Embedded - 298: In the Cow Case

Episode Date: August 9, 2019

Eric Brunning (@deeplycloudy) returns to talk about doing science in the field in this crossover episode with the Don’t Panic GeoCast’s John Leeman (@geo_leeman).  Eric is a Professor of Atmospheric Sciences at Texas Tech University specializing in storm electrification and lightning. We spoke with Eric on 268: Cakepan Interferometry about lightning and using baking goods as measurement devices. Eric was also on GeoCast 134: Launching Balloons out of a UHaul. We spoke with John about his Phd research in 169: Sit on Top of a Volcano. The previous Don’t Panic GeoCast crossover was with John and Sridhar Anandakrishnan in 206: Crushing Amounts of Snow. John’s company is Leeman Geophysical. The paper was Reconstructing David Huffman’s Legacy in Curved-Crease Folding by Erik D. Demaine, Martin L. Demaine and Duks Koschitz. Elecia is working her way through Erik Demaine’s Phd thesis on the same topic as well as Jun Mitani’s excellent book Curved-Folding Origami Design. Geology also has folds. For 3D printed origami, Eric mentioned Henny Seggerman’s twitter @henryseg.

Welcome to Embedded. I am Elysia White, and I'm here with Christopher White. And welcome to the Don't Panic Geocast. I'm John Lehman, here co-hosting another crossover show, and this week we've got special guest Dr. Eric Bruning. Hi, everybody. Hi, everybody. Hi, everyone. Eric, since you're the guest, let's start with you. Could you tell us about yourself? Yeah, I'm a professor at Texas Tech University, associate professor, and I study atmospheric electricity and lightning and how that relates to meteorology and the rest of the way the thunderstorm works.

And you've been on our show before. We asked you many lightning-based questions for Lightning Round. You've also been on John's show, right? I have, yeah. Been on both and glad to be visiting with you again. And let's see, John, since you aren't always on our show, could you introduce yourself? Sure. So I'm John Lehman. I co-host the Don't Panic Geocast with Shannon Doolin. And I also run a small custom scientific and

industrial instrumentation and automation company creatively called Lehman Geophysical. So I couldn't decide what I wanted to do with life. So I stayed in school for as long as possible and ended up with a PhD in geoscience from Penn State and spent a lot of time working in an experimental lab, which gave me lots of fun opportunities to play with things, do all kinds of hardware and software hacking, if you will, and left academia, worked in software for a couple of years. And then my company got large enough to support me full-time, so I've been doing that for the last few months. That's great to hear that it's going well.

Thanks. Christopher, will you introduce yourself? Yeah. As though I didn't know you? As though you didn't know me. Well, I'm a pretty cool guy. No, I am an embedded software engineer, and I've done a whole bunch of different

kinds of products, a lot of medical devices, a lot of medical imaging devices. So I've played with optical physics and things like that, and laser tissue interaction, which sounds about like you think it is. And I formally worked at Fitbit more recently, and now I'm back to consulting full-time. And yes, that's a change. You also have an advanced degree. I have a degree in physics, yes, a master's degree. I'm the least educated one here, but I wrote a book called Making Embedded Systems, and I am an embedded systems engineer, a consultant. And let's see, a podcast, a book, an engineer.

Two dogs and a pile of folded papers. Right, which we'll be talking about later. Before we get to that, the real idea behind this show came up when we had Eric on our show, and it seemed like a lot of his gear, like his field gear, was made out of cookery. And I know that John and Eric have worked together quite a bit. And so I have a question for you both. Is this how physical science is done?

Yes. Yes. That was short. All right. Show's's over and we're done uh will you tell me about things times you did crazy things and got them to work or didn't get them to work sure uh i i guess i'll uh i'll start um there's's an instrument that I flew as an undergraduate and graduate student on very large weather balloons called an electric field meter. And it does what it says, measures the electrostatic field inside thunderstorms. And my main memory of doing work on that instrument was the amount of tape that we used to hold the thing together. as well as electric tape, like regular high-quality vinyl electrical tape,

to hold together the two aluminum spheres that did the actual sensing of the electric field. And so, you know, you'd get an instrument back, tear it all off, and, you know, fix any damage, grab the data, and tape it back up and fly it again. That sounds like a perfectly reasonable use of tape. Yeah, I don't think folks realize how much of science is accomplished with various forms of tape. All of the samples that we built in our lab to do earthquake studies started out with lots and lots of scotch tape. We bought it by the gross and used that to contain our sample material

and lots of razorblading scotch tape to very precise heights with titanium jigs and fun stuff. But at the end of the day, it was really just scotch tape. I think the tape in the medical device lab was always Kapton tape. I'm surprised you guys aren't just using that all over. the uh the exotic one that we used in the uh uh in the electric field meter was a uh was a certain kind of conductive tape that uh that 3m makes probably other manufacturers too um and we were having uh so there's a radio transmitter in the instrument and it uh it sends out uh 400 megahertz rf and as you might imagine, dangling an antenna out of the instrument in 100 kilovolts per meter is maybe not the most forgiving environment. And we were trying to deal with some interference problems. this conducting tape to create a proper Faraday cage across the individual seams in that case that we built. I have this idea of crossing chibitronics, you know, the little books you can buy and make like collages that are made out of this conductive tape and weather balloons.

Is that about right? I think that could work. Is it because you end up just having to improvise in the field? You have good plans, but as soon as you have first contact with a real weather. Or is it about the budget? Yeah, yeah. Which is it? Is it just improvising because you didn't anticipate something or just it's, hey, this

is cheap and it works? A lot of it comes from, you know, every few years when you go out and do another field experiment, you've tried to make some improvement to the instrument, which means you then, as you said, are up there in the storm for the first time with that part, and you have to iterate on it a little bit. And so you need something that you can do quickly. You can't rebuild the whole instrument. So that's part of it. Other parts of it is budget as well. And when you're doing things on dozens at most scale, it's not worth investing in some sort of hugely custom part. Yeah, no need to do vacuum forming and die manufacturing with 10 units. Yeah, yeah.

Well, and I think a lot of these, you don't always know if you're going to get a balloon-based instrument back. So you don't want to have something that's too expensive that you're going to put up in the first place. But a lot of these sort of low-quantity instruments, there are all kinds of little tricks to get them to work that resort in interesting field bodges. Because you are always modifying the instrument or you're in a different environment than you anticipated. Nobody probably anticipated trying to deploy this thing from the back of a U-Haul truck with a big balloon like Eric was doing.

But there's a lot of laying on of hands, sometimes literally, to get these things to work. Yeah, that's right. The laying on of hands was something we ran into with, let's see, it was the, which part of the there was some sort of capacitive thing that literally required for the instrument to start for certain instruments, not all of them for you to touch the instrument, and then it would magically start working. I've met so many of those devices. Yeah, usually it's me not touching them because I have this field around me that breaks electronics.

But once I take them to Christopher, they work again. Is that true? Usually. I just swap them with a new copy. Yeah, probably. In grad school, we call that the advisor exclusion zone. Speaking of grad school, you got ad uh is it two years ago now yeah so i've got coming up on three years ago pretty soon that i finished time passes and uh

you have a story about uh bodging things as part of your PhD thesis? Oh, there were lots of bodges as part of my PhD thesis. The instrument that I worked on, especially, it was called the BIAX, and listeners of our show will be familiar with that, but it was just a two-axis large hydraulic press. And so each cylinder produced somewhere in the neighborhood of a quarter million pounds force. Say that again. Just a little. Yeah. And so we would put rock samples in there, powder samples in there, and shear them to study their frictional properties and learn about how earthquakes work from that. And that whole machine was built on continuously because in research, you're always trying to

do something new. So we would always need a new capability. So somebody would modify the control electronics. It was all servo hydraulic controls. But the first thing that came to mind when we talked about this is the hashtag overly honest methods on Twitter. It's a great one to go follow and look up sometime. Because in all of my papers, you'll see me talk about impermeable plastic membranes that we use to humidify the samples. Those are Ziploc baggies. And we did this because we started engineering this humidity control chamber that was going to go inside. And we had all kinds of things that we were trying to do to seal where the pistons go through to exert force on our samples

and a whole control methodology. And then we realized that just by using a simple chemical reaction in a Ziploc baggie, we could do the entire thing much cheaper and just as well. I once had a company come to me and they were having a bunch of trouble because they needed to have many of their devices in their office working with all their engineers. This is a radio problem. So they had like 82 BLE devices all crowded together and they couldn't make anything work. And they wanted help. They wanted software to fix this. And I said, well, I'm pretty sure I can fix it. And if I come up there, it will cost some exorbitant amount of money because they were in San Francisco and I don't like to travel. Or I can just tell you, but you won't believe me.

And they said, of course, they said, well, just tell us. And so I said, okay, you take the Amazon box you got today, the little one, you put your phone inside, you put your device inside and all of the debugging stuff, and then you wrap it in aluminum foil. Yes, you do. And it took them about two weeks before they seriously tried it. And they just kept fighting it and fighting it. And I said, well, have you tried, you know, the Faraday cage? And well, we don't think that's the problem. Have you tried, you know, the Faraday cage? And well, we don't think that's the problem. Have you tried it?

Oh, yes. There are several different places that I've gone where we were doing everything from high energy physics to just piezoelectric time of flight measurements and rocks. There's ample use of aluminum foil everywhere. Okay, so we have tape, electrical tape, and packing tape, and scotch tape, and aluminum foil as our things we must take to Mars. And Ziplocs. And Amazon boxes. What else is in your field kit john uh you know i i make use of a lot of five gallon buckets for housing things out of the weather since we're doing sensing out in the environment

these are one of the more inexpensive ways to keep your sensors and your electronics out of the weather. The first time I did this, we were actually deploying, it was during my PhD, but not related, deploying some infrasound sensors out in the auspicious sounding experimental forest, which was in central Pennsylvania. Infrasound sensors are long wavelength sensors. And like some animals use infrasound to communicate, like elephants and crocodiles. What were you looking for? I don't recall there being either of those in Pennsylvania. No, so I was actually looking at microbarons. So these are things like as the waves break on the coast, they create very small

pressure fluctuations. So that's sort of 0.1 hertz or 10 second period signal. And we had some storms that were at that time coming up towards the eastern seaboard and going. And so we were using a little infrasound network and doing some fun things with noise cross-cor cross correlation on them. But we put these things in buckets, put them out and the, the experimental forest was open to hikers. So we had to try to camouflage them a little bit. So nobody got too curious. So we ended up putting the, we'll just put this branch over here.

Not the orange Home Depot. Well, they were the orange Home Depot buckets. That was the problem. And it was during deer season, so we had to have orange on when we were out there. You never know who's around hunting. But we put these out and then dressed the orange buckets in camouflage t-shirts we got from Walmart and then bought a little child's tent and put all of the recording electronics in that and draped it with camouflage as well.

So we had lots of Home Depot buckets and camouflage t-shirts in the forest. That's pretty awesome. What about you, Eric? What do you have in your toolbox? Well, one of the things we like to use a lot is tie wraps and hose clamps and all sorts of things like that. I run a lightning mapping network out here in West Texas. And when we were installing that, of course, you're building relationships with local farmers and things because they have these nice big open fields with no noise. And so, you know, a farmer follows us along to do the install one day. And, of course, we've got an antenna up on a big pole and coax cables running down from the pole into our electronics enclosure. And, of course, you've got to get all that stuff held down in the West Texas

wind somehow. So, you know, lots of tie wraps and hose clamps to get it all to stay right tight to the pole. And I was proud we got complimented by the farmer actually said, you know, you do great farming out here in West Texas because you got to make use of the materials you have to to get the job done. What about you, Christopher? You don't go in the field very much. Yeah, I don't have a field kit. Sorry. I'm a software developer. I have a toolbox. My field is my chair. I have a toolbox. I just don't use it. So Alicia, what are some of the things that you carry with you everywhere when you're going to work? I have a multimeter because, of course, the first thing you need to know about any piece of electronics that's broken is, does it have power? And then there's the make sure it doesn't have power and then power it again and the classic have you turned it off and back on again

uh tape of course now that the logic analyzers are small i usually have a logic analyzer too um just because you can throw it in uh i have the good tweezers and I have the bad tweezers. The bad tweezers are regularly punished. What makes the bad tweezers bad? Because they are regularly punished by being used as little tiny pry bars. Gotcha. What else is really important? I mean, tape.

I totally agree with tape. Definitely electrical and probably duct tape. Although sometimes they put in packing tape instead of duct tape. Although that does remind me, we were talking at some point about attaching instruments to vehicles. And at one company, we were doing an inertial measurement unit and the quality engineer had gone flown out to see the customer and I was on call to make sure that this very unhappy customer would get whatever they needed for software and maybe I'd even go with them if it came to that. And the QA guy landed and he couldn't have been there for more than five minutes

before he called me and said, never mind, you don't have to come. And it turned out that what you really have to do to make an inertial measurement work is to tape it down so it doesn't tumble all over the vehicle. Ah, yes. Yeah. Picture is worth a thousand words in situations like that. You have a four-wheeler and drone stories too, don't you? Oh yeah. As a geophysicist, we were attaching all kinds of instruments to four-wheelers for land-based things like magnetometers. So we're using aluminum pipe and pipe clamps and plywood and mounting extra batteries on the front of this four-wheeler,

and then realizing we need to mount them on the other end so it doesn't tip over when we're going downhill because of weight distribution. Using a lot of electrical conduit because it's easy to bend with a bender that you always have with you when you're doing things like this to make nice little brackets and things. On the drone side, it's really just been atmospheric profilers and that's involving a lot of another kind of tape uh double stick tape how do you put a magnetometer onto something with an engineer with an engine and expect it to work

so you need to put it pretty far away so we're looking at a few meter long boom, which also makes going up or downhill interesting. How many times have you knocked that boom off? A few. And it's generally something that's rather worrisome because there's a several thousand to $15,000 sensor out at the end. So you want to design some sort of a weak point into the system. So maybe have a piece of plywood that will break before the sensor breaks and bends your boom all up and ruins your day. In field work, it's really important to remember that at the end of the day, when it's been 105 degrees outside all day, you sort of have, I've heard it referred to as field brain. Oh, yes. And you will mess up and you have to provide safe failure points for your system i like that you have these extremely uh unique and expensive sensors that you're attaching to vehicles and and generally integrating into systems with

tape and plywood and buckets and then doing donuts of course, once you have a four-wheel dune buggy, you're going to do a donut. Well, of course. And the meteorologists do this a lot, too. There were recently some field experiments where you've got anemometers and temperature sensors and pressure sensors attached to vans and trucks and even semis that have radars on them. Hey, I've seen the movie Twister. Okay, sorry. A little bit different than that. And Eric's probably a better one to speak here. But I will just say that measuring pressure from a moving vehicle is a really hard problem.

It definitely is. There's a colleague of mine at the National Severe Storms Laboratory that spent a lot of time working with different arrangements mechanically positioned relative to the airflow on the car. And let's see, also different sort of housings that surrounded the sensor. For pressure, there's sort of two parallel plates that tend to work well for the airflow through those. Other problematic things are things like temperature measurement. So you're going to get, you know, radiative heating coming up off of the car and all sorts of other effects. You want to make sure that you're really sampling the airflow. And so a lot of that was actually done with large PVC plumbing pipe, and the sensors were then housed in that, and there was a forced airflow through that to ensure that the measurement

was unbiased. And how far off the vehicle was this? I just can imagine this poor temperature sensor on a boom going up. It's pretty much right up above the roof, maybe hanging out a few feet ahead and maybe three or four feet up above the car. There's a fun figure in a paper in the literature, actually, from when these were first being developed, I want to say in the late 80s, early 90s, where they actually took a, like a Ford Escort or something and put it in a wind tunnel and looked at the airflow around the car and how that interacted with the sensor boom. I'm picturing a van with a horn like a narwhal. Yeah, it's, let's see, it sort of looks like a snorkel, actually, the way the PVC tube kind of sits up there.

Thank you. I needed a better picture. Snorkel's different. Yep, yep. And the anemometer is just high enough, you know, about a meter off the top of the vehicle, if it is a minivan, it's just a little too high to go through a drive-thru. Oh, man, that's like the worst. This was empirically discovered? Yes, of course. This was empirically discovered. I like that empirically discovered means, yeah, we damaged the roof of this van. What you want to do is you want to have the VHF radio antenna that you use to do vehicle-to-vehicle communications taller than all of your instruments.

So you hear that thing whack first and you know that you're getting close. Yeah, yeah. And how are parking structures? You avoid them, yeah. And how are parking structures? You avoid them, mostly. So what has been the most expensive instrument you've potentially damaged that you're willing to admit? There's a lot of qualifications. Oh, that's tough because

if you admit it you're going to be sanctioned or there's too many to choose from yes so there's a gravimeter so we measure tiny changes in gravity you can tell differences in density of material below you because we're measuring gravity out to six, ten decimal places.

These are roughly $150,000 for a used instrument. And they're about the size of a little toaster oven, but they stand up on end instead of lay down like the toaster oven does. And when they are unlocked, they have a zero-length quartz spring inside, so very delicate. And so when the mass is unlocked and you're actually taking a measurement or you haven't relocked it after taking a measurement, if it tips over, it's a several tens of thousands of dollars mistake. And we have had very close calls with starting to have a little scree slope disappear out from underneath us and us and the instrument start sliding down. There were cutbacks and cut arms, but the instrument remained upright. I'm imagining this right now. Desperate scrambling around to make sure the instrument remained upright. I'm imagining this right now, the desperate scrambling around to make sure the instrument is up. But it doesn't matter if we break our necks.

It doesn't matter if our elbows are bleeding. Get onto it. I don't care. And that same instrument was later in that same field season involved in an incident where there was a hailstorm that required sheltering that might or might not have used the instrument housing. Oh. Checking the density of the human belief? Yes. Yeah, the instruments that, you know, I've largely worked with have been, you know, maybe a little bit lower cost than that.

And thankfully, we haven't fried any of the lightning mapping sensors. So that's been a big plus. But back to the balloon-borne electric field meter, one of the big investments of cost with those is in just getting everybody out to the storm and your whole crew and all of the travel involved just to get to the point where you can launch it. And this is one of those cases where you its normal orientation to measure the vertical electric field, and then had turned one of them on its side to measure the horizontal electric field. So we have this giant tens of feet long lash up of instruments that we're trying to chuck up into the sky. And I guess we hadn't all read our Boy Scout manuals quite well enough, but it got about, I want to say,

30, 40 feet off the ground, and then the instrument chain spontaneously disassembled, and we had two very broken instruments on the ground. Spontaneously disassembled. A red, rapid, unscheduled disassembly. Yes. That's the technical term, yeah. And these balloons are huge, by the way, that Eric's talking about.

They're not a few-foot-in-diameter weather balloon. Yeah, they're manufactured by a custom outfit. I think the ones we got were from somewhere up in South Dakota, and it's a giant polyethylene bag, so a giant garbage bag filled with helium gas that you've uh chased around the plains with and uh gets inflated inside a u-haul truck that also serves as a mobile shelter for uh for everyone if there's uh lightning nearby why do they have to be so large

or they're they're not i mean they have the ones that go up to like you know a hundred thousand feet but that's not the point here, right? These do get up to 100,000 feet as well. We're going all the way up through the storm into the stratosphere, but 100,000 feet might be overdoing it a little bit, but 60, 70, somewhere in there. But they need to be big because the instruments that we're flying are about four pounds or so. If they aren't rapidly disassembling. That's right. Yeah.

I mean, that's a massive increase in buoyancy. And, you know, you get through the storm in a hurry, that flight. I'm just going to imagine the report from this. Massive increase in buoyancy. What did they, what happened such that it disassembled itself? Note the complete lack of blame on anybody. Yeah, it's – my guess is that whatever not was chosen wasn't quite right. So this was a Boy Scout manual failure.

Oh, it absolutely was. Yeah. I mean, there was a book of knots that was part of the research group that everyone sort of sat down with and learned how to tie a knot so the knot sat in the right place so the instrument would be balanced and all of that. I thought you had zip ties or is the zip ties to reinforce the knots now? Yeah, zip ties. Let's see. I don't think we used any for that. Yeah, the primary rigging was a waxed nylon line, and you need waxed nylon because you don't want it to absorb water and become conductive. So there's all sorts of interesting things when you're up in that electrical environment. All of which have been discovered by trial and error. Empirically. Empirically is the word.

Am I to assume that several went up with non-waxed line at some point? That was before my time. I mean, the measurement dates back to before my birth. So, you know, I inherited from my advisor what he inherited from his advisor and science advances. I like that there's this whole collection of tribal knowledge and lore associated with doing these field experiments that aren't necessarily part of, you know, the scientific literature. It's like, this is how we take these measurements. And it's not written in your paper that you took all these steps to get this all to work. It's just stuff passed down from advisor to advisor.

No, you get a rapid increase in buoyancy. And it's your advisor that explains what happened. I know, but there's no manual for this. No manual for decoding what science actually happened? Do you guys have a wiki that you put all this stuff up on for posterity? You know, there probably should be. I think in the past, method sections of papers really got cut because you wanted to get as much in about the research, the cool thing that you discovered as you could.

So when you started hitting up against that length limit for a specific journal, methods were the first things that got knocked out. So I think sort of it comes from that, and that's starting to change as people are doing more supplemental material. But even in lab work, there is a lot of this, we do it this way because it took us five years to figure out that this is the best way to do it that we've found so far we keep trying new things too i mean that happens it happens everywhere i mean in in my lab work which is just a couple of computers we had built up a large store of voodoo and i spent

a week killing the voodoo but that was just because somebody discovered that if you typed A and then B and then C, it worked. And I'm like, yeah, those things shouldn't matter. Order should not be dependent. Here, let me try it all the different ways and verify that you can stop circling the computer twice clockwise in order to make it work. So at some point, knowing that your devices are made of tape and very expensive. And hope. Very expensive sensors and hope. Do you worry about the accuracy of your field bodges? That's what the instruments are for.

Well, I mean, I've definitely gone, you know, with ShotSpotter, we would go out and take data, which is to say we would go fire large weapons at various locations. And some of the field bodges... In various locations. In various, not at, in various locations. Okay, just want to make that clear. Sorry. Yes. Oh, and speaking of field bodges there, that was the time that I needed duct tape

to keep my hiking boots together. Yes. Yeah. Anyway. Accuracy. Accuracy. I definitely have been in the field where accuracy suffered

because we figured out that tape was the only thing we could do. I mean, as long as it's a calibrated bodge, that's probably going to be pretty good. And I know that, you know, just a few weeks ago, I believe you said, everything's a temperature sensor or something since other things do. I love to say that. Right. And it's very true.

And it's the same with a lot of our instruments. As long as you calibrate it and calibrate it against all of the things that we can conceptualize that have major impacts on it, maybe being a bodge isn't necessarily a bad thing. Some of these stay in place for years. The machine I did all my grad work on, that big hydraulic press, it was bodged over and over and over for around 30 years. They were the original schematics, but they were almost completely unintelligible because of all the markings that were dated from late 80s and early 90s and mid 90s and late 90s and early 2000s. And eventually it did become a problem. So we just recently finished redoing it completely. So now it's clean and bodge free, and that'll probably last for maybe 12 months

before somebody wants to try something totally new and they have to add something else on. So I don't worry necessarily about the accuracy unless it's something that we've bodged something where it's going to have some kind of time dependent thing. And I've even seen people try to calibrate that out. And I think it's at some point you get to a point like that with a lot of instruments where you have a maintenance procedure or an engineering approach that gets the job done until the budget allows you to do it right. So for instance, our lightning sensors that are out in the field right now have been there closing in on 10 years. And thermal expansion and whatnot gets after some socketed chips that are part of the digitization circuitry. And the maintenance approach for that right now is you send know, compared to replacing a $10,000 board or

whatever, that's, you know, preferable for a time. But that's something we're certainly looking at now in the next next year or so is to try and get that get that replaced to a proper soldered situation. In a lot of these instruments, you know, I said they were in service for a long time, you would not believe the number of labs that I walk into to work on equipment. And when you open up the control panel for this, there is a Z80 in there running. Or I have seen a couple labs where an Apple IIe was communicating with an instrument still because it works. We're not going to update the instrument, and there have been a few bodges along the way. I mean, that seems very reasonable, as long as they're keeping up with updates to their operating system. Yeah, I'm pretty sure the last security patch for the 2E is a little out of date. how does it go from a bodge to a device how does what triggers the engineering is this like oh my god we got a grant and now we can finally do this right or this is so broken we're saying

you're not doing it right? No, of course. Doing it better, right? Better, better. Less human cost in the future. I mean, Eric, as somebody who has to go out and find this grant funding, that's probably kind of a great question for you. Yeah, I mean, it's a mix of, to be frugal so that if small things break, you can replace them properly at the time. grant money that comes in as well as trying to seek out grants that will allow you to make the replacement or repair that you would like to make. But how do you decide that's how you're going to use your money? I mean, you only get so much money in grants and you have to pay people to go out to the field. So at what point do you say it is better to pay for this to be rebuilt and do the same thing it does right now, or maybe just a little better versus we're going to use this one more time and we're just going to

make it work, add some more tape and it will be fine. Yeah, I think with this board that we're about to replace in the lightning sensor, you look at the amount of time the technician is having to spend and the amount of time it takes him to go make that replacement. And it's pretty easy to come up with a cost-benefit analysis that it's worth doing the change to the instrument to save him all that time. But in other cases, there's painful decisions that sometimes have to be made where like, well, we're just not going to make the measurement anymore and we'll move on to a different science question that uses a different instrument that we have that we

can afford to run right now. And that's maybe a bit of a painful truth about doing research in this kind of a setting. Or even sometimes just the part for that instrument or the company that made that instrument originally is gone. And so when that instrument reaches end of life, you're done with that line of research. That's got to be really painful. Yeah, there are a lot of different instruments out in the field that when they cease working, there are going to be some major pain points. Or when they have firmware issues,

like the recent GPS epic rollover that affected so many seismometers. And there's nobody there to update it. Did it really affect that many? I mean, I heard it was happening and I was like, okay, so only a few things are going to matter because most people handle that fine. Most people aren't upgrading their seismometers every year. Right? Most people aren't upgrading their seismometers every year, right? I mean, these are old instruments that were huge capital expenses at one time, and they're supposed to work forever. Right, and so they should be designing in the epic rollover. Except it was probably 1998 when they were made.

Yeah, early to mid-90s would be a sure thing. I would say you're exactly right. Probably early to mid-90s is a lot of the vintage that got affected by this. And I'm not going to say the vast majority of the seismic network, because it wasn't. But there were definitely some older stations that have been kind of standby solid stations where, yeah, this instrument was put down a borehole and it's been there for 30 years and nobody's touched it. And then suddenly it has major issues on April 6th. And I guarantee you the people who made it weren't thinking, well, this is going to be in use in 2020. Well, they were just so excited to have any GPS at that point. Great. Okay, so have there been times that the bodges have worked better

than the engineered instruments? I mean, have there been times when you went back to the old hacked up system and said, oh, that actually gave us better data? I'm going to say not in the long term. There have been times where we've tried it the new, the improved, the fancier, the engineered way, and we didn't get as good of a result. And we had to go back and figure out what about our bodge was different than what we were doing now and why it was not better when we knew it should be. In the rock mechanics lab, a lot of times that came down to a signal integrity issue or a pressure sealing issue or something like that that we just hadn't thought of or hadn't realized

that our bodge was helping when it, it was actually having an effect on that thing that we didn't even initially consider in our redesign. Yeah. I think that that's, that's always a concern when you have a, when you have a tested, a tested design that you have a lot of miles on. And, you know, let's say you switch out your prop and vane anemometer to a newer like ultrasonic or resonant cavity type anemometer that is now measuring the wind in this totally different way. And you know, your old one works well in this old environment, other than like maybe a little water intrusion problem on one of the boxes. So you switch to this new one thinking it'll solve all those problems. And a lot of times you don't know until you're back out in the field whether or not that investment you've made to change to modern technology was fully worth it.

And that's how we lead to having tape in our toolboxes to find out that now the field is different. Whoops. I understand that. And when I'm engineering something, when I'm building something, sometimes with software, I get irritated when I have to take a step back and deal with my build tools. Because it's just one of those things that I don't enjoy.

Do either of you, both of you, get frustrated when you have to stop doing science and take care of the engineering? Yeah, I mean, I can certainly speak to that from recent experience. So I wrote this code as a postdoc, you know, 10 years ago that I thought was for my own research purposes. And there's now essentially operational processing running on this code for the geostationary lightning mapper instrument. And it's processing, uh, processing those data and delivering images to the national weather service for their operational use. And so, um, that's become a bit of a, of an engineering burden on the software side where I now get like trouble tickets filed by someone in Washington, DC. And, uh, you know, it's like I'm diving back into whatever, you know, nine-year-old, nine years ago me wrote and trying to excavate my logic from that point in time. It's thoroughly unit tested though, right?

Absolutely. Absolutely. As in it worked on my unit. In fact, that was one of the bugs that I had to fix. The data that I take in decided to change its units from square meters to square kilometers. And so that was one of the hot fixes that had to get written real quick is like, check to see if unit, if not rescale. The new definition of unit tests. I mean, unit is a one thing, right? So I've always thought of unit tests as it worked once, and you're done.

Oh, breaking the rules. Yeah, and I mean, I'll say I don't really do science anymore other than as a hobby when something cool happens, and I'll go pull some data and look at it. I mostly do now the engineering things to help scientists do what they need to do. And so they don't have to do the engineering things. But yeah, I mean, hardware, software, you can actually write tests. With hardware, it's hard to write a test for does this instrument measure the same thing that the other instrument did.

So it's a really interesting challenge. And as now we're getting more and more advanced sensing methods, we're able to store and process more data. And now we're measuring things just because we can, not necessarily because we know that we need to measure X, Y, and Z. Well, now we have the capability to measure A and B as well for very little more cost. We're doing that, which introduces complexity to the system. And that's been a fun problem to grapple with. Yeah.

Some of my clients say they need an accelerometer, or we determine that what they need is an accelerometer for what they're doing. And then they look around and it's like, oh, well, but we can get gyros too for the same cost. And, you know, the magnetometers aren't that much more. We'll get that too just in case we want the data. And now they've got 3x the data they have to store and they're suddenly wanting quaternions and Kalman filters

with all they really needed was a tilt sensor. But it's so, I mean, these sensors are amazing and they, they just, it's easy to want more. Are there sensors that you've been working with, John? I mean, IMUs have been out forever. So are there new sensors? Nothing a novel, totally new thing comes to mind. You know, Eric mentioned doing ultrasonic measurement of wind speed now, so get to deal with some of that. But more so just now we're everything we're measuring all the state variables that we can about our lab or our atmosphere. I've been dealing a lot recently with displacement measurement and very precise displacement measurements. So using eddy current

sensors, capacitive displacement sensors, and a few of the older DCDTs and LVDTs, the little variable transformer sensors. And so those all have their own challenges. They all do things a little differently, have different bandwidths. And a lot of times the scientists may not know exactly what they're looking for. I want to subject my material to these boundary conditions and see what happens. So if you say, well, what's the maximum frequency that you expect? It's unknown at the time, maybe because there's not a model that can accurately reproduce that, so we don't know exactly what to expect.

So some of it's sensor challenge, some of it is not knowing what we're trying to sense challenge. Yeah. We were at the MBARI Open House, the Monterey Bay Aquarium Research Institute. And they were talking about their benthic event detection system and how they thought they knew what they needed. event where this little rock-like thing was carried by mud down the canyon, it went so much faster than they expected and it had so much more force than they expected. But I mean, you're going down under the sea.

What do you, I don't know what's down there. Now we know a little bit more, but it would have been nice if it hadn't broken. Right. And next time you can design it and, you know, we'll iteratively sort of find that point. But a lot of times you don't know what to expect. You don't know what the conditions are. Or the thing that you completely didn't think about, like some of our tilt meters. So I make very micro-radiant and sub-micro-radiant tilt meters for geophysical tilt applications. And some of them went down to Antarctica. And we had to think a lot about things like, how does the cable respond to temperatures that it's going to experience there? Does it get really stiff? And then when it gets windy, is that now stiff cable going to telegraph too much

vibration to our sensor so there's the thing that you need to sense and we didn't know really how big what we were trying to look for there and the tilt signal was and then we also had this whole other host of things that weren't what we were trying to sense and we're trying to keep out and they included things like what happens to cables and zip ties at minus 30. So what do you do? Go down and get some dry ice and really chill up your lab? Or how do you figure it out? Yeah, so I actually went eBay searching, one of my favorite things to do,

and looking for bits of scientific equipment. And we bought a thermal chamber. And we put these things in a thermal chamber that would go down to minus 60 and so we just had a bunch of cable samples in there and then took them all out and started whacking them on the table to see which ones got the least rigid after being exposed to those temperatures then we put the instrument in and calibrated at these different temperatures uh just the things that you end up having to figure out that you never dreamed of. Do you have to do similar things, Eric?

Because, I mean, a balloon going up to 60,000 feet, that's a harsh environment as well. It is. There was, my first job as an undergraduate was to take these motors that drove the spinning apparatus. The measurement works by inducing charge on these spheres as they spin in the electric field. And so this motor has got grease in it to lubricate it, and that's designed for sitting in a manufacturing floor somewhere at roughly room temperature. Well, that grease gets awfully stiff when it goes up to minus 40, minus 60 Celsius. So my first job was to disassemble all of those motors and all of their little gears, degrease them, and put in some sort of low-temperature oil that would work way up high in the sky. Seems inefficient. Couldn't you just order them the way you wanted them?

Or is this because people make these things, but they don't make the 10 you need? You know, I'm not sure. I have not gone back to look if you could find a spec sheet that would give a temperature rating on a motor like that. it's actually cheaper to just pay an undergraduate to do the work and, you know, or do the work yourself and wind up, you know, just taking care of it that way. How are ultrasonic wind sensors different than what I usually think of as an anemometer? No moving parts. You can do that in such a way that you can even do a three-dimensional wind measurement and a wind measurement at a much faster rate, up to, say, 50 hertz in some cases. So no moving parts, which is nice, and maybe some different issues to deal with in terms of electrical environment and things underneath the thunderstorm that might cause a little different kinds of interference on your measurement. Okay, I'm not understanding. When I think of ultrasonic things going out and sounds coming back, I'm thinking like bats and echolocation and that sort of thing, but that wouldn't work on

wind, would it? Right, so do I understand how the measurement works well enough to explain this? This is interesting. I didn't mean to quiz you. No, no, it feels like I should, right? I have a,

there's a tower that we, that we run that has 10 of these up to 200 meters in height. So, and then the realization, this can't work. This has never worked. Wait a minute. That's just noise. We've been measuring all these years.

But at very high bit depth. That's right. Absolutely. And at 50 hertz. So it's many hard drives full of noise. How does that work? I'm going to have to go do some reading after we're done recording, I think. But it's, uh, but

it's, it's definitely a speaker and a, and a receiver type setup. And it might, it might be something similar to like a Doppler, uh, Doppler type process where, um, where the, the signal is shifted, uh, because, because the medium that it's traveling through is moving. That makes sense. Okay. I think some of them use Doppler, and then there's some of the cheaper in quotes that are sub-several thousand dollar sensors. I think they just use time of flight difference and assume that the sound is advected by the medium. Oh, okay.

But what are they bouncing off of? I mean, is it, they have to, are they bouncing off of something? So they emit in one spot, and then the sound is carried by the air. Oh, so there's air in between them. It's not that it's bouncing off and seeing a car in front of you. It's bouncing off. It's casting it into the air and then it's measuring the difference.

Okay, so it looks more like a capacitor or a... Because you know the speed of sound right, for the given conditions probably, and then you can figure out if there's some additional and temperature yeah yeah 0.3 meters per second per degree c yeah so a lot of these actually look like um

two sets of three fingers that are staring at each other across a maybe one foot gap or something like that okay i have a much better picture this makes sense to me. And Doppler and shift both make sense. Okay. That's makes so much more sense than what I had in my head. John, I hear you've been putting sensors on cows. the fun things about making sensors for scientists now is they come up with lots of interesting things they want to measure. And one of the more recent jobs that just came in was we want to measure the jaw motion of a cow. And we want to track this cow and then telemeter that data back. And so we're looking at doing some various piezo sensing methods, but I never dreamed of interrupting. She's over here pantomiming with her jaw being a cow for some reason. And I just thought everyone should know. Well, it's not something that I ever dreamed that I would have these,

you know, little PCBAs with a programmer and debugger hooked up to them connected to a harness for a cow sitting on my desk. And then the cow's sitting on your desk? But actually, that leads me to an actual question. How do you prototype this? You don't have a cow in your lab, presumably. And tell me if you do.

He does like eBay. I don't know. I know he has trouble not buying things on eBay. Right? No. So there, there's no cow in the lab. I did get a building and it does have a nice little, you know, courtyard area that's fenced in, but I have not put a cow out there for this contract.

So in this case, I'm just moving the harness around and making sure that I record a faithful, at least what I think is probably faithful, uh, representation of the motion that I put in or the accelerations that I record a faithful, at least what I think is probably faithful, representation of the motion that I put in or the accelerations that I put in. He's been doing the chewing cow thing too. Yeah. And they're going to actually put this on the animal. That's good. Do the testing that way.

Yeah. Well, how do you find people who need these sorts of sensor solutions? They find him. So it's a difficult thing because there's not a lot of people that do just the sensing side. And sort of, and if it's in the geoscience side, I can know enough about what they're trying to do scientifically to make some intelligent decisions about what we need to do on the electronics and hardware side. In the cow case, I'm learning along with them on what's the best way to detect jaw motion. But so people find me a lot through other work that I've done. So they've talked to somebody that's used products that I've developed for them, or they'll ask somebody to conference and somebody's heard of that I really enjoy working in because it's a different problem.

Not quite every day, but it's a different problem pretty often. And there's a big need for it because a lot of times the scientists are writing grants. They're managing the actual project. They're doing data analysis. They're working with their grad students. They don't have time to figure out what changed in the latest version of whatever CAD program or figure out where their PCB layout went wrong and why they're getting massive crosstalk between these signals. And that's where I come in and take that work away from the research groups and let them focus on the research that they enjoy. And I focus on supporting the science,

which I found out I enjoy more than actually doing the science. That sort of thing is really a godsend to have access to because, you know, again, deploying our lightning sensors in the field, I had come up with some sort of elaborate tripod design and spent a whole bunch of my time trying to come up with this. And we have a technician on our staff here who previously had a career as like a welder in Alabama and then did some engineering support for the V-22 Osprey tilt rotor helicopter aircraft. And anyway, lots of, lots of experience. And so I described my problem for him to him, you know, what, what I wanted to have done. And he goes to

bed overnight and said, I woke up this morning and I think I can get some steel and be about $500. And I'll get do all 12 of your sites and it'll be all done. Just, you know, give me a, give me a credit card and I'll, I'll go get started. So it's, it's fun to have, it makes the resource more fun when you have a team that can, can take on some of those, those ingenious mechanical and, and electrical or whatever you need kinds of projects. And they enjoy it. I mean, you enjoy doing the science and they enjoy doing the engineering. Everybody's happier than when you're swapped.

That's exactly right. And I mean, it's really delightful to watch people with expertise that I don't have doing the thing that they're good at. And so a diverse team from all walks of life can be really useful. You have mentioned the Global Lightning Mapper. Could you tell me more about what that is? Yeah, everyone's familiar with the weather satellites that NOAA launches.

They're the ones that give you the pictures of the clouds that you see on the evening news. And the most recent series of these that NOAA designed and launched a couple years ago now have a lightning sensor on it. And the way this thing works is it looks for the optical pulse of lightning produced even against a bright cloud top. And so you might imagine there's a very sensitive CCD and some interesting electronics that go into extracting that very tiny signal against a bright sunlit background. And it does that well enough that it's about 70% of the lightning, 70% of the lightning flashes that are created are detected. And there was something about CCDs and meteors that I should ask you about. Yeah.

So it turns out that when you, when you design a sensor that's good at detecting small optical pulses that are transient, that same sensor is very good at looking at meteorites when they reenter the atmosphere and explode. It's called a bolide when it does that. Sometimes these are very, very bright, but they, you know, they come in all shapes and sizes. And there's been probably dozens now that have been seen, you know, since the, since GLM was launched, and the the bolide community is very excited about this, because they have a whole new hemispheric scale data set to look at to try and build up a sense of what, for instance, the size distribution of these

meteors are and the energy with which they re-enter the atmosphere. Even things like the light curve, as the bolides re-entering and heating up and then exploding, that this lightning sensor runs at 500 hertz. And so there's some detail there in that light curve that they can retrieve. Did you expect it? I don't think so. I know I certainly didn't personally. And then someone came to our instrument science team meeting from a bolide group,

and they were very excited. And so it's been to a couple more meetings with those folks. And as John was kind of saying before, it's fun to get to learn about these new areas that you come across in the course of the work. Yes, that's one of the best things for me as being a consultant is getting to just learn about all of the things and see what's working and learning about the different technologies and then finding the client who needs a technology that they didn't know and being able to say, oh, no, you don't really need to hire me. You just buy this entirely finished product. And it's, I don't know. I like doing that. Yeah. Okay. So, Geocast does this thing with papers. John, could you introduce?

The segment. The segment. Yes, the segment the segment yes the segment right so as our listeners will know it's time for everybody's favorite segment of the show fun paper friday and in this segment we select a academic in nature paper and find some interesting connections that it makes to the real world or it's just a bizarre paper. And in this case, Alicia, you picked a really interesting paper for us. Yes. I mean, it was interesting to me. There's a guy named David Huffman,

and for software engineers or embedded engineers, this is Huffman of Huffman Encoding, which is a good way to make things a little more dense and to make sure you have good error checking. So I can link to Huffman Encoding, but that's not what he's famous for in this paper. Eric Domaine and a couple of other authors, although I know Domaine did his PhD research on this, wrote about Huffman's work in

curved crease models. This is origami. And when you fold something in half with a straight line, you get a piece of paper that is now flat again, right? But if you, instead of a straight line, you draw a U on it with a, say a ballpoint pen and you write really hard so that you get a pre-crease so that it will fold along that line. If you fold that, it becomes 3D. It becomes a curve and it has to be 3D. You can't fold it into a flat piece of paper again if you are also folding this curved line. And that one U in a piece of paper is the tiniest piece. That's addition, and he's doing advanced calculus with this amazing stuff.

Yeah, and I had never heard of this before you sent this paper over. And it was really fascinating, the different pretty much conical sections that he used to make these shapes and in the paper they sort of go through building up to more complex shapes by taking these and they said he used this little sort of a circle spring-loaded pen tool like he's a kind of sort of like a ballpoint pen and vinyl to do this. But in the paper, the authors do it with normal paper. They also replicate his method and they do some simulations, some solid modeling as well to try to understand exactly how some of these worked and just sort of play with the advanced math. And it is math.

I mean, if you just do, like the U-curve will work because you'll get some shape out of it. But there are many things you can do that don't work. And you can just end up with a crumpled ball of paper. Trust me, I know. And yet there's one here that looks super hard called tessellations that you end up with ellipses that go in and out of a paper. And I folded it and it actually wasn't that bad. You just have to do the creases. He uses the pin tool. I actually use a metal mechanical pencil without the lead in, and a lot of people say that ballpoint pens work. And if you let it fold the way it wants to fold, it curves and the paper moves. And at the end, you have something that doesn't look like paper.

I encourage people to go look at this because at least has been doing this and it's kind of just these weird things that you don't think paper should be able to do. And that sounds kind of like, what are you talking about? Um, but, uh, when you see these, it's, it's, you think it's like, uh, cast plaster of Paris or something with intricate, uh, curvilinear designs built into it. But it's just paper that once you score it a certain way and manipulate it, it just kind of wants to be in this new shape, which doesn't make a lot of sense. Yeah, it looks very sculptural, like something you might even see, you know, attached to a building at great expense and, you know, cast

in concrete or something. And, you know, they provide templates in the paper, it looks like, that you can just kind of trace out. I haven't done this yet, and maybe Alicia can clarify if I'm right on this. But basically, you just trace along these lines, and then the shape sort of pops out? Yeah, there are little dashed lines and then dashed dots. And one of those should go in and one of those should go out. And if you're going to choose one, there's one that makes a flower called inflated vertices or sort of a swirl. What would you call it, Christopher? Pinwheel kind of thing?

Yeah, that one is probably the easiest one of the ones I've tried. Okay. And it gives really satisfying results. Have you just been doing this with printer paper, or do you use special paper, or do you use vinyl? What have you found that works well? I have origami paper, and I have to say it does not work that well. You want to use something a little thicker. If you have resume paper around, you know, the nice stuff.ap construction paper works pretty well. But it has to be something that when you push down on it,

you can feel the line, but you aren't ripping the paper. Okay. So he said in there that the vinyl that he used was 10 thou thick, so I guess it's time to pull out the calipers and go to the craft store. I've had pretty good luck on anything that feels beyond thin paper. Like it doesn't have to be super thick, but it shouldn't be, wow, this is really cheap paper. Okay. And so looking at some of these shapes, some of them look very not natural, like the tessellations, though I guess you could sort of

think of it as a Horst and Graben structure in geology, but that's not what it would really look like. Some of these things to me looked geological in shape, like I could see an alluvial fan up in figure three, or some of these things. I'm wondering how much of a connection there would be to things like plate boundary collisions or geological folding, because you're getting some really strange shapes from something that was effectively a flat sheet. That was why I wanted to suggest it, because it seemed very much like that. And if you look at that pinwheel one, if you add some more lines which aren't on here it ends up looking kind of like a mountain if you add enough lines in there and it becomes this very craggy mountainous shape and i was like okay so is that how mountains work yeah and i just looking i said some of these it really does

to me look like a geological folding or plate boundary collision, which would make those mountains. You're taking this solid thing and forcing it into a three-dimensional shape. So I would imagine that some of the mathematics overlaps pretty heavily. This is more of a structural geologist's area. But I could definitely see some plunging folds that looked like this. And I was wondering with those even, if you've got different strata and you bend it, sometimes you can get movement in between some of those different units effectively faulting. So would that be like doing this with a stack of paper where the sheets can move and slide on each other. So you can have these subparallel faults, and then you're forcing this into a more complicated shape even then.

A lot of these complex shapes that sort of pop out of what seems like a simple structure into three-dimensional space reminds me of a guy I follow on Twitter named Henry Segerman, who's also a professor at Oklahoma State. And his Twitter feed is full of all these sorts of unusual mathematical objects. I've come across videos that someone's put together of actually like flying through hyperbolic space as well, which is one of the conic sections. And there's a bunch of conic sections in these paper folding things. And so it's he seems to have a lot of neat work about, you know, how to take these maybe,

I'm wondering if some of the stuff he studies is related to the mathematics behind this paper folding even. So far it's been computer scientists and mathematicians that seem to be studying it or even doing it in their free time. I agree with the, I mean, the time of arrival and the hyperbolas make sense because you're fitting least, least what curves? At least, like a sphere is the least amount, least area. least what curves? Like a sphere is the least area. Right, right. Sorry.

There must be words. There must be words. It's a minimization problem. It is a minimization problem. Exactly, yeah. Yeah, it's actually solved as a least squares minimization problem. Yeah. And many of the things, I don't think it says, no, in this paper it doesn't say, these are computer generated, don't try to do them by hand.

But I get a book that is, this paper led me to a book that basically said this is a math book, so get with the program. And it said, don't even try this on your own. You have to do a computer rendering of it because the shapes are very specific. The angles are very specific. They lied, though. I did it by hand. I'm tough that way. Ha ha. Yeah, in figure 14 in this paper, you see two lines that look very similar.

You can see subtle differences between them, but very similar. But they say if you crease them tightly enough, you see a pretty dramatic difference in the actual folded work. This is the one column design. Oh, yes. I tried that one and it was weird because it is just two lines that, I mean, it's straight, straight, a little bit of a curve, straight, straight. And then underneath it's the same curve just split by a little bit. at making this pleat in the paper that somehow leads you to looking like you have a, like a part of a flower in a vase. It's the stem of a flower in a vase. Wait, that inner column is part of the paper, not something else? Oh yeah, that's part of the paper.

That's not, some of these are intuitive. You look at them and go, okay, I can see how this moves into that shape. But with that one, these two very simple curves, that doesn't make any sense at all so i'm sorry this doesn't this doesn't work not physically possible this is this is this all lies i made that one didn't i show it to you i don't think you showed me that one i think i made that one out of origami paper so it's pretty crappy well even there that's one where you you start thinking about taking just a pretty simple you know sedimentary section and applying some forces to it and you get all these crazy compound looking folds that are a nightmare to try to map in the field and figure out what happened and to see just these small differences in these two lines that

got creased and what the dramatic effect was you you start saying, okay, maybe I can see how this geology got so complicated so quickly. Yeah, it doesn't, it could be instantaneous almost, right? It's something happens to curve these things in a certain way, and then suddenly it settles into this position very quickly. Right. But are there sheets that act like paper where they, they want to stay attached? Definitely so. I mean, you've got a pretty, they may violate some, I'm sure they violate some assumptions of the math of adequately thin media compared to its length and width.

But I would say if you've got a nice, thick, chunky rock unit, a big sed section or a big piece of carbonate or even a tectonic plate, if you want to go to that level and you're looking at something that's 10 to 40 kilometers thick, that in mass is going to hang together as one sheet. Even if it gets folded. Right. So there are some great, I would definitely encourage folks to look up geologic folds, both from the aerial and sort of the conceptual side view. And you can get these folds that range from, I've got some in hand sample that are a few centimeter scale, all the way up to folds that are many kilometers scale or even larger for tectonic scale. And yeah, these rock units behave as one entity. They don't necessarily split apart. Sure, there is some faulting along

the edges of the fold sometimes, but in especially the smaller ones, you don't always see that. I typed it into Google and now I'm looking at a rock that's zigzagged back and forth. It does not look real at all. it to 100,000 years of erosion. And now you've got a completely different cut through this three-dimensional surface that was made in a complex way that you don't know. And you have to go back and figure out how that happened. So when you get a fold that is not perfectly horizontal, the axis of the fold plunges into or out of, comes out of the ground, which is relatively common, you can get some pretty weird surface expressions of that after things have eroded away. Hey, that one looks just like the cone one that's in the paper.

I'll put those in the show. The cone reflected seven times? Yeah. Yeah. And a lot of these geologic features, you know, I said they're on all scales. But when I say a fault, a lot of times you might think of a tectonic fault, right? Say the San Andreas fault. That's a very atypical fault.

But you would think of something pretty large with lots of displacement on it. Faults can be a few tens of microns thick and have a half a centimeter or a centimeter of displacement on them as well. There's all kinds of interesting things about the fractal nature of geologic features, but you can have folds large and small and you can get some things that you shouldn't think rock should be able to do. But over a long enough period of time, rock is able to do it. That's how you get salt plumes that rise, salt rises buoyantly over geologic time and creates all kinds of interesting features in the Gulf that we use as mechanisms to trap oil and extract those resources. So rocks can do some pretty interesting things. And I

think sets of geologic units isn't really that much different from sheets of paper. extract those resources. So rocks can do some pretty interesting things. And I think sets of geologic units isn't really that much different from sheets of paper. Sweet. That means I win fun paper Friday, right? Yeah, I think you win. This is certainly one of the, I wasn't expecting to be able to tie this to geology initially, papers. And i always try to tie them too to some sort of i wonder if there's an industrial application some kind of interesting manufacturing thing where you could use this to to speed up some manufacturing process i would think the folks that do uh you know automotive body design would have to be really familiar with this

when they're doing the really odd geometries that you'd have in a sports car or something. I think they just stamp them. Do they? Okay. I don't know if the original, they'd have to do something like this, but I think they just, because they have to make thousands a day.

And you can stretch sheet metal too, right? I guess you can stretch rock too. So maybe you can get away with not having exact curves and creases. I did see that there was some packaging that used some of these principles, but it was to make the packaging prettier. I mean, the goal was to make the packaging prettier. I mean, the goal was to make it look fancy. Right. I would say if you're making some kind of a shock-absorbent packaging, maybe this wouldn't be the most efficient way to do it, but maybe it would. I don't know.

Buying a flat sheet would certainly be more economical, I would think, than making some kind of custom mold. And I'm sure bending it is easy. Right. You just need a lot of ballpoint pens. It's been really fun to talk to both of you. Are there any things we should have talked about that we didn't? No, I think this was a great discussion of how science actually happens. Yeah, absolutely. Learned a lot from everyone.

John, do you have any thoughts you'd like to leave us with? Well, I guess I would just say, going back to some of the things we were talking about earlier, there's really no substitute for trying something or doing something. You may think you know that it should work, or you think you know what you're looking for, but even that model is going to be incorrect in some way because models are only that, models. So just, I would say, go out and try to measure something or try to do something, even if that's not making a widget that you're going to sell 100 million of or

a scientific instrument or whatever, just start on a project or if you're me, try to finish one as a goal. Eric, what about you? You know, the advice from John there, I think is really solid and something I need to remind myself of all the time. You know, I have overthinking and overplanning tendencies, which means I have the perfect job. But, you know, there's certainly a time for thinking carefully. But that I think comes more when it's time to share your work with others and really formalize it. And at this experimental stage, it's so helpful to get moving and even, you know, try doing some of that writing down for others along the way. I've noticed that when I can get into that mode, I'm actually more efficient at

discovering the next thing I need to do because I've actually tried to explain it to someone as well. So yeah, give it a shot and get started. I think that's usually very good advice, both finishing things and getting started. This has been a joint show between the Don't Panic Geocast and Embedded. The hosts of the Geocast are John Lehman, who is with us today, and Shannon Doolin. John is also the founder of Lehman Geophysical. The hosts of Embedded are Alicia White, that's me, and Christopher White. Our guest has been Eric Brennan, Associate Professor of Atmospheric Science at Texas Tech University. It's good to talk to you, John and Eric. Thanks.

Thank you. Thank you. I would like to remind those listeners of Embedded that we're having a party, real-life, physical, in-person party in Aptos, California. That's near Santa Cruz, California, on September 7th. Embedded300.eventbrite.com, or you can email us at show at embedded.fm or hit the contact link on embedded.fm and I will send you the link. And now thank you to Christopher for producing and co-hosting and thank you for listening. And thank you for listening to our show. You can contact us at show at don'tpanicgeocast.com.

Thank you all and have a good week. Embedded is an independently produced radio show that focuses on the many aspects of engineering. It is a production of Logical Elegance, an embedded software consulting company in California. If there are advertisements in the show, we did not put them there and do not receive money from them. At this time, our sponsors are Logical Elegance and listeners like you.