can imagine and starts growing and growing and growing and then kind of takes over it. You can't get your things back because they're in the middle of the ocean and debugging can be exciting when you can never get it back. So there's a lot of interesting problems. Yeah, but you don't have to worry about rock slides or,

you know, hard things running into it, except for maybe, I guess a boat could hit it, but yeah. Icebergs. I, hopefully those come slow enough that it would be more of a,

is there any reason why that was the job you took? Oh, well, so, um, previously, so previously, so let me recap what's happened since I was last on here. I was here right after I quit Apple, and I think it was at Planet Labs, right? Maybe? Yes. Well, that was the first time I was on.

So I joined Planet Labs, and like six months later, I quit because it wasn't a great fit. And then I took what I call my self-funded sabbatical. It was originally supposed to be like a month or two just to kind of recap, like get my act together and find a job, but then it was too much fun. So I just started traveling and meeting up with friends, that kind of thing. And then I joined a company called Verily,

which used to be Google Life Sciences with Ben Cross and I was there. and it was a lot of fun. And I was there for almost a couple of years. And then I kind of did a switcheroo to Project Loon, which was also within Alphabet. So it was kind of like switching jobs within the same company. It wasn't, it was different companies, but it was all still under Alphabet. So I went to Project Loon, and then I was there doing ground station stuff for the balloons. And last year in March, I had the brilliant idea of quitting my job to go traveling. I didn't have that idea in March, but my last day was like March 4th or something. So I quit my job, and like two weeks later, I was locked down

in my apartment. So that didn't work out as expected. And you know, that's like for everyone else as well. And so yeah, I didn't really have a job. I was just kind of not doing anything, working on projects at home, watching lots of movies, walking around the hills. And then a recruiter from Sofar reached out and it seemed like a really interesting project. And for me, my favorite part was the chance to work on something from scratch because I'm working on the next generation of the Buoy. And it was a full redesign. So do the processor selection, start the firmware.

I don't want to say from scratch, but a lot of it from scratch, because it's an entirely different system. You know, some of the algorithms will port over, but the underlying system is totally different. So it was a chance for me to kind of do the whole thing. And, you know, I got to play with FreeRTOS. Well, I ended up choosing free RTOS, but I had to choose like which RTOS we want to use.

Do we want to use an RTOS? Which tool chain do we want to use? And then going through the whole process of like writing the drivers, optimizing power consumption, debugging, all sorts of different things. So it was a really fun puzzle, if you want to call it problem for me

that I've never done the entire or never owned that entire thing. I usually come into a project that's already started and somebody already made all of these decisions and then I just have to live with them. And this was an opportunity to do that. And also it's really cool that we're doing remote ocean measurements and stuff. I like both the company and the opportunity that is the most it can be really fun to come into a company and and be able to kind of start things from scratch because you know the mistakes people have made before so you feel like you can kind of skip some of those and you know what things are important what aren't yeah i'm sure i'll make some myself right but at least i least I can not do what I've had other people and be like, oh, this is stupid, and then I don't want to do that.

So now I have a chance not to do that. I've noticed that you've worked at space companies and air companies and ocean companies, but you need to do one more company to own all kinds of things. Do you need an ag tech company? A land company of some kind. Mining or ag tech?

I've worked in trains before. Technically. That's okay. But worked in trains? Does that mean you were working while on bar? The trains went on the things that I worked on. So my first internship in college was at a company called Railcom, and they did remote track switching equipment.

So I was writing the code that communicated with the radio and actually moved these giant track switches, like did the actuation. And somehow my code made it out to like lots of csx train yards so yeah that was that was a fun one so i have worked i guess in sort of it's like the embedded version of an e-god right air land sea space you gotta catch them all haven't done underground stuff i guess um seismic i don't know what i would do we just saw godzilla so they're those tunnels that go to the center of the earth so it's probably some stuff you can do there godzilla versus hong versus king kong and versus hong kong actually

i've seen i've seen the other ones okay so work aside because we're not quite sure what details I can ask you about that. Oh, you can ask. I just don't know if I can answer. Right. That'd get boring real quick. Oh, it's just going to be fun for Chris during editing, right? We can talk. Whatever.

It's great. I'm not going to get in trouble if I don't edit it out. Cheese. You've been. Let's go somewhere where there's cheese. Let's go somewhere where there's cheese let's go somewhere where there's cheese um so i guess i guess uh wait i should have asked him if the moon was made of wensleydale wasn't it stilton i don't know anyway you've been making cheese um do you know that you can buy that in stores? I've heard.

No, I do buy it in stores because they have really tasty versions that I can't make yet. Take us through what making cheese is all about. Like, I don't really understand. The most I've seen about making cheese is when I your, your short Instagram live broadcast of making bread, which was interesting. It was very interesting. Sorry. Yes.

It seems like it's a lot of biochemistry. So can you explain what cheese is and why I never want to eat it again after hearing you tell me what it is? Well, okay. I will tell you about a specific type of cheese that i am currently making i can say a couple but but yeah basically the the one that you saw me making in my instagram live the one and only time i've done that uh featured my parents even uh which

is great but i you basically get your your ingredients that you need are basically milk and heavy cream if your milk isn't fatty enough. But basically you have your dairy, your milk. So I got a bunch of milk. Then you have cultures you introduce. So this is bacterial culture and then also molds. And your coagulant which is rennet and then depending which milk you use you might add citric acid as well just to play with the ph i think but what you do is you warm up the milk you add the cultures you let them kind of rehydrate because you buy them powdered

form from the the web of cheese and then you add the rennet, which will coagulate it. And then you kind of let it sit. And you have now kind of a jello, a big jello thing of milk. And that's what you call your curds. And so you cut it. You cut it all up, chop it all up into smaller pieces. And you stir it up. They squeak?

Not in this case the you you're probably thinking of the cheddar ones and those i believe they they wash them with hot water after this step they do something else to it depends on the cheese you do different things once the curds are set um but for brie it's a soft cheese after the curds are set you Um, but for Brie, it's a soft cheese. After the curds are set, you cut it up and then you put them in these molds. And they're kind of like, imagine a cup, but it's sort of a strainer. So it lets the liquid out in the form of the way, uh, W H E Y, the way, and it kind of, the way kind of falls out, and then you flip the cheese a couple times.

And then you take it to someone, and you say, this is the way. Sorry. This is the way of the cheese. I don't know why that reminded me of Star Wars, but. A Mandalorian. Oh, right, yes. That is the way. I just watched it last weekend for the first time.

But so you have the cheese curds, you put them in these molds. The way kind of falls out and and drains out. That's a better, more technical term. You turn a couple of times and this over the course of hours. So you might do 12 hours, leave it overnight, kind of a room temperature while the basically the cheese forms can hold themselves up without the molds. And you take them out of the molds, you put them in your cheese cave or aging container, which for me is a plastic Tupperware box or Tupperware container inside of a small

fridge. And that is where the mold grows. So if you've ever had brie, it has this white rind. That's actually mold. And if you see it grow, and I can actually share, I have a little time lapse that I took of the mold growing. It's just kind of fuzzy, cotton-like mold that grows out. And the roots, the roots of that mold are what actually age the cheese. That's what's actually changing the cheese from just milk flavor to whatever different flavors you're going to get. So after about 10 days or so, you're kind of turning it every day just so the mold doesn't grow into the surface that it's sitting on. Just turn it every day and then you can wrap

it. And that's when the rind for the brie gets smooth. And that's why you don't eat a brie that's kind of like fuzzy, but it's smooth. So you wrap it, and then you age it for another like four to six weeks, depending on the temperature. And then you have cheese, or brie cheese, or camembert, or something like that. Okay, setting aside what rennet is made out of, and that there was a lot of mold in this story, which I thought wasn't what you wanted on cheese.

Well, it depends. And actually, most rennet in use nowadays is not animal-based. Oh, really? Yes. So there's a few different kinds. And yes, originally, rennet is made from, I don't know, the inside of the stomach of a baby cow or something. Something like that.

It's animal parts. So technically, like cheese isn't vegan or vegetarian. It's definitely not vegan. Yeah, it's not vegan. Well, there is cashew. I'll come back to that. I'm not going to go into that. But now there is microbial rennet.

And then there's also, so they actually use, I forget which critter. I don't know if it's a yeast or a bacteria or something or a virus to actually make the same proteins that are in regular rennet without using the animal. So I guess they got the original, like the, the, um, the specifications, the user manual, they basically made a copy of that.

I think it's protein, but whatever chemical composition, they're able to reproduce it now without using the animals. And most, I think it's over like 60% or 90%, something crazy in the U US of commercial cheeses actually don't use animal rennet anymore. If you go to like France and you go with the old school diehards, they're going to tell

you that it makes a big difference and you got to use the old school stuff, but it doesn't seem to be the case for most places. Okay, I stand corrected, but I still don't know where the electronics comes into this picture or into this cheese as the case may be what electronics i don't know what you're talking about uh yeah so this i i have a problem yeah i it's an electronic cheese problem or a personal problem i do have a personal problem. side projects in ways that teach me something I don't know yet. So I don't know, this was 2017, maybe I started thinking about making cheese. And I went, I went into it with the perspective of being an engineer. And I started designing like a circuit board with temperature and humidity

sensors and like relay output so i could turn the fridge on and off to control the temperature and then i had another relay to control a little motor motor for a water pump so i could introduce humidity i had no idea what i was actually going to need but i kind of just started making circuit boards because why not and then i actually started making cheese and found out engineer Engineer it first. Figure out what you need later. Yeah, yeah, yeah. I was like, I'm just going to do all the things. And I needed to do this.

I wanted to play with the STM32, I don't know, F0, or some microcontroller, right? And so I wrote some code. I got to work with USB and that thing. And then later I realized that I don't need to control this so much. I just need to measure it. So I started making my circuit board smaller. And just because I don't need a fan controller,

I don't need a relay controller for the fridge. So I made it smaller. And then somebody, I don't know if it was Wendell from Evil Mad Scientist, or someone recommended it. It's like, hey what if your your boards were shaped like like cheese and i was like what so so then that's a great idea so then these sensor boards themselves that had a temperature humidity sensor i started making them like little wedge shape like pizza slice or cheese shaped but these were osh park boards so they were purple so then i found out how can i make these yellow?

And then I made yellow cheese-shaped cheese temperature and humidity monitors. And then I kept going, because why not? Because there's a lot of wires. I thought, why not make this wireless? Why not make this Bluetooth? So I made these kind of standalone little cheese wedges. And they have an NRF52, a little 2032 coin cell battery, and the sensor and Bluetooth.

So now I've learned how to do weird shape circuit boards. I've learned about STM32s. I've learned about NRF52s. I decided to teach myself my newt, the operating system, the RTOS for playing with Bluetooth. And then I wanted to learn low power design, right? Like how low can I go? And then I was able to get it to like one to two microamps average current. And I'm transmitting every minute the temperature and humidity, which I thought was pretty good.

But it took a lot of work to get there. And then, yeah, so these are just excuses to learn things and, you know, monitor the cheese. So when you have Bluetooth on a wedge-shaped board, does that make it blue cheese? Oh, no. No, it's yellow. I'm not going to acknowledge that. Did you put extra large vias or extra large drilled holes through the board so that it could look like swiss cheese i thought about it but i didn't i i i i think the most i

would do is silk like weird silk screen the main problem is the battery the cr 22 2032 is i guess i could probably get away with some holes but the bottom is mostly just battery so you can't really see much and the top's components so there's not a lot of room to cut out of any shape that would look good I guess do you have all this online can somebody else make your cheese board that's not github it's not like I don't have like a page describing it all. But yeah, it's all my stuff is open hardware, open source. And I have pictures. I've also done a crowd supply teardown. I gave a talk about it a couple of years.

Well, OK, no, maybe more than a couple of years ago. But all of the don't do this lessons learned. So all the mistakes I made, I made a talk about that and how i ended up where where i was at the time so you're sending temperature and humidity data every minute yeah while your cheese is being what's the word here while the cheese is slowly digested by the forming uh while your cheese is cheesing. Mushrooms? Mushrooms. Are you sitting here watching a big display and seeing if there's, you know, an error or something? How do you use this?

Yeah, I have a projector on the wall with the live, no, no. Alerts coming into your phone, emergency cheese humidity. I get text messages every three minutes. No, so I don't really look at it very much, to be honest. So the boards transmit temperature humidity, and right now it goes to a Raspberry Pi that just saves them to an SQLite database. And then I do have a web interface I could use,

but I don't really look at it unless something's wrong. So I'll check it every few days just to make sure the temperatures are what they're supposed to be and the humidities aren't crazy. But the main goal right now for me, since I'm able to make it without the sensors checking them too much, is if something goes horribly wrong, I have kind of the history the history the data to go back and see like what happened or if something goes very right i'll have this an example of you know what environment it was in so it's mostly kind of a backup of if something goes wrong right now and just the current the current measurements when you started out making cheese When you started out making cheese, you started out making

possibly over-engineered sensors. Yes. And then you got more into the actual product, and now you're barely monitoring them unless something happens. Yeah. Do you draw any conclusions from that i mean the monitoring was always the excuse right like people have a journey cheese but was it about the cheese or was it about the electronics

that's what i'm trying to get to the heart of yes it was about both uh really the they kind of drove each other, right? Like sometimes the cheese, there was like a year and a half where the cheese was failing repeatedly. Like we couldn't get it right. And so I worked more on the sensors because that was actually working.

And then eventually the cheese start working again. So this year I've been making lots and lots of cheese, but I haven't done anything with the sensor. I did change the, for MSQ Lite i i'm playing with influx db which is a different type of database just because why not because it's fun uh but but yeah it kind of goes back and forth and and and like i said it's just an excuse to learn new things do you think your adventures with your cheese board helped you get your current job? But with the whole monitoring and low power part? Yeah. So the, uh, so far has a pretty neat

interview process where you do your regular kind of interviews and I guess this, we're all remote, but, uh, one of the, the last steps of the interview process is to do a presentation to the entire company. Uh, it's a very small process is to do a presentation to the entire company. It's a very small company, but do a presentation to anyone who wants to show up about anything you're passionate about. They don't tell you what to give a talk on. And so I did like a five minute, how do you make cheese? And then I went on after about how i made all my cheese boards because why not and then and then i talked about my weather station after and these were all very relevant

projects to to a remote weather measuring company so i don't know if they that's the reason i got it but i i think they helped it's, did the skill building that you were attempting succeed in getting you a new job? Oh, absolutely. Yes, like 100%. And this has been kind of historical. Like, this has helped me get jobs in the past as well. Like, when I first showed up for my interview at Apple, it was for firmware engineering, but I brought some

circuit boards I made for fun just to play with. And they weren't super relevant to the writing code part of it, but they were like, okay, that's cool. And then that also means that you can know your way around hardware. And as you both know, if you're a firmware engineer that works with hardware, if you're comfortable with hardware, that's with hardware, if you can, if you're comfortable with hardware, that's a lot more debugging you can do, right? If something's wrong, you don't just go hand the board to the hardware engineer and say it's not working. You can actually dive in and, and like pull out the oscilloscope logic analyzer and do some more debugging without needing anyone else's help. And showing up with my hobby projects that were hardware related for a former

engineering interview was impressive to them because it's like, oh, okay, you know, you know, more than more than that. And same, you know, with Verily, you know, I showed up and I had these other side projects that were related to the job, but not specifically what I was wanting to go do. So I think it absolutely has helped. Long story short. Do you consider it a portfolio or are you just taking random projects to interviews? Do you do the formal thing? No, I just do it for myself, basically for my own. One of my high school teachers was my personal edification. And yeah, just I do it because I think it's fun and I want to learn these things. For example, with some of the cheese projects and some of the kind of in-between projects, I wanted to learn surface mount technology, right?

And I had done surface mount before, but with the soldering iron, but I wanted to do reflow. So for my personal projects, I started getting stencils made, and I got some solder paste, and I got a little reflow oven. And my second job at Verily was doing a lot of hardware prototyping, and I was doing all that myself. And I got it because they knew that I could do it, and they knew because i had my little side projects that i'd been doing just for myself but i didn't go and say oh i should learn how to do surface mount solder paste whatever stuff and because it was going to help me get a job um i could have done that but that was not my primary driver i guess it's mostly my own curiosity. Do you have any tips for people who don't have the energy, motivation, ambition to be able to do both work and projects at home? I mean, that's a tough one, right? Because I am that crazy person that likes this so much that does it for fun as well as for work. But if I'm going to be perfectly honest, I don't always do the exact same thing at work and at home. Like whenever my job is

more firmware related, I might do my hobby projects, more hardware or something else, right? If I'm doing electronics at work, I might go home and do more software and firmware. This year, I've been doing a lot more mechanical. I got myself 3d printer. I've been learning like free CAD, just doing kind of 3d modeling stuff. So, so that's kind of a way to, so your brain doesn't get totally tired, but as far as making time, anything like that, I just, you know, the only thing, and I'm sure y'all have told people before is you just need a project, right? You need something that you're excited about, and then you learn the tools you need to do it. So if, and in this case for me, it was like making cheese, right?

Like I thought that was cool. And I was like, I can make a tool to make my cheese making easier or something. And then that was my excuse. And that's what kind of got me motivated to do it. But how to get that motivation is kind of a hard, hard question to answer. Yeah, I think if we could bottle that, we could really make a lot of money. Yeah, well, then we don't have to work anymore, right?

But also, I don't want to because of the bottling. Well, because, yeah, that's true. That's so true. But, yeah, no, and this year or the past year and a half, I would say, actually, has been not very productive for me. Just even though I do have all these side projects and it looks like I do a lot of stuff, I haven't. I've been trying to stay off the computer because I have a lot of repetitive strain injuries and stuff. So I've been trying to find other hobbies, which is really hard to do without using your hands. So to say, even if you do this for like a few hours a week,

it's still like something, I guess. There is some value to realizing that you didn't do this all in the last five months. No, no, this has been very long long term yeah um and yeah that's why i i mentioned earlier that one of my talks was all the things i did wrong because you see yeah a lot of video a lot of youtube videos blog posts you see those amazing projects that are super intricate and then it looks like they just did them really fast and some people do work really fast but but they don't you don't get the perspective of you know all the hours that were put in all the suffering the pain and suffering of things not working and yeah like for example the cheese thing i think i started in 2017

so it's been like four some years on and off just doing these things a little bit at a time. But it's never been I have to put in 20 hour weeks in addition to my job for the next four months. No, absolutely not. If you can look at I watch a lot of movies. So if you just add up the amount of time I spent watching movies, you'll that i don't spend most of my time outside of work uh working on projects because i do keep track i can you can do things while watching movies you know that's makes the movie i'm kidding i'm kidding i don't multitask at all like if i'm watching a movie i'm watching a movie i'm not but yeah i do keep track of all the movies i watch, so I know exactly how many hours I watch.

So I can tell you. Okay, so one of your other major projects, as you mentioned, is weather station things. Yeah. What do you do? I mean, I've read some of your blog posts, but... Andy brought it to the Embedded Party which which was that's right which was only a year and a half ago oh no disturbingly enough the second one right yeah yeah yeah i feel like it was eight years ago but no i feel like it was two weeks ago like the vast year has been a blur year and a half feels like a century um i think i brought my cheese boards to that one too but but yeah so how did i end up

doing the weather station uh that that i i found a different excuse to work on a project and that excuse was my dad because he's a nerd just like me and his father before him and his father before him we were all nerds and my dad's family has a thing of, they love measuring things, measuring everything. Like my, my grandpa had his own little rain gauge in the back of his house and he would measure every day and write down how much did it rain. He would write down temperatures. We would log the, the car mileage. Every time you get service, the car service, or you fuel up or anything, they would write down a little logbook and that kind of thing. So I thought it'd be neat for my dad's birthday to build him a little weather station. And I could just buy him a weather station, but then I don't learn anything. So I

was like, why don't I design a weather station and give it to him for his birthday? So that's kind of how it started, like, I don't know, three years ago, I think. And again, it was an excuse to learn new technology. So let me learn about radio stuff. So I started playing with ZigBees. Well, actually, so I should go back and this will be relevant to what we were talking about earlier. I started this project and I went and I got a little radio module and I built kind of a sort of working weather station, but didn't start working. And then I just got tired of it and gave up and didn't touch the project for a year. And then I realized that I, I, the project suffered from premature optimization as a lot

of people do, where I try to get it ultra low power, ultra small, super efficient, everything all at once. And then it wasn't quite working. And then I gave up. And so I realized like that's stupid. I need to do this better. And I kind of gave myself a break and said, look, just build it as big as it has to be with as many wires and development boards. Just get it working. Make something work.

And then once you have something working, start optimizing it, start making smaller, but always having something that works. And then you still have that kind of a shot of, of, uh, joy that you have something that works and, and, and then you can keep iterating and while still having something that's working. So then I started, you know, making smaller, I, I was using, I think, yeah, XB radios and an STM 32 something and a big solar panel and a big battery. And I learned about weatherproofing because that was my excuse to like learn how do I make electronics to survive outside? How do I make solar? I've never done solar charging stuff.

I've never done weather measuring besides temperatures and humidity inside fridges. And yeah, so then I got it to my dad in Mexico and we put it on his roof in his house and then it would send the data to a Raspberry Pi and then the Raspberry Pi then has a web interface and then you can see that. And then over the following year, I started making it smaller.

I switched from from was it Zigbee to Bluetooth because Bluetooth was long enough range. But then I switched now to LoRa and then I switched from NRF 52 to STM 32. And I've just been kind of making a smaller lower power different technologies and and it's just an excuse again to learn all these different things laura versus ble versus zigbee i mean those all have really interesting use cases but they're not always the same use cases they kind of are a spectrum right yeah what did you think of them what would you use them for in the future so the nice thing with the the xv modules is that they just work um they're not cheap i mean relatively you know for one offices whatever but um they're you do have kind of to set them up with their little um gui program, unless you want to read the data sheet.

But you basically just you can set them to do a mesh network. You can set them to do point to point or one to many. And the radio just works. You can kind of do a serial bridge, if you will. So you just send data through your yard and then appears somewhere else. And it's easy to use. So that was really cool. cool unfortunately i think most of them now are 2.4 gigahertz which is fine it's just the range is not as much they used to have a 900

megahertz one that would reach a little bit further then you have bluetooth and i think you're the most familiar with with bluetooth and it's it's a nightmare um because they're for for depending what you want to do, if you want to talk to a phone, then yay. But if you're just trying to communicate between two devices, I feel like it's a little too much. It's a lot of overhead. But the reason I wanted to use it is that the NRF52 is pretty cheap, and then the NRF52-811 and the NRF52-840 supported the new protocol.

What was the... The BLE mesh one? Not BLE mesh, but the long-range one. It's a different kind of coding they do. Yeah, I don't remember what it's called. But it's basically Bluetooth, but long range. So they actually drop the bit rate significantly and can get much further range.

And it's still 2.4 gigahertz, but they do more error correction, I think. And then I was like, oh, that's really cool. I can use the same chips I've been doing and get much longer range. But I never actually got there. Because it was working well enough without trying to figure out. And also,

because Bluetooth is complicated and I didn't want to learn all of the I've already forgot all the names, but you have to make your connection and then you have the things. I just do broadcast packets. So I just send advertisements with the data and that's very easy to scan for advertisement packets and you don't have to actually make connections that's that's not i mean you're you're making fun of yourself but that's that's that's i mean

nothing wrong that's how the g sensors work too they just send advertisements with like hey here's my like serial number and temperature and humidity and battery level like do what you want and then i don't have to write in any fancy phone apps or anything i just kind of scan for scan for broadcast and and read it in so so that's bluetooth and then laura is much much longer range and it's a very different protocol again it's not quite as cheap because Semtech, the company that basically has, I don't know if it's a patent, so the rights to the actual encoding, any other company that wants to make lower radios

has to license it from them. So there's a little bit of a tax on that. But the protocol is designed to be very long range, very low bandwidth, but long range, which is kind of what the weather station is great for. And then you can do LoRa just point to point, kind of what I was doing with Bluetooth and with ZigBee. But then there's LoRaWAN,

which is kind of the protocol on top of of that which kind of you can send it straight to the internet if there's there's gateways all over the place and then you don't need to have kind of your own uh gateway because there's lots of public ones do you have your own gateway or or do you use a a wan point uh right now i'm doing myself just point to point because it's easier again um i haven't i do have a laura it's called the things network and it's kind of an open laura network and then if you have your own gateway connects to their network then you can any of your devices can also use their network for free around the world um now amazon snuck uh laura gateways into everyone's houses with the echoes excuse me what yeah uh most amazon echoes have a laura gateway in them that they're turning on

yeah surprise um why why would they do that? For all of their Internet of Things widgets for all around the house. I mean, Zigbee doesn't work that well around a house. Bluetooth. Except we have 700 Zigbee notes in our house. Well, yes. Yes, but they're not all on the same Hive and channel. Or a Pan-ID and channel. So they don't necessarily talk to each other.

And LoRa is a little nicer in that you can talk to strange devices and still expect things to get to where they're supposed to go. Yeah, if you're using LoRaWAN, then it'll route it to where it needs to get. But yeah, so I think it's called Amazon Sidewalk. And I think what they're doing is they snuck these gateways into everyone's houses and then they're going to sell the service. So if you have any device or any connected device that you need to connect to a LoRa network, then you can use Amazon's and they're probably going to have global coverage already. Well, and since they're probably providing the backend where you can monitor, it's part of their AWS services.

Yeah, okay. Makes sense. AWS IoT or something. Yeah, AWS IoT. But yeah, if you're developing Gizmo, that's very convenient because then you don't have to worry about the other end. You can just make your device, make sure it speaks LoRaWAN, and then register it with the Amazon network,

and it'll work almost anywhere, which is convenient. You're not using that yet? No, not yet. Do you think you will? I'm thinking about it. It's just more code. I just have to spend the time to kind of write the code Do you think you will? I'm thinking about it. It's just more code. I just have to spend the time to kind of write the code or port the open source code to my device,

and it's a little bit more testing. But that introduces more uncertainty, just because, yeah, it's now bidirectional communications. I'm not just transmitting out into the void and then hoping it gets there. I actually have to leave my device on for a bit, wait for response, or turn on every so often, and then the power consumption goes up a lot.

I don't want to say a lot more, but I have to do more profiling just to see how much it's going to last. And again, it's not a huge deal. I can always put a bigger battery, but I'm very obsessive with, like, oh, I need to be the most efficient okay so cheese and weather stations do you have any other micro obsessions or plans i thought you said microbe obsessions well we did talk about cheese

so that's fair um do you have anything that you're starting or thinking about doing? Well, I did do over the break. Break? It's not a break. I don't think that's what people consider. While I was on vacation. So while I was unemployed last year, I found my, for some reason, I had my dad's, one of my dad's kind of lab

reports from college for one of his digital systems classes and he had made this was like in the late 70s he made a digital clock out of logic 8 so it's like a 555 timer as a clock source and some uh counters and and then uh hex encoders or decoders, I guess. And so I found my dad's like notes. And then I thought, hey, I have some of these logic gates. I should see if I can put this together. So I breadboarded the entire clock. And then I kind of tweeted about it, which was fun, just as people kind of reminiscing. And then thought why don't i make a soldering kit out of this so i made a circuit board with my dad's original design and then i sent him a soldering kit to put together which was which was fun did he yeah did it work yeah he's got a working clock

that he designed uh-some years ago. That's a great idea. That's just a great idea. Yeah, but new, new projects, I'm trying not to. So you always ask the start a dozen or finish one kind of thing. And I'm trying to finish these two before I take on any more ambitious things. I've been doing 3D printing stuff, just learning. And mostly it's an excuse to make enclosures and kind of things to hang other things in, like for storage. I do have my obsessive storage system where I have all my components cataloged and labeled and database. But that's another condition that I have.

But yeah, nothing new. I really do want to finish these before I move on to something else. I have lots of ideas, but I don't... I'm going to talk to you at some point about FreeCAD and stuff, because I haven't been doing much 3D printing, but I'll probably get back and do it a little bit.

It's fun once the printer is working i always have trouble with the cards yeah they just released free cad uh 19 finally or well 0.19 uh it was in beta forever but it's it's pretty good and i can i can recommend some tutorials as well some pretty good youtube videos cool i did um so sophie wong back i don't know how many years ago maybe two years ago did a 30 days of fusion 360 where she was modeling uh uh doing a new 3d model in fusion 360 every day of the month oh right and i saw you doing something like that so i did that for her well then gregvel, who's the wizard of the electronics, was doing one circuit board every day, which is mental.

And then that was also very inspiring. So I think in January or February, I decided to do a FreeCat model every day just for practice because that's the best way to learn is to do something. And then I put the ridiculous and arbitrary condition of i'm gonna make a 3d model every day and so yeah i did do that how long did you spend on them it was a question i meant to ask you while you were doing it but i didn't so like is this three hours of homework or like a half an hour or it depended i i in some cases it was less than half an hour

when it's like okay it's 11 o'clock and I just want to get something out quick. And yeah, it'd be like half an hour, 25 minutes. But others, I did spend like a few hours because it was something useful that I wanted to do. But yeah, it was mostly an excuse, again, to practice. Yeah, and that's a very arbitrary excuse, but you know, it's, yeah. And that's a very arbitrary excuse, but, you know, it's something. I really like that idea of doing something every day, even if it's small,

to keep you kind of engaged with the skills that you're supposed to be learning. And then you could just choose, yeah, to do different, like, even if it's not something useful, it's like, maybe I'll use this different tool inside of, or different feature of the tool. And at the beginning, because I'm terrible at ideas, I did find this one YouTuber who posted kind of 3D modeling challenges, if you will, like puzzles, where he'll post the mechanical drawing, like the 2D drawing with dimensions, and then you just have to draw it or make it in CAD. And that was actually pretty helpful because it was some, you know,

some pieces that, you know, aren't super interesting if you just look at it. But going from the drawing to the 3D model is kind of a different way of thinking. And also he designs them to kind of force you to use different parts of the tool, which was cool. I like the 100 days of art, 100 days of code to get you into the habit. 100? This is just one month. Those were more general habit-forming things.

Oh, yeah. Where 30 days of free CAD is more learning a tool all the way down and not learning the first 10 minutes 30 times as you do it over a year. Yeah. Well, and if you're trying to solve problems, you're... Exactly, yeah. There's a necessity to find, oh, I cannot do this with my first 10 minutes of knowledge. What's the other tool I need, right?

And if you have a 3D printer, you can actually print stuff and kind of hold that thing you made, right? Yeah. But yeah, there was a couple that i took several days to do so so i did a 3d model of my headphones so one day i just did kind of the the arc the cups and and then the other day i did the the headband and and that still counted because i still modeled every day well it's been wonderful to talk to you and thank you very much for um standing in for a guest who couldn't make it i really appreciate the the hey i need help and you saying okay do you have any thoughts you'd like to leave us with um yay have some cheese cheese will make you happy unless you're lactose intolerant, don't have cheese. Please.

I mean, there's still some lactose intolerant people who still like cheese. with projects, uh, uh, the biggest lesson I learned from the weather station one was just get something working. Don't get fancy at first, um, because you might get discouraged early. So just get something working, even if it's not elegant, even if it's kind of unwieldy or hacked together,

once it's working, you'll be happy. You got something working and then you can build up from there. I'm so glad we didn't have to end the show on flatulence jokes. Nobody said anything about flatulence. It was implied. Well, now we have. Our guest has been Alvaro Prieto, electrical engineer, firmware engineer,

maker, traveler, and co-host of the unnamed reverse engineering show. And cheesemonger. Oh, show and cheese monger oh and cheese monger sure thanks avaro oh no problem thank you to christer for producing and co-hosting thank you for listening and oh i do have a quote here i. Oh, if you want to contact us, show at Embedded.fm or hit the contact link on Embedded.fm. My quote is from Patrick Rothfuss from his book, The Wise Man's Fear. I was one of those. I meddled with dark powers. I summoned demons. I ate the entire little cheese, including the rind.

Yes.