Episode #2

October 2020

​

Hi, everybody. Thanks for being here. And welcome to ocean solutions, a NOISE Lab podcast. I'm Dr. Morgan Reed Raven, biogeochemist and professor at the University of California in Santa Barbara. In this podcast, we're talking with inspirational individuals who are working on some of the largest issues of our time. Problems that threaten our human civilization and living things across our planet, that exist at the intersection of climate, ocean conservation, and human wellbeing. Last episode, we talked with Gavin MacDonald, who told us about some incredibly important work he's been doing alongside global fishing watch and others to quantify and understand abusive labor practices on deep sea fishing vessels.

Today's guest is Dr. Isabel Houghton, data scientist with Sofar Ocean Technologies based in San Francisco, California. Here, we're continuing on our theme of data science. This is huge right now as our ability to collect and process and maybe even understand giant datasets is growing exponentially. In this case, our guests is a data scientist at a company that is doing the engineering and the design of autonomous open ocean sensors that can build some fundamental data sets about ocean waves and water conditions with all sorts of important applications.

These devices can guide ships. They can track winds inside hurricanes and they can provide real time alerts to coral bleaching events. And they're small and cute and yellow and much cheaper than anything else that currently exists. All of which I approve of. But there's more, they also build drone submarines that can explore dangerous wrecks and caves while you fly them around from the surface.

And on top of that, we also get to hear about life in a fast paced San Francisco startup company. That's trying to make money while providing free data to researchers at the same time. Let's do it.

 

Morgan: [00:02:33] hi, Isabelle.

Isabel_Houghton: [00:02:35] Hi, Morgan. Thanks for having me.

Morgan: [00:02:37] So Isabelle, I am really excited to learn more about the technologies that your company has been building for ocean data collection. I hear the words, drone submarines, and I get this uncontrollable urge to play with one. Of course your devices are actually intended for greater things than just being super awesome.

So to get us started, could you just tell us a little bit about the big issue or issues that motivate your work?

Isabel_Houghton: [00:03:05] Absolutely.   one of the leading issues really is this idea of under sampling of the ocean.  one of the,  leaders in ocean science, Walter monk  once said that. the previous century really has just been a century of sampling of the ocean. And so that's really the issue trying to address at Sofar, whether it's with these drone submarines or with the more recent wave buoys that we're developing.

And what we're trying to do is essentially get measurements or videos or,  time series of what is going on in the ocean.

So one of the things that we measure is waves . So whenever there's a storm, the winds will turn up massive ocean waves, much like you would see at the. The beach, but out in the open ocean and because the oceans are so massive and we don't have many instruments out there, it's a pretty rugged place. We just don't have an, idea of how large the waves are and how they move around the ocean in time.

 

Morgan: [00:04:04] Got it. And so this under sampling I think is really fascinating and it seems really important on a lot of levels. We think about what we do and do not understand about the ocean and how it works today. It's really behind all of oceanography, right? Is this lack of complete datasets.  so for example, with waves, how do people normally measure waves?

Isabel_Houghton: [00:04:29] Yeah, so  it's fairly challenging to measure waves. Uh, we have  buoys that are typically,  Maintained and deployed by government institutions.  it could be tens of meters, tall of a booth and cost tens to hundreds of thousands of dollars.

And that we'll be able to measure the wave height in a single location. So we have a number of these, deployed along the coastlines of.  relevant areas where it's really, really important to know the way of height at a given time, but on the open ocean, a way from coastlines, we have very little information from buoys.

 So satellites,  there's methods with satellites to measure wave height as well. But once again, you have to have a satellite passing over the region of interest. And there just aren't, there's a lot of satellites out there, but there aren't enough to actually be measuring waves across the entire ocean all at once.

Morgan: [00:05:25] And I suppose a buoy is expensive, but a satellite is even more expensive.

Isabel_Houghton: [00:05:30] absolutely. Yeah. So massive campaigns to gather the satellite data. And right now state-of-the-art, you know, really relies on satellites as providing truth to, to models for all sorts of variables of the ocean.

Morgan: [00:05:45] Got it. And so the satellite is a new piece, but have these buoys then with us for a really long time, it seems like  just on a boat waves matter, right? Is fairly intuitive to me. And it seems like sailors would have wanted this kind of information for hundreds.

If not thousands of years, how did we do it way back in the day, understand waves.

Isabel_Houghton: [00:06:07] That's a great question. And I don't think that we did, I think in the last century has really been a massive amount of progress in trying to understand waves. And a lot of inspiration actually at so far comes from  some previous projects specifically, we called waves across the Pacific, and it's actually a fairly wild product,  where they were trying to understand how waves move around the ocean.

And no one had ever actually gone out and measured this. But this was the early 19 hundreds. And so I  they built a ship that could, turn 90 degrees and become vertical in the water. And that allowed it to measure waves at the exact location where it was at. So it became its kind of own portable,  wave buoy.

 that was how the design was made.

Okay.

Okay. So I had to look this up waves across the Pacific was an experiment in 1957, that demonstrated that waves could travel the entire breadth of the Pacific ocean. They had stations from New Zealand to the South Pacific islands to Hawaii and Alaska. And then in the middle of the North Pacific, where there are no islands, they use this incredibly crazy ship that yes does. In fact, tip on its end 90 degrees.

Any generally not advised ship position. Which puts part of it so deep that it doesn't move around and waves it's called the Navy's flip ship. And the last I knew it was hanging out at the Scripps institution of oceanography pier in San Diego. I had the chance to go inside once. And it is like a submarine in there.

 it's tiny and a little bit creepy and everything can turn sideways  in the tiny bathrooms or I should say heads, there's like two sinks, one on the wall and one on the ceiling for when it flips. Which seems like it would be an absolutely terrifying thing to do in the middle of the North Pacific. But anyway, hooray for crazy ideas in ocean technology. Back to Isabelle. 

Isabel_Houghton: [00:08:03] and there's a pretty wild movie out there of these scientists , trying to go out and measure these waves with such little amount of technology available and incredibly challenging logistics. But because they didn't have, you know, a small buoy that they could throw off of a ship or out of a plane and, and that could measure, the waves. They had to have a massive ship, . And that actually. Was pretty fundamental work in the field of understanding waves, but at the same time  was incredibly expensive and actually obtained a very small amount of data.

Morgan: [00:08:38] Yeah. Wow. That seems like you said, that sounds both incredibly challenging. And like, at the end of the day, you still have one – one ship that turns on its end in a terrifying direction. Um, but you have one data point there and you're trying to understand something that is stretching as a process from Tahiti to Alaska.

That is an enormously difficult challenge. Okay. So you guys have for dealing with this, but before we talk about its details, I would just like to back up a little and talk about why it matters to understand waves.

Isabel_Houghton: [00:09:12] Yeah. So  someone on a boat definitely cares about what is happening with the waves and global shipping is actually a massive industry. 

 And if you're able to accurately measure where the waves are, then you can choose essentially the most optimal route,  much like taking Google maps and you pick the fastest road rather than necessarily the shortest road.  you take this move to rather than the shortest path.

Morgan: [00:09:39] our ships and transport companies, then the main thing that you think is the application for the sensor? Are they your customer? So to say at this company,

Isabel_Houghton: [00:09:49] Yeah. So there's kind of several levels of how, sampling the ocean can be beneficial to,  all sorts of interested stakeholders. So to the scientific community, they want the data, they want the raw data because they have lots of questions as well beyond just internally at our company. But we also have a lot of stakeholders who are really just interested in.

, you know, a shipping company, I don't really care about the raw data they care about. Okay. Tell us which way to go.  should I head North right now and how quickly should I go? And so. How we're using the data is kind of at several different levels. We'll, we'll share the data itself. if researchers are interested, but we'll also build essentially intelligence products on top of it, because we fully understand what we're measuring,

and then we're able to provide,  solutions that we believe in,  to these ship running companies.

Morgan: [00:10:42] Very cool. Okay. So let's actually define what this thing is that you built. What does it look like? How big is it? Is it really crazy expensive or it sounds like much less than crazy expensive.  and what exactly does it do?

Isabel_Houghton: [00:11:01] So our product is essentially something called a spotter buoy, and it is very low cost, which is.  trying to address this issue of under sampling. You can't sample,  in mass, if you have to spend $100,000 to buy a single instrument. So it's supposed to be less than about $5,000 for a single buoy, which is, you know, an order of magnitude decrease from typical wave buoys.

And it's like less than half a meter in diameter. So little over a foot in diameter, . It's the size of maybe a beach ball. And so you could walk around with that beach ball and you can throw it off the edge of a ship out into the open ocean and it will float there and it will measure wave Heights and it will, connect to a satellite and ping back, essentially all of the information it's collecting.

So wave Heights, along with sea surface temperature, and hopefully soon. Barometric pressure. And you can log into your computer and see wherever in the world that buoy is. You can see what the weather is like there.

Morgan: [00:12:08] Wait real time.

Isabel_Houghton: [00:12:10] Definitely. Yep.

Morgan: [00:12:11] Wow. That's a whole different level because now you have something floating around. That's actually connected to a satellite potentially in a really remote part of the ocean. Maybe even like the Southern ocean or someplace where I can't imagine you have much data at all. and you can see it from the comfort of your living room.

Isabel_Houghton: [00:12:31] Exactly. And we're at the point where we have several hundred of these buoys out in the open ocean. And so you can look at this dashboard essentially, like you said, from the comfort of your living room or from your computer is you're trying to understand something and see data from several hundred buoys in real time.

  You can start to really look at things as they're happening. So if you see a massive storm churning up in the Southern ocean, you can say, Oh wow, I see there's a buoy right there. What's about to happen there. we're also trying to pursue some more scientific the questions.

And so one idea is dropping them in front of hurricanes, um, getting wave measurements. Around hurricanes is really challenging because they're very small, very confined, although they are massive impacts. And so rather than just trying to have a buoy in the precise, lucky location that the hurricane hits it, knowing that we can compact spiritual buoys, that we can actually throw them from pretty great Heights into the ocean that we could actually intentionally seed a bunch of.

Instruments right in front of a massive event and get a ton of information that you just otherwise wouldn't have access to.

Morgan: [00:13:46] That is so cool. And you can throw them. So you're like in a helicopter whooshing by, in the front of a hurricane throwing out buoys. Is that, am I picturing the right thing here?

Isabel_Houghton: [00:13:55] Yup. So I think , maybe last year they did some, we did some tests, using a , so a small plane and threw some buoys out over Monterey Bay and they were able to land in the ocean and start immediately, you know, recording data, sending it to the satellite and you're up and running from there.

Morgan: [00:14:15] Wow. Well, and I imagine their low cost is helpful in this too, because I imagine it's hard on a buoy to go through a hurricane.

Isabel_Houghton: [00:14:25] Exactly. And even in hardware development, uh, if you're developing a buoy, that's hundreds of thousands of dollars, you're not going to go out and drop test it. you're not going to, you know, come up with these wild scenarios where, there's a 50, 50 chance at survives, but if you get that data, it'd be really valuable.

So that's another kind of really exciting. A line of expansion, as you mentioned, the Southern ocean is incredibly under sampled, but it's also a very severe location for instruments. Hence it being under sampled. But if you have buoys where you can, have a survival rate of a couple of them out there, then you're, you can start getting data, whereas you just wouldn't have any.

Morgan: [00:15:06] That is so cool. so the buoys that already exist in the ocean, where are they and who put them there? Do they belong to so far or are they working for specific researchers? How does that work?

Isabel_Houghton: [00:15:18] So it's a whole mix. we have a variety of programs as we try and see the ocean. And there's a number of, of privately owned because they are low cost. There's a lot of interests. In using them for research or any sort of application, whether it's, you know, monitoring wave height outside of  your Cove or outside of your port, or, we have a broader effort to see the Pacific right now and ultimately next year, see the entire ocean.

And so those are less exclusively owned by  so far.  if you. Are trying to conduct research and want to use that data. It's free for access for research, but otherwise the global grid is intended to be kind of our data that we're able to build products on top of.  and as a result are able to fund this project,  sampling the ocean.

Morgan: [00:16:11] Wow. How many buoys do you need to see the Pacific?

Isabel_Houghton: [00:16:16] So we started in the Northern Pacific and I believe we had about a hundred by, mid 2020. And that was enough to really start seeing markable improvements and our.  forecasting models.

Morgan: [00:16:31] Nice.  You were talking about applications and something that jumps to mind is that these are really as accessible as they seem. It seems like you would put one outside of surf zones and that, that would be really useful information actually.

Isabel_Houghton: [00:16:46] Yep. Absolutely. They have all sorts of coastal applications as well as open ocean

Morgan: [00:16:51] very cool. So something we haven't talked about with these buoys that I saw when I was  on the internet, was that you can actually monitor not just the surface, but sometimes things that are under the surface too. And that seems like a really potentially powerful application.

Isabel_Houghton: [00:17:08] Definitely.   the surface and the subsurface are very much connected and if you're able to understand. Both what's happening at the surface, as well as the subsurface, you can really start to describe a lot of the ocean,  and where we're trying to measure with, especially these subsurface measurements is near coral reefs.

And as you'll probably have read things like coral bleaching is. Increasing, and that has to do a lot with ocean temperature. So if you're able to measure temperature, especially at depth and at the surface, you can really start to understand what processes are happening and if you can really monitor it in real time and have longterm data.

So you can start to see when things are changing, you can  try capture the process that's happening and potentially start to think about interventions.

Morgan: [00:17:58] Very cool. so this device we've been talking about, you called it spotter.

Is spotter the only thing you guys make, or can I hear more about this underwater drone?

Isabel_Houghton: [00:18:11] Definitely  Yes, I could definitely talk about the, the underwater drone. And that's a really interesting part yeah. Of this kind of low cost fleet of devices that we have where we're trying to address under sampling and sampling can be taking a photo of something.

It can be measuring the temperature at an exact location. And so while the spotter is. Measuring wave height over time or temperature at a specific location. This underwater drone is going down and taking video of things. And so it's a whole different type of information, but at the same time, it's a low cost way to sample the ocean

Morgan: [00:18:47] Just me too, this drone. What does it, what does it look like? How big is it? What does it do? Why is it cool and how do I get to play with one?

Isabel_Houghton: [00:18:56] So the,  Trident is again, , another small device, so less than half a meter, easily carried looks like a really aerodynamic underwater drone, essentially. And it has a high resolution camera on it. And it has a tether because communicating through waters is really challenging. But with that tether, you can essentially drive it much like a remote control vehicle or a video game essentially.

And you can get a live feed. It looks like underwater, wherever you are driving this, this Trident, which. You know, you could maybe do the same thing if you were scuba diving, but you probably wouldn't want to go as deep for as long or into as risky of locations as this low cost, a little underwater drone.

Morgan: [00:19:46] Cool. So what are some examples of places that have been explored with this drone? It seems like you mentioned like dangerous places. I was talking about like shipwrecks and caves.

Isabel_Houghton: [00:19:57] Yeah, exactly. And I think that was actually one of the motivating or the inspirational explorations that they wanted to do when the Trident was first created was to go into a cave, which is not necessarily fun if you're scuba diving, but if you have a small underwater drone with a light and a camera on it can lead to some pretty incredible exploration.

So exactly it can go into caves. It could go into a shipwreck and explore all sorts of nooks and crannies and other things that just wouldn't be accessible to a scuba diver. And it can also just go deeper where your air wouldn't last for very long. If you wanted to scuba dive

Morgan: [00:20:36] yeah.  Can you tell us just a little bit more about how Trident was invented and how this turned into a company and the came, something that now makes ocean buoys that talk to satellites.

Isabel_Houghton: [00:20:51] So I think it really started as a project that was intended to be.  opening accessibility to ocean exploration. And so they created this incredible device, found that both scientists and recreationalists loved using it.  they had their following and, and that grew, and then actually separately, wave scientists also developed this.

Spotter buoy. And it kind of had the same driving ideas of make it low cost, make it modular and make it accessible for people to use. And so both Trident and spotter I think are fulfilled by that kind of checklist. And so about a year and a half ago, the two companies merged together to form one, which is now called so far and ever since then, Everything has kind of come together and people who are experts at  developing new hardware, whether it's an underwater, drone or a buoy, are all working on all of these projects together, with kind of all of the same, motivations and love of the ocean at the end of the day.

Morgan: [00:21:58] So what happens at this company then?  What is your, what is your job? What do you do?

Isabel_Houghton: [00:22:05] Yeah. So I'm a data scientist by title. and my background is, is ocean science. So it's kind of this perfect combination of an understanding of the physics and the processes in the ocean, and also an ability to analyze data. And we're getting all sorts of data off of all of the sensors that we're deploying.

And so we need people to look at that data and. Create these products on top of the data. What is the data actually telling us?  is it good data?  are we measuring what we think we're measuring? Are we able to validate that? And then how do we actually, incorporate that into useful products? So  what I do is is a lot of analysis, which is  a combination of coding, and then  problem solving  but there's. A number of other people at the company who are building hardware, who decided that it would be incredible if we could measure subsurface temperatures.

And so they spent months coming up with the proper material to connect a temperature sensor, to a surface buoy that was both low cost and durable and easy to obtain and deploy.

Morgan: [00:23:19]  in one of your normal days, what does it look like? How do you spend your time?

Isabel_Houghton: [00:23:24] , I would say it's a whole mix of things. And I think that's a product of working at a startup, which is what we're really technically still are and will be for a while, as we're trying to. grow in this industry and because we're a startup, that means we're pretty, pretty scrappy in what we do.

Morgan: [00:23:40] What is it like to work at a startup company? Right. Okay. First of all, how big are you? How sturdy are you right now?

Isabel_Houghton: [00:23:47] Right now we're about 30 people and we are kind of growing every week. So I would say six months ago we were maybe 20 people.

Morgan: [00:23:56] Wow. Okay, cool. So we got jobs. Yes.

Isabel_Houghton: [00:24:01] definitely. There are jobs in ocean science.

Morgan: [00:24:05] great news. Um, okay, so you are, you're not so small, you're like 30 people, but it's growing really fast is what I'm hearing. So, what is it like to be at a startup as opposed to say being an academia?  you did a PhD before this. We haven't talked at all about that yet.  how does it, does this compare to being in an academic setting?

Isabel_Houghton: [00:24:27] It's definitely different, but I think there's some aspects that remain the same . It varies very much based on, you know, what company, what startup you're at, but because we are really interested in the ocean, it's a lot of people who also. No care about the science and the research behind it.

And so I came from academia, I did a PhD and all you're thinking about are the research questions. What's something interesting we can ask. And I think  one of the biggest switches to a startup is, okay, how do we solve this problem?  we really want to be able to gather all this data, how do we make that feasible?

Both economically in terms of hardware, in terms of people and resources.

 

Morgan: [00:25:07] Fantastic. I'd love to hear a little bit more about your background too. Now that we've brought this up.

 did you grow up around the ocean where you, why did you originally get interested in ocean and atmosphere science?

Isabel_Houghton: [00:25:22] So I grew up in Northern California, actually pretty far from the ocean up in the mountains. And I can't point exactly to, you know, how I got into ocean and atmospheric science other than I like being outside a lot. And the ocean's a pretty incredible place to be outside near.  So I think the ocean has always been really intriguing. And I would say it's definitely having a, a bit of a moment right now thinking about the open ocean and the deep ocean and all the crazy creatures down there being involved in that I think has just always been exciting. And that's why I've kind of stayed down this path.

So I actually started out in atmospheric science, which sounds different, but at the end of the day, in my mind, very similar, my focus has always been fluid dynamics, which is essentially how does water or air move? So in atmospheric science, you're thinking about air in oceanography, you're thinking about water.

 at the end of the day, the, the equations are the same.  And then as I moved into my PhD, really focused on physics in the ocean and, researching questions in experimental fluid dynamics.

Morgan: [00:26:29] So what made you decide to work for a startup company? 

Isabel_Houghton: [00:26:32] when I finished my PhD, there's a lot of interesting routes. You can go   I knew I definitely wanted to stay in ocean and science, but I wasn't entirely sure in what environment  but I was really interested also in kind of the, the shorter runway, the really immediate applications of industry. And I would say, especially at a startup where you're. You're thriving, but at the same time, very much trying to stay afloat also.

the timelines are really short. And so, whereas a PhD is on the order of years to really start to see the outcomes.  In industry, you can very much see something that you developed six months ago, be put on board, a small three that's then thrown out into the ocean and you know what, the next day it's sending data back to you.

So I think being able to work on a, a large team of people who are all really excited about the exact same thing and seeing all of that work move at such a really rapid pace is a really exciting thing about working in industry.

Morgan: [00:27:39] Yeah. Well, and it sounds like you guys have a lot of things, exciting momentum too, with. Even just the fact that you were founded a year and a half ago, and you have these new devices and you're hiring new people and you're expanding it. You're seeding the Pacific with your buoys. That's all really exciting stuff.

Isabel_Houghton: [00:27:56] Yeah, definitely. And it, it really is. It, it is for me personally, I would say to at the end of my PhD, I wasn't really sure if there was a place in industry for science and ocean science necessarily. And it has become pretty clear to me that there really is. A rapidly growing kind of ocean technology and ocean solutions industry that isn't just government funded, but can  keep the lights on in a variety of different ways while still addressing issues in ocean science.

Morgan: [00:28:25] Absolutely. So where do you see this going in the next, I don't know, pick your number of years, five to 10 years. Where do you see your fields in ocean technology in general going or so far more specifically?

Isabel_Houghton: [00:28:40] I think we're still in the under sampling phase. So I think where we're going in the next maybe five years is really incredible development of. Sensors and the ability to actually collect all that data and do something meaningful with it. So, and so maybe the do something meaningful with it is, you know, five to 10 years also because it's one thing to collect a ton of data.

And we're really at kind of what you mentioned of a huge time for data science. Because we have so much data, there's also a responsibility to actually find something of value in it. It doesn't matter that you have, you know, gigabytes of terabytes of data spewing out of sensors. You need to start finding patterns in that data and then be able to make actual.

Logical decisions based on that data. And so maybe also in the five to 10 years is interventions,  based on what we're able to collect. So specifically for something like the,  the smart Moring projects where we're collecting the,  temperature or at depth as well as at the surface, if we're able to start to.

Intervene in situations where we know that bleaching is happening or we have incredibly at risk ecological regions.  is there something that we can do now that we're able to measure and diagnose what's happening?

Morgan: [00:30:03] so almost a rapid response for stressful events on coral reefs, for example,

Isabel_Houghton: [00:30:09] Exactly. Yeah.

Morgan: [00:30:10] possible. Very cool.  All right. So if. Somebody is listening to this and they are thinking, this is the coolest thing I would love to design accessible, small ocean technology objects that can address really big questions and

fill in these enormous gaps in the data that we have about the ocean. What would you want to say to this person? How do you get involved in something like this? Would you have any advice?

Isabel_Houghton: [00:30:37] I would say, figure out what it is in that domain that gets you really excited in the day to day. Do you like analyzing data and maybe coding, or do you like building things and tinkering with temperature sensors? Or do you like maybe thinking  about policy and thinking about how you motivate people to think about these problems.

And so there's  all sorts of different ways is to get involved even a, you know, a small startup, like where I work. People come from all different aspects, 

Morgan: [00:31:11] thank you so much for talking with us today. Really appreciate your time and expertise.

Isabel_Houghton: [00:31:16] Thank you so much for having me.

 

  Thank you all so much for listening. I want to make sure I give special thanks to Eleanor Duran and to dust on the radio for our theme song. One way trip to Mars. Next week, we're going to be talking with Juan Carlos Villaseñor-Derbez about communidad y biodiversidad, also called COBI, an NGO based in Mexico city that works to promote sustainable fisheries management in close consultation with coastal communities throughout Mexico.

It's going to be a really fun time. See you there.