[Announcer] This conference will now be recorded.

Good morning, everybody. Welcome to our 2020 fall EcoFOCI seminar series. I am Heather Tabisola and I co-lead the series with Jens Nielsen. This seminar is part of NOAA's EcoFOCI bi-annual seminar series that's focused on the ecosystems of the North Pacific Ocean, the Bering Sea and the US Arctic to improve understanding of ecosystem dynamics and applications. Sorry. Whoever that is, can we mute? Caller three, okay, that should fix it. And the applications of that understanding to the management of living marine resources. The seminar has been going on since October 21, 1986, and has provided an opportunity for research scientists and practitioners to meet, present, develop their ideas and provoke conversations on subjects pertaining to fisheries-oceanography, or regional issues in Alaska's marine ecosystems, including the US Arctic. And you can visit the EcoFOCI webpage for more information at www.ecofoci.noaa.gov. Once again, thank you so much for joining us here today as we continue this all virtual seminar series. You can find our lineup by the OneNOAA Science Seminar Series online and also the NOAA PMEL calendar of events. We are here every Wednesday through next week, which is December 16. Please double check that your microphones have been muted that you are not using video. And during the talk, please feel free to type your questions into the chat. Jens and I will both be monitoring to the chat and we will address those with Kim at the end of the talk. So today we have Dr. Kim Martini, who is a senior oceanographer at Sea-Bird Science, which is locally here in Bellevue, Washington. And in her role she makes oceanographic data better through sensor development, technical product support and customer training. And if you don't know Kim is actually part of our EcoFOCI family. So after she received her PhD from the University of Washington in 2010, she attended the University of Alaska Fairbanks as a postdoc, and then spent three years there and then came back to Seattle, completed her postdoc through JISAO, which is now CICOES, and then also, as a research associate with us at EcoFOCI. Kim's other passion is new and novel ocean observing technologies. And quite simply, she likes to say that her goal in life is to throw expensive, I'll say stuff today for the matter of who we're talking to stuff into the ocean. She also believes that broad scale data dissemination is vital, but often overlooked, and part of oceanographic research. And her goal is to create simple and universal tools to not only quickly process data, but for researchers to easily share this data using standardized public repositories. Kim is also a really amazing science communicator. And if you don't follow her on Twitter, you definitely should. I always look forward to the Twitter wars with internal waves, pictures and discussions, one of my favorites. And with that I'm gonna turn this over to Kim will be sharing her work today titled "The Science of Sea-Bird Scientific: "How design, calibration, and data mining are being used "to continually improve oceanographic sensors."

All right, well, thanks Heather. Thanks for that great introduction. So I'm just gonna pop up my title slide here. So that I mean really, today I'm just gonna be talking about the science of Sea-Bird Scientific. And this is actually a talk that I've been wanting to give for a long time. Heather asked me to do it. If you don't know our daughters actually go to daycare together on NOAA campus. And so she was like, "Do want to give a talk?" And I was like, "What am I gonna talk about?" And this is something that, you know, I've been at Sea-Bird now for just over four years, and I do a lot of training, and I've learned a lot of stuff. And it really, it's just kinda how we do things and getting that across. And this is I do a lot of seminars, and a lot of it is just, you know, either training or telling people how, you know, how their sensors are performing. And this is kind of just the backbone of the understanding of it. And so I'm really excited to give this talk. So thanks, Heather, a lot. Heather did go over my background a little bit, but I do want to give you a brief introduction, as well. You know, I've done a lot before I actually entered Sea-Bird I actually did a lot of work, integrating and working with the instruments that I actually helped build now. And so, you know, as part of these groups, I went to see a bunch. I put lots of different instruments and sensors on different sorts of platforms and dealt with a lot of different data. So I feel like you know, in a lot of ways that I have a lot of the field experience and a lot of just the technical experience of the data analysis experience as well. And so I joined Sea-Bird in 2016, and I've been there since. For those of you that don't know the science of, and let me just talk about this is really what I talked about now, the science of Sea-Bird Scientific isn't just science. It's there's a lot of pieces that actually go into this. And one of the things is that we are a company. So we're a company that makes ocean instrumentation, we make sensors, we make systems. But you know, it's not just science, there's kind of three main parts of it that I'm really gonna talk about here. And the first part, I mean, really is the science. That's kind of what the backbone, that's where we came from. But a lot of it has to do with the manufacturing. So being able to actually make instruments repeatedly that can do the things that you want them to do in the ocean. And then the last part, I think that is really important as well, it's the key of this is calibration. So that when you throw a sensor into the ocean, it's actually giving you the data that you want. So you know that you know it is measuring one degree C that's what it's giving you and our calibration facilities are really kind of what makes that possible as well. So today, I'm just gonna kind of go talk about I'm just gonna give you the outline, I'm going to talk a little bit about the history of Sea-Bird what we are as a company, you may know us from across the lake in Bellevue. But we're actually I'll talk a little bit more about our history. Then I'm gonna talk a little bit about the ins and outs of sensor design. What makes a good sensor? Like what do I think what do we think about when we look at sensor performance and how we're designing it and how to make it better? The next part is calibration. What is metrology? So it's not meteorology, it's metrology. And that's actually just the using standards to really understand what your your measurements mean. And then finally, I'm gonna talk a little bit about our manufacturing process. Because I think as part of when I came actually from, one of the most interesting parts for me for actually moving from NOAA to this, to Sea-Bird was actually that you know, it is manufacturing and we use lots of manufacturing techniques and really how that works in to actually make all these sensors and it's I think it's really cool. It's a really interesting process. And it's always it's a whole kind of different sort of learning and process structure. Okay, so who is Sea-Bird Scientific? So we're actually right now, we're actually a conglomeration of three companies. And so we were based. We're actually right now, we have merged three companies that make three kind of different sensor groups into one giant company. So that's Sea-Bird Scientific. So the first one was Sea-Bird Electronics. It actually was founded in 1974. And it came out of UW in the Applied Physics Lab and that was in over in Bellevue. So that's probably what you're familiar with as well. The other company was Satlantic. They made chemical sensors, and that was founded in 1990. And it was in in Halifax, Nova Scotia. And then the last piece is WET Labs, which is down in Philomath. And so it was founded in 1992. And it also joined Sea-Bird Scientific and so all these companies are actually bought out by a larger company called Danaher, you've probably never heard of them, they're a Fortune 500 company. But we operate kind of in that larger corporate umbrella as an independent company. And it's actually a really good partnership, because there's a lot of things we gained from that. And there's a lot of things we bring to the table as well. And so we kind of have a lot of benefits of having a huge company. But we can also be really independent as well. So it's pretty good. So kind of I like to kind of put up this plot of this, I did a little digging, and I found about the history of Sea-Bird Scientific. And you can see some of these little snippets of history from all these companies. I like the one on the left, the first location of WET Labs, it's that building on the left behind the red car. So you know, everything's coming from pretty humble beginnings. We have some early WET Labs marketing, which is one of their fluorometers, which I think is next to an abandoned pool, and it looks like a bottle of Michelob. We also have the start of Satlantic, so they actually really they're Satlantic, they were doing a lot of satellite data. And then also the start of Sea-Bird when they were putting the first early conductivity, you know, the the precursors to what we use in CTDs now. The early conductivity sensors on SPURV, which actually came out of a APL, which is pretty much just a repurposed torpedo, and that was used to do turbulence, and then also, the history that, you know, really one of the biggest ways that you know, Sea-Bird CTDs really came in and that's typically what use on boats was, through that technology was really really developed through WOCE, which was the World Ocean Circulation Experiment. So that was sort of one of the biggest coordinated efforts to actually look at the physical properties of the ocean in a really coordinated way and have really standards to make sure that we can actually do this kind of stuff. Okay, so I just kind of really want to briefly just 'cause I know that not everyone knows what Sea-Bird is. So what do we make? So we make a lot of things. One of the most common things that you've probably dealt with is we make the CTD package, we make the instrumentation on there. We make the sensors, and then we'll put together you know, the big bottles and stuff. And we'll put that together for you. We also make all the little CTDs that go on.

[Heather] Hey Kim sorry, this is Heather. Hold on one second.

Sure.

[Heather] Something when really crazy with audio. And, can you hear that? Is that you? I don't [laughs]. You can hear it. So it's probably not.

I cannot hear it.

[Deep Voice] I can hear it.

[Ned] Yeah Heather, this is Ned, I can hear it too. Somebody must have just come on and they don't have their phone or their computer muted

[Heather] Thanks Ned, I could see that other folks could hear. I just want to make sure that Kim can get that sorted, nope. I'm just looking at the new check your audio?

Yeah, I'll check my audio.

[Deep Voice] I think it's your audio Kim.

Hold on a second--

[Heather] Yeah, 'cause it's only when you're talking now. Sorry. Oh yeah totally, it's just...

Let's try. Can you hear me?

[Heather] Yeah, that's good.

Okay. Sorry. I'm sorry to interrupt you, but you sound much better now. Can you hear us?

I think I can hear you. Can you hear me?

[Heather] Yeah, you're good.

Okay, is that better? Okay.

[Heather] Much better. Okay, continue. I'm so sorry.

No, I'm glad you stopped me. Do I need to go back a little bit?

[Heather] You might want to, yeah.

Okay. So okay, so what does Sea-Bird make? So I think the product that most people are familiar with for us are CTDs. So that's the big shipboard conductivity temperature depth sensor that you typically use on lots of ships. I kind of included some pictures here. I thought it was really fun. I actually got to I have lots of pictures from when I was at EcoFOCI. I have all the Sea-Bird instrumentation that I was using at the time. So it was kind of fun. The live pictures are from that I don't actually have other pictures as well. But that's kind of what we make. And then we make chemical sensors. So we make oxygen sensors. We make pH sensors. We also have this phosphate sensor, which I really like it has these canisters that are based upon what you use for Xerox printers, we make active optics. So this is like fluorometers, backscatter, we make passive optics to measure the amount of ambient light in the ocean and we also make systems so we also make Argo floats, which I know some people are familiar on. And this is actually a Sea-Bird float that I watched deployed before I actually worked at Sea-Bird, which is really cool. And now I actually do a lot of work with these as well. So it's kind of a really nice to see that progression for me, you know, having been one of the people there that puts these on the ocean, okay. So I really wanna start, you know, kinda what is a sensor? and so that we are obstensibly a sensor company, that's where we came from, you know, we were making all these kinds of different types of sensors, we're making, you know, fluorometers and this. And so what is a sensor? What does it do? And a sensor really is just a really basic description of that is a sensor is a device that allows a physical characteristic of the environment to be converted into electrical signal. So that's all we're doing, okay. So it's composed of kind of two things. It's an active element that has a property that changes in response to a physical characteristic, and then a circuit that kind of changes the signal into something that may be measured. I think, we actually, you know, we encounter sensors all the time in our life, like I think this is a really good example of something that we just have to do every day, right? Which is temperature check. So that's a sensor. This thermometer, and these are all things we live within our lives. You know, it's even for like our COT sensors in our house. And we do encounter them all the time, but they really just don't have the same requirements that you need for an oceanographic equipment. So one thing that's really easy, like, if your thermometer breaks, you can just go to the drugstore and get a new thermometer. If your ocean thermometer breaks, it's probably a year before you actually have to recover it. And so we take we really do that into our design process. We know that these things aren't replaceable, and they're also actually put in really, really harsh environments as well. Ocean is big and mean and angry and it just breaks stuff. And we know that and so we design a lot of that around that as well. And so, really, what does a good sensor look like? And this is a really you know, you can have this idea of like what your perfect sensor would be. If I was to make the most perfect sensor, it would only react to one physical characteristic of the environment, right? You know, if you're measuring temperature, that's all it does, if you have like for, say, like a salinity sensor, it only responds to salinity. And if you have pressure, it's only responding to pressure. That would be the ideal, right? And then you have a response. So response to that physical characteristic that you can just easily model mathematically, a really simple one is if it's a line, all right? Often it's not a line, though. The reality is, it's just it's not the world isn't like that. Even if no matter what you do with your sensor, it may react to more than one physical characteristic of the environment, I think a really good example of this is that you're we have conductivity sensors, they're sensitive to both the salinity of the ocean and the temperature of the ocean. So we have to take that into consideration. And then also, the response of the sensor might be nonlinear, and or parametric, and they might be really complex. And so and then the terms that reflect its reaction to physical characteristics other than the one of interest. And so we see this a lot in our actual sensors themselves. And we understand this, and we have to figure out how to understand this, and how to make these corrections and how to design the sensor about this. Because reality is, is that once you throw your instrument in the ocean, all of these things are changing, you have pressure, you have temperature, you know, you might be depending on like what you're doing, there's just a lot of different variables. you know, you can have interference, other chemical species, if you're doing chemical analysis, you have to really take this all into consideration when you're designing a sensor. It's always a really interesting problem putting stuff in the ocean, because sometimes you just really, you know, when you're really in the process of designing something new, you always find out new things like it kind of interacts with things that you never thought it would be able to, and then you have to work your way around it and design your way around that. Okay, so I really wanna go through kind of what characteristics define a sensor. And I like to use thermometers as a really easy way to understand this. And how do you define what it is. And so I think a really easy way is just the medical thermometer that you use at your house, 'cause we're all really familiar, I'm actually going with the really old school one that you have to put in your mouth. And the difference between kind of what happens when we make a thermometer, okay. So we have something like a range, right? So the minimum and maximum values where the measurement attains your stated accuracy, so you wanna use something that's appropriate. For a medical thermometer, you're gonna do like 35 to 42 degrees C, right? 'Cause that's typically the range of human body and outside of that, it doesn't matter, 'cause you're probably dead. And then you have a resolution, you don't need a really big resolution, you need a pretty low resolution, but you need to know maybe when you have a fever, or you don't have a fever, it's different for our thermometers. So we actually have a range from minus five to 35 degrees C. And that's really based upon what we see in the ocean. And then we have a resolution. And you know, we're getting down to the fourth decimal place here in temperature. And so it's a really big difference. So ours are really have a high resolution. So these are the smallest changes that something can detect, you actually really need to know this, when you're looking at small changes in the deep ocean, because as you know that those are really small, and they don't really change that much. And then you really need that kind of high accuracy. The other one that you really talk about accuracy with defines a good sensor. So this is just a really great example that I've actually pulled off the internet, you may have seen that. So the difference between accuracy and precision. So accuracy is a measurement that gives you the true value. And the precision is that you can make the repeated measurements to produce the same results. So we really try to do this. You can see up here, if you have what the goal for Sea-Bird is is high precision and high accuracy. But for some applications, you know, you could get away with perhaps high precision or low accuracy, or you know, low precision and high accuracy, depending really what your application is. But that really is, you really wanna take those trade offs depending on what your application is, as well. And the getting to this is a really big part of our calibration processes as well. And I'll talk a little bit more about that later. Okay, so and then the other one is response time. So how long does it take a sensor to respond to a change in the property measurement? I think the medical thermometer and the difference between our thermometer and a medical thermometer is great here. So if you think about it, you know, I'm just gonna tell you that for our thermometers, so this is kinda our fastest responding here. It's in the milliseconds, all right? It's 65 milliseconds, which is super super fast. If you think about your medical response to actually the time if you remember, if you actually ever use one of these, it takes about 10 minutes. Like I don't know if this is really the truth or this is how much my mom actually told me I needed to sit in bed to actually so I wouldn't get out. But that's how long it would take to actually get an accurate temperature. And so that's really the big, that's not gonna work in the ocean. And that's why you have to really know a lot about response times to really understand that if you have things that are changing fast, are you really measuring that? Or is that something, you're still just waiting for the sensor to come back. And then the other one is stability, that's actually that's also really important. It's really understanding if your measurement is changing over time. So you know, if you go back to the same place, and your thermometer is just constantly taking the same temperature, is it always gonna be the same? Or are there changes? The reality is for all sensors, and I will say that all sensors do drift. There is no sensor that doesn't drift, materials age and things change. But the real trick is, is to really get things to drift less than what the signal you're interested in. And that's really the hard part. And the way that we get around that is also for repeat calibration. So we know that our sensors do drift in time. But if you come back for the calibration, you can go back and you can shore it back up. And you know, you're always getting the accurate temperature that you want. Okay, so I'm gonna kind of talk a little bit about the design of just some really basic sensors. Again, I'm starting with temperature, because I think we use temperatures all around us all the time. And I think it's really, really easy to understand. I actually have I forgot to pull them out. But I have actually a bunch of temperature sensors that we actually use for all of our Sea-Bird products, preferred for all our Sea-Bird products. And so really, what we're using is we use a thermistor-based thermometer. So a thermistor is just a semiconductor that changes resistance when temperature changes, and then we measure that we use an you know, the way that we you know, we have to be really careful with this, you can't just put a thermistor in the ocean. Because if you actually squeeze the bead, so you can see it's just like a little bead with two little wires coming out of it and the semiconductor is in there. If you squeeze it, that'll actually change the value. So you actually always want to make sure your thermistor they're pressure protected. And so what we do is we make all these metal pressure protection sheaths on them. And then also, the thing that you also want to do in the design process is you want to make sure that the heat that the thermistor itself is the same temperature as the actual water that you're measuring. And so what we use in that, and so the thermistor is inside this metal sheath, but it also has a thermally conductive compound so that you know that the temperature is much. And so we have all sorts of different sizes, you can see I have down in the bottom right, we can have this little, this picture of all the ones so 3F is our fastest one or 3plus, and you can see it has this tiny little sheath. And that's why it's so fast, because it's just the amount of time that it takes for the heat from the seawater to actually get to that semiconductor. We use a bigger a bigger kind of casing for different kinds of sensors, because they don't need to respond that fast. So like that's SBE 37 is a modern one, you're only taking a temperature every, you know, maybe every 10 minutes or something. And so you don't need something as fast as opposed to with 3F if you're kind of sampling really fast. So 24 times a second instead of once every, you know, 3000 seconds. And really the size of that really helps as well. And the other thing is that you know pretty, one of the things that I really find amazing about this design is that these designs are actually they look like a little simple metal tube, but they're not, they're actually designed so that when they get squeezed and you go down to pressure, it's not actually squeezing the thermistor. And it's not so that thermistor is not sensitive to changes in pressure. So that little tiny little pin, it's really just the size looks like a pin needle, right? There's a thermistor in there. And you can take that to full ocean depth. So that's 10,000 meters, and it will squeeze it and it will not change, it should not have a pressure component. Some have a very tiny one. But it's really designed that it's less than the accuracy of the actual instrument. So it's actually a pretty amazing piece of manufacturing to actually get that in there. It's really simple, but it works. And so that's part of the mechanical design of that. All right, the other one that we do is also we also you know, I'm kind of just being really focused about CTDs, 'cause that really is my area of expertise. And I'll talk a little bit more about the other ones as well. The one that really other people are familiar with is the conductivity sensor and the conductivity sensor you need. And you need that to actually get the salinity of seawater. And so this is actually just pretty much a glass tube. And it just measures the electrical resistance of water. And then using the temperature pressure and a really long equation, you figure out the salinity of seawater. And so this is actually one of the main kind of this is the main technology of what Sea-Bird does. It's this conductivity sensor, it has these three platinized electrodes it has an electric field in there and all the seawater that's in that tube, you can measure the conductivity, and you can get the salinity. And that's actually how you get this really, really, really accurate. And it's really simple. I think that's one of the things about this, it's designed to be simple, it's designed to be a tube. So you don't have a, um...You don't have um... So it makes it easier to kind of figure out these corrections. And really, what you're doing when you're calibrating this is you're actually just calibrating for the volume of that tube. So how much water is in those electric field lines? And it's really, really simple. There's a lot of things that actually affect this sensor itself. It's the fluid flow through the cell actually affects that, which is one of the big reasons why we actually use Sea-Bird uses actually pumped flow. So we actually force water through this using a mechanical pump. And by doing this, you actually change, you can actually reduce the response time and you have a way better response. We also have kind of different sorts of all little sizes. We have about three major sizes of the actual conductivity cell that we use. And you can see that there's different kinds, there's two at the top, they have a black urethane coating. And then the other one, you have a it's a conductivity cell. It's with an epoxy coating. And actually, that is actually the original one that was designed at Sea-Bird. And we still make that for our 911 and 25plus, and so using the epoxy, and the reason why we do that is actually really interesting, it's actually harder to manufacturer. And so we actually haven't changed from that design, because it actually keeps the climatological record for all of our sensors over time. So we've been making these for about 30 years. And so kind of you have a really close comparison, you don't have any major changes from CTDs, that we designed 30 years ago. And to keep that as well. And we really think about that too, we think about how we change our sensors with time and how that really can affect the data that you're using. Particularly relying on long timescales when you're looking at stuff like climatology. And then the other thing to kind of looking at too is, these are really fun. And I really like talking about pressure sensors too that's to know how deep you're in the ocean, we actually use two different kinds, typically, one is called the Digiquartz, and we get that from Paroscientific, it's super accurate, they have to make it there. And then the other one that we get is from we actually don't make these pressure sensors at all, actually, we buy them from other manufacturers, the strain gauge pressure sensor. And so there is a strain gauge is more affordable, it's not quite as accurate, but you can definitely use it more. So the strain gauge chip, it basically just has a semiconductor and it pushes on it, the water pressure pushes on that and then you can actually measure the strain against that chip. The one thing I really like talking about this is because we do actually do our calibrations of strain gauge pressure sensors at Sea-Bird but we're not able to actually do it the one for the Paroscientific. And the reason why is because it's just such high accuracy and because the standards involved. Paroscientific is another company that's out in Bellevue, and they actually have a gravitational map around their entire factory to do that level of pressure sensor. So they it's a totally different ballgame, depending on what kind of pressure sensors but we use this again, this is something we use from another manufacturer, and we're able to integrate, as opposed to the temperature and the conductivity stuff or stuff that we make, we build actually and assemble at Sea-Bird. Okay, and I kind of just want to talk a little bit, you know, and I talked to sort of about the physical effects of other sensors and how we design but we have a lot of other sensors in our portfolio. And so, you know, we do a lot of things like, you know, one thing we understand is right now, you know, we're trying to understand like pressure compensation, pH sensors, so we're continually refining our pH sensor building, and how do you do pressure compensation? Because pressure does affect pH? Is it better to make a mechanical design? Or is it better to do a calibration type design? And what's the best way to do that? Other things, you know, we do work with and try to understand about our sensors is that, you know, if you ever used a SUNA nitrate sensor, it's basically a tiny spectrophotometer and how do you understand what interference by other constituents and how do you work that out to make sure you're actually only measuring what nitrate is and not other instruments, things in the water. And then also, you know, for oxygen sensors if you ever used the membrane type sensor. We do things like the science of it is really understanding why the, one problem we had and there was a paper written about it is why the oxygen sensor changed below 2000 meters. And it turns out that there's a Teflon membrane. And once you get to those kind of pressures, that the crystalline structure of that oxygen sensor actually changes. And so we make a correction for that. And we decided, we figured out what that correction is, and how you make that correction. And that's kind of all the science pieces here. And that's a lot of the stuff that I use, not myself, but I'll introduce part of the other science team, that's kind of the science aspect of that, that's really exciting and really fun to talk about. And then the last thing is, you know, I think one of our biggest things that we like to do is really controlling our environment that the sensors were gonna go in. The ocean is wild, there's lots of changes. And one of the things that we really bring to the table is that we really, I really, I guess I'm just a convert at this point. But really pumped flow is the best way to give you your best data. And the reason why is because you're controlling the environment past your sensor. So a stable environment will give you the best possible calibration. A stable environment gives you the best possible data. But that's not always possible, because you're just in the ocean, right? You're like throwing stuff in and there's waves, there's changes in temperature, and then pumped flow for us, it gives the most stable environment. And therefore the most best possible data, which is why so many of our sensors are pumped, because if you want to get into that high accuracy and not have to make lots of corrections, that's the best way to do it. Okay, so I'm gonna move on, like kind of just the design process and kind of a little bit of an overview of some of the design things we talked about, and that we've done. And the next part of it is really calibration and accuracy. This is a picture of our calibration, or one of our calibration labs in Sea-Bird, we actually have a whole bunch of them. Each one of these green bins has probably about six to eight sensors in there, and they're in there for the night getting calibrated. So a calibration bath is pretty much a extremely stable bath that you put your instrument in, and you try to compare it against a reference. And then you try to understand what the calibration is and you try to figure out okay, these are the coefficients I needed to make to the calibration equation to get one degree C or one PSU or whatever it needs to be. And so all of these are calibrated and computer controlled salt and sea water baths. They typically run overnight. So what happens is that our technicians will put these in the evening and they'll wake up the next morning and they'll come and they'll unload it and they'll do the calibration. Oxygen calibration take a little bit longer, they're about 48 hours. And then also for these calibration coefficients as well. The accuracy specs after we get the calibrations are always for the use of specifications of the instrument, which is typically full ocean depth. And our calibration, our accuracy specs are always based upon what we know following the sensors long term in the lab, and then also looking at data from the field. So we've actually adjusted some of these, depending on you know, how we know things change or like the more information, it doesn't change a lot. But it's always we always take that into consideration when you give accuracy specs and that is what you should be expecting to get in the field from our product. Okay, so really the force I brought this up this is actually I don't know, I like making Photoshop. This is our metrologist Laura, she is the force behind the accuracy. This is something just made up for fun for MayThe4th. That is actually one of her reference thermometers she is carefully wielding, but really metrology so that's the science of standards and measurements. And that really is the force behind the accuracy. Having a really awesome metrology lab is really how we get that accuracy. And I cannot like talk enough about how important this is and how painstaking it is. So this is what our metrology lab looks like. It's not very big. But we have this is kind of where we maintain where we maintain the physical standards used to calibrate the sensors. All these temperature standards and pressure standards are NIST traceable. It takes a lot of upkeep, it takes a lot of long calibration records to do this. You probably see we can see on the left hand side, we have the temperature standard area there. And the right hand side we have the conductivity. And you can even see even on the right right side, we actually have our oxygen standards as well. And so how we do all of this, I'm gonna go kind of a little bit through that as well of what this looks like. But this is we have somebody that's dedicated to keeping this in check and making sure that we're using the right kind of references to give you the calibration and to give you the accuracy you need on your sensors. So for temperature sensors, the standard is kind of it's how you ref... it's what you actually reference it to. And so we actually reference the two primary standards that are based on physical properties of two different substances. So the first is you have a low point. So that's actually the triple point of water. And so that's the triple point of water is where you can actually have water exist, coexist in three phases. So that's a vapor, no sorry, a gas, a liquid and a solid. And that's a really specific point. And we keep that in these quartz crystal, you can see the quartz crystal tube over here, this one is actually doesn't have it in there. But if I actually do tours of the lab, and you can see those, and you can take it out, and you can actually see that it's really, really cool. We've had them certified at NIST, and then their old metrologist actually took a train all the way from Washington, DC, back to Seattle, to get it back to us after it was certified. So, because at the time, he didn't want to put it, actually, because that's when they actually you couldn't actually bring it with you into the airplane. So that's what he had to do spent a week on a train where he decided to do that. And then the other one for the high point is the gallium melt point. So those are kind of our two physical standards. And that's at 29. It's about 29.764 C, and we use that those two melting points to actually calibrate another sensor. So this is called a transfer standard. So what we're doing is we're gonna take two physical standards. So that's the triple point of water and the gallium melt point, use it to calibrate a secondary standard, which is called it's a standard platinum resistance thermometer. And then we're going to use that to calibrate another bunch of reference sensors that are gonna be used in the bath, to actually calibrate what the sensors that you use in the ocean. And so this is a really, it's a process that you actually really need to be have great bookkeeping on. And you really need to know how the standards transfer because as you have the errors, if you transfer from one to the other, you don't want to take that error, you don't want to propagate it. And the SPRT which is here, you can see this isn't actually wise, I couldn't find a picture online. But this is pretty much what it looks like. It has a really it has a well known calibration curve. And you know that if you do these two, if you calibrate it against the gallium melt point and the triple point of water, you can have a great calibration for this, and then you can use it to pass over to the actual sensors in the bath. And that's actually how we do that. And so we keep these standards really, really well. Okay, and then the next one we do is salinity standards. This is a little bit different. In the we don't particularly What we do here is when these are in the bath. So every time we do a calibration, we take a water sample from that bath. And then we have a reference sensors as well. And then we check the count the salinity of the bath against the what, OSIL standard seawater. So that's actually what we buy from OSIL. And using a Guildline salinometer that we've actually so are using two Guildline salinometers. The OSIL standard is a primary standard for seawater conductivity measurements worldwide. And then we compare the bath and the standard seawater, we actually use so much of this OSIL standard seawater that we actually also produce our own seawater, to get a little bit to do that the standard as well, because we could not afford as much as we use. But it works out pretty well. So we're able to kind of track all those standards and know that we're always keeping to the OSIL standards and what they need to be. And then the last thing is for pressure standards, like I was talking about as well. For those digiquartz sensors, we can't actually calibrate them. Because we just don't have that kind of capability, it's a really big capability to actually get to the accuracy. So we actually just use a digiquartz pressure sensor to calibrate a lot of our state strain gauge sensors. And that goes back and they're able to do that calibration as well. We use that for calibrating our strain gauge sensors. And then we also, if you do actually send a digiquartz back to us, we do actually check it against the Sea-Bird standard, so we're able to verify, but we're not able to recalibrate. And then we also have more, I don't have a picture here 'cause I don't know what's allowed anymore, but we actually have also we have more intensive pressure calibration using a temperature control thermotron. And now we use this for the strain gauge sensors for the Argo program 'cause they need a little bit more accuracy for the temperature, particularly where it's cold and deep. All right, so the last one is kind of last piece of this is manufacturing and production. I don't know if you've ever been in our factory, but there's a lot going on. I'd really say it is a factory it's a factory with highly skilled technicians. And I feel and I really want to say that our manufacturing capabilities are really tied to the process and the skills and the people that are making them so we have people that are really good at what they do. And they're able to do that. This is just a picture of our service instrument storage racks. This is actually one of my favorite places in Sea-Bird. Because we pretty much get like every single instrument you can possibly think of. We get 30 year old instruments, we get, you know, we have brand new ones. And it's, you can see really, it's kind of like, this history of just how things are changed. It's always really useful for me to just go back and be able to see how things have changed and have designs have changed over the years and as well. But really, there's kind of four teams that work together and make a sensor, instrument or any system. So the first, you know, is the science piece, like what's really the basic functionality of that sensor? How does the science behind it work? And then we have R&D engineering, so they help us to package it, they help us to make the electronics, they help us to make the mechanical design to make a sensor. That's something we can throw at the ocean. And then the next one is production engineering. So these are the people that kind of go through and they say, Okay, this is what are needed, how are we going to make the jigs and everything how to build this, how are we gonna be able to make a lot of these the same way every time. And I think the last really important piece of this is supply chain as well, they actually find the parts for us that need to be done. So this is always really, really tricky. It particularly, you know, just finding the right parts to make really highly accurate sensors, you can't just use any some any kind of chip off the shelf, you have to use something that's stable, and you have to use something that's reliable, and it has to fall within a lot of tolerances that have been brought down from R&D, engineering and science. Okay, you know, we talk about production. But I think one thing for me is it's really I want to talk a little bit is about the scales of production. You know, I talked about manufacturing, but you have to really think about like, where does Sea-Bird kind of fit in, in the world of production, and I thought there was some pretty good common themes. And so really, you know, I think we can go from the top left to the bottom right, you can go from sort of scales from like big to small. So one thing is like that, we kind of know that we've been really ramping up production right now, is surgical masks. The amount, I think it's 110 million produced every day, which is an unbelievable amount. And the reason why is because you have machine assembly, machine assembly, so the machine kind of basically assembles most of it with a little bit of help of human operators, and then each of the machine only actually makes one model. So it's really fast. That's why you can actually produce so many. Kind of the next step down, I think, in a little bit more complexity, but you still make a lot of them is an iPhone. There's 15 million built annually, all right, so that's a lot less. I looked on the site, you have about five models. And each one, it actually is assembled by a human. This is actually really I didn't realize this, but there, Apple cannot actually move away from human assembly, it is just too complex. But they still are able to you can just see how many people are working there actually building these. The next one is a you know, car, I think a car is also a really complex system. So a Toyota Corolla that's your most popular car that's made in the world. So they make about 200,000, they only have seven models, but it's kind of a mix of human and machine assembly as well. And then I'm gonna go through the low end. And so the low end is instrumentation built as a research institution. So this is something that, you know, you're kind of making a one off, right? You maybe make one to 40 built annually, I'm saying maybe there's some people that make more typically, when you're building this infinite iterations, and you have a lot of human assembly. So Sea-Bird's kind of in the middle of this in terms of production, we probably build about 1000s of sensors, annually, we have 150 models, it's actually a lot of complexity involved. And we have human assembly. So you know, I think a lot of our production is sure we make a lot compared to what you can do at a research institution. But we don't make as many as a typical production. And so that really, that's what it kind of puts us in an interesting space. Because you know, to make changes, you can't always, you know, you have to be really mindful of your changes, and also mindful of your changes 'cause things can't break. You know, if your car breaks, your iPhone breaks, or your surgical mask breaks, you can get a new one. But that actually doesn't work with the way with Sea-Bird scientifically, it breaks it's really, you know, it can really have effect on people's work and people's research. And so we really understand that. So we try to make everything really, really durable. And so you don't have that. So we have a lot of additional design constraints. So it's a really interesting space to be in, because we're kind of pulling from both of this R&D world, but also from this production world. So really what defines manufacturing at Sea-Bird? So this is actually just a picture of one of our little cells, where we actually make one kind of instruments. So this is a MicroCAT. So all the MicroCATs that we build come out of this one area, it's a desk. You can see on the left, there's these little bins and bins have each of each of these bins is labeled with a part. And you actually the bins are arranged in the way that you build them. So you just kind of go from left to right and then down as you're building them. And this is part of production, right? This is to make sure you do everything the right way, and to make sure you're doing it as fast as possible. Really what defines manufacturing is that we really are supercharging parts we buy from ordinary suppliers. We make sure we pick and choose our parts. But we're really doing a lot of the calibration work really does that. We have a really high level of quality control. We do a lot of extensive electrical testing, we have a whole e-test area to ensure stability of our sensors. We also have a really well defined manufacturing process. And one thing is we also have adopted what's called the Danaher business system. And I'm not going to go over into this too much. But it really is based on the Toyota Production System like really understanding you know, how to make how to produce things in a really regular way. And make sure that you can identify like if you have a problem or production really how to solve that quickly. And so you have a high quality and reliable product, and service and calibration also follow this model. And then the kind of last thing I want to talk about too, as well as one of the things I really love about working at Sea-Bird is that I'm actually part of a really big science team. There are eight of us on the team right now. So we all kind of... I'm really the area of expertise is CTDs but we actually have a lot of people with huge scientific backgrounds, really doing a lot of these problems. And it's really, really great. We have a huge it's a really great... I really love everyone that I work with, we always have really, we have good conversations and we have good problem solving and we work together. And so that's the kind of the last piece of the science, Sea-Bird science. So really what I want oops, oh, no, I missed, I lost the slide. All right, so my last slide that I managed to get deleted, which I'm really bummed that I lost. But really just the key takeaways is that, you know, manufacturing is, you know, really Sea-Bird is really a combination of science manufacturing calibration. I really do. I really, actually, really, and I can't stress this enough, I actually really enjoy working there because from scientifically it's, there's always really hard...

Kim, you just went mute. Oh, my goodness. Sorry, didn't that pick up on anybody else? Or was it just me? Nope. Wait. Now I'm getting your feedback from before. Darn it.

Hi, can you hear me again?

I can. I might interrupt you again. 'Cause it sounds like it's kind of going back and forth with that. Very loud noise from before. But...

Oh, is it? I'm sorry.

It seems okay. We'll right it up. We'll see. Okay, I'm so sorry. But we probably missed like the last 30 seconds of your close. [Kim chuckles]

It's okay. I lost the slide somewhere, too. So I'm not entirely sure what happened. [Keyboard clicking]

But you sound okay. I mean, we've got about eight minutes left before 11.

Okay. Close up and we can do Q&A.

Yeah, I do actually see one question from Ned. Yeah, does Sea-Bird use real or artificial seawater for its salinity calibrations? Yeah, we actually use both for our most accurate sensors, which is what goes on the shipboard CTD. So that's the three. And the four, we actually we do use seawater. And we actually have had been in a partnership with NOAA to actually get the seawater. But now we actually have a bigger source as well from a Hawaiian facility. For our MicroCATs that have lower accuracy specs, we use just artificial seawater. So we just, we mix that up from bags of known salt, and we're able to dilute that to our specification.

I'm just gonna join you here at the end. Thank you, Kim. I'm sorry to end... I don't know what happened, why you went mute. But I'm giving the round of applause from everybody, even though you already answered the next question. And thank you, it's so fascinating. I mean, like I get to work with the engineers often in obviously, not even in intensive capacity, but just always learning all the intricacies that go behind every kind of instrument and design is so much more in depth than like you're trained to know, right? It's just not your expertise and like as you get to it. So I love hearing you talk about all the different pieces on the equipment and kind of what goes into that. Ned says Kim, thank you very much very informative. Ned I feel like this was speaking to your heartstrings today. [Kim laughs] Yeah, everybody says thank you. Any other questions from folks, please drop them in the chat box so we can ask Kim, Noah says great talk, way to carry on through tech difficulties. I think total pro that she is. Yeah. Okay, there's no questions yet. So Ned here's your chance keep going. I said Noah was on... And oh Kyle, Kyle says can you mention your new O2 system?

Our new O2 system? Well, we are working with the Argo community to make, it's not new, it's a different flavor of O2 sensor that's going to go on BGC-Argo Floats. So this is work that we've just been doing in collaboration with BGC Argo. And it is part of a grant to actually just... We do make an O2 sensor, but this is going to be an in air calibratable one, just a little bit of a different flavor.

And Noah is here, Noah asked, how do we make these sensors more affordable? So we can buy/deploy them in larger numbers.

It's really hard. You know, I think we always try to make them more affordable. You know, I think there's a lot of things we do try to make them as affordable as possible. But like I said, you know, we have really well trained technicians. And there are scales, but typically the scales of, like I was saying, you know, the scale of actually of building them, they're not huge scales, you know, we're building like 100 to 1000. And if you really think about, typical manufacturing, when you say like bulk parts, it ends up going from like maybe like 1000 to 10,000. And that's when you start kind of getting that bulk pricing. And it's really hard to design a lot of these sensors for the ocean. But I'm gonna say, so there are, you know, there are people making sensors, we make really highly accurate sensors that can get the accuracy, really high accuracy, to really deep depths and in really extreme conditions. And so that's kind of what we do. But there are, you know, if you don't need as high tolerances, there are other sensors out there, and people are making them as well. And so there are possibilities for that. But there are trade offs. I mean, there really are, if you want a cheaper instrument, you either have to make a lot of them to really scale up the production, or you have to really take some hits in whether the quality or where you can use it. And so that actually brings a lot of infinite variation. So there is there is room for that. So yes, it can happen. But it really needs kind of a different you have to really understand what you need for that kind of stuff.

So let's say lots of questions now. So Al asked does Sea-Bird deal with micronutrient iron sensors?

No.

Okay, Ned, you showed the MicroCAT desk. How long does it take to make one of these?

Gosh, you know, I don't know. I'd actually have to ask downstairs, I'm not really sure I know we make a lot so but I'm not gonna say I, So Ned there's actually a lot of steps in the production process. So that's just the assembly. So one of the first things that you don't see that so that's mechanical assembly, but there's actually a whole another area. So what you see at that desk is like the pressure case, and then the boards. And before the boards actually ever go into that MicroCAT they go to e-test and they go in electrical task. And so they're actually tested for all of their electrical components before they go in there. So there's a lot of different pieces to that. So they'll go through e-test, then they'll get put together, then they're actually go through calibration to check if they're okay. And then finally, actually they're going to get a pressure test. So there's a lot of and then you get a final test at the end. So they actually go into another bath where we have final test technicians that actually test whether they actually do sort of it's not for calibration, but it's just to make sure like if it turns on or off really basic stuff before we ship it to you and everything is integrated together. There's a lot of steps.

Yeah. I think most things have a lot of steps when you sit down and you break it down. Okay, Ned says good explanation for a complicated process. Thank you. [Kim chuckles] Lisa asked, oh, no, hold on. Okay, do the Sea-Bird scientists/engineers go to sea to see instrument performance in the field?

Yes, we do. So we just actually had we've been a little bit late up this year. We actually typically do a lot of testing off of Hawaii. It's really easy to get stuff in the water and really deep, really nice, really easily. So it's a really good testing bed. But this year has been kind of a not. As with all of you, we have travel restrictions as well. So we've been dealing with that. We actually do test cruises, I was actually supposed to go to New Zealand in March, but I'm not going anymore, but other people are going to do instrument testing. And we also actually do instrument testing with partners in the field so often we'll work with people like the GO-SHIP program to put sensors on that particularly when you have salinity samples to actually do that. So we actually have a lot of different avenues. You don't have a.. Actually no we do you have a pontoon boat of science I found out but that's really just for lake testing. That's not for deep ocean testing. That's sweet. They use that in full width for some things, but yes, we do, we do send out and also, you know, if you have like local cruises and you need people we actually do like to encourage like our technicians and stuff 'cause a lot of our technicians aren't our technicians. And so if there's opportunities for them to go out for like a one day cruise or something and you want somebody along if you get a room, we'd love to send them out with you too.

Noted, all right.

Dave Butterfield here and he asked does Sea-Bird have a sensor for hydrogen sulfide?

I don't believe so. But I believe hydrogen sulfide does interfere with some of our, from some of our sensors, but for off the top of my head. I don't remember which one.

Let's see. And so I can take probably one more question after Jens. And then we'll be, we are at 11 o'clock right now. So Jens asks, O2 sensors, many applications are now on floats, or profiling crawlers, etc. The sensors are not stationary but move. Do you test sensor performance on moving objects, particularly for understanding sensor delays?

Y-y-yes. But that's kind of a trick question. So we pump we love to pump our oxygen sensors, because then it doesn't really matter how fast your oxygen sensor your platform is moving, that takes a lot of the ambiguity out of having different flow rates, because the response time so your response time is always dependent upon the flow rate over the sensor itself. Because if you have that flow rate, it changes the boundary layer. And that changes how fast oxygen can get there. Which is why we love pump sensors, because that really controls a lot of it. Pump sensors I know people don't use them, and it makes a lot harder. So when we're actually doing that design work, we think about that as well. And our yeah, and so we always are trying to figure out how to control flow in that case.

Heard Ned talk a lot about that with Saildrone. Yep, he answered...

It's hard.

Yeah, things I never knew, right? Any other questions? It's 11. So if people need to sign off, please feel free. Thank you for joining us. But I can take a couple more questions if people would like. Kim, thank you, again, for doing this. It's so fun to have you back here with us. Even though I know you're not far away, but it's nice to bring you back and talk with the group. So I'm sorry that it couldn't be in person.

It was a fun talk to give. I mean--

It's been it's been fun hearing, you know, seeing you at the conference and nerding out on all the new years. And it's truly just like, your perfect place. Like I feel fun. Yeah, and for those of you, we did record this talk today, it may or may not be up we'll decide shortly, I guess. But all of our other talks have been recorded. You can find them on the PMEL YouTube page. We do have a separate link now for EcoFOCI seminars. And we have one more talk next week. So make sure to join us for the last talk of the season. And then we won't see you until next spring. So any other closing remarks, Kim that you'd like to give today before I close out?

Well, it's kind of good seeing everyone virtually. [Kim laughs] But thank you everyone for coming by. If you do have questions, please feel free to ask and email me I mean I'm pretty they try to open them. I try to answer as soon as possible. But everyone kind of asked me really complicated questions. So sometimes takes me a while I never get an easy question.

You like the complicated questions though. I know you like questions. [Kim laughs] Always looking for a challenge. [Kim chuckling] All right, everybody. Thank you so much for joining us today.