Episode #47 — Daniel Bailey the Co-Founder and CTO of Astraea talks about the changing world of eath observation and some of the challenges facing the industry at the moment. We also discuss remote sensing space with regards to the Gartner Hype Cycle and discuss the role of non-traditional players in the earth observation space and what that might mean for the industry
Link to a free data discovery platform mentioned in the interview
More information on the Gartner Hype Cycle
This podcast is sponsored by OpenCage Geocoder
OpenCageData does not make maps or provide routing information, they focus on turning address or place name to coordinates and coordinates into addresses or place names and they are amazing at it! If you need help with geocoding OpenCage Geocoder what you are looking for.
We made some socks for you!
You are more than welcome to reach out to me on social media, I would love to hear from you!
In Conversation
Infected by Satellites
Daniel: Hey Daniel, welcome to the podcast. Today we’re going to be talking about earth observation — the past, present and future. You are the co-founder and CTO of a company called Astraea. You’re dealing with earth observation data and building a platform around that, and we’re definitely going to dive into that later on. But before we get into that, it would be really useful if you could give the audience a brief introduction of how you got involved in earth observation and remote sensing.
Daniel Bailey: Absolutely, excited to be here and really passionate about this discussion and this topic. I got infected, really, with the power of satellites and extracting information from satellites as a young soldier in the late 90s. It was kind of heady times where you could pick up a phone and requisition a satellite in support of operational intelligence support centers, where we were supporting warfighters and leveraging that information — multiple data feeds from drone, aerial, satellite, human intelligence, what they call all-source intelligence — to extract information. That was kind of my first aha moment: wow, this is incredible, the information and data that is being collected. At that time in the late 90s it was pretty much isolated to the government, to the Department of Defense and intelligence type operations, specifically for earth observing satellite imagery.
Daniel Bailey: Kind of in the mid 2000s I transitioned out of the government into the private sector and was focused on helping individuals and organizations make sense of operational data. My math and stats background led me to the AI/ML field — it was just the burgeoning field that it is today. I was fortunate enough to work across a number of different industries: insurance, oil and gas, telecom, software, all around making sense of data, leveraging AI and ML techniques to do so.
Daniel: I just want to back up a second, because first of all I want to point out to the listeners that you are the very first infected geographer that we’ve had on the show. We’ve had lots of accidental geographers, but you’re the very first one who talks about being infected by this, and I think that was worth pointing out. Secondly, you talked about a maths and stats background. So does that mean you’re not coming at this as a geographer, someone with an interest in geospatial — you came at it from the side, using your understanding of maths and statistical analysis?
Daniel Bailey: Absolutely. I’m very much non-traditional. I didn’t have a background in geospatial or geography. I come at it from a math and stats, computer programming type of approach.
What Earth Observation Looked Like in the 90s
Daniel: It would be really useful for the listeners to understand what those early days looked like in earth observation. You talked about being in the military, looking at earth observation data and trying to get some information out of that. What did that process look like back then? Are we talking heads-up digitizing, where we’re drawing circles around interesting objects and georeferencing those?
Daniel Bailey: It involved all that. It’s a very manual process, from teams of people — it’s throwing bodies at it. We didn’t have a lot of the advanced techniques, the deep learning and a lot of those computer vision techniques that we’re seeing applied today to the space. So it was drawing, digitizing, cross-referencing that with other platforms and intelligence assets. Very, very manual process. I remember many nights going home with headaches — pretty much every night — because it’s just a deluge of information that you’re trying to make sense of.
Daniel: I could imagine. Identifying those features would be one thing, but tracking them across time would be a whole other thing, and then making sense of that with your team members — did you get that also, or is that the same object there? I could imagine those kinds of things that we perhaps take a little bit more for granted today would have been incredibly difficult.
Daniel Bailey: Yeah, there wasn’t a ton of great collaboration tools. We often think about satellite imagery and aerial imagery — that’s really where it started, in photogrammetry — but now everything’s sensored. And those sensors — the only thing we were doing with the imagery is really still a measurement about the earth. We’re observing the earth and measuring the earth. Now we can do it on a number of different platforms, and it’s just increasing.
Space 2.0
Daniel: Let’s stay with that theme, because one of the things I really wanted to talk about was the future of earth observation. We got a little snapshot of what it looked like in the past — that heads-up digitizing, and the solution back then was just more people. Get more people to do it and then we can do it faster. The world looks quite different today, and in a recent post on Medium you talked about a concept called Space 2.0. Could you give us an overview of what that means for you?
Daniel Bailey: I certainly didn’t coin, and Astraea didn’t coin, the concept of Space 2.0, but we certainly have been in the middle of it. You said something a second ago that was really interesting — it’s certainly not the way it used to be. Unfortunately it still is a lot of that way. I mean, you take the National Geospatial-Intelligence Agency, NGA — just three years ago they kicked off a huge modernization effort to try to get away from the manual processing of this data. So some of the leading organizations that extract information from this are still struggling with modernizing beyond throwing humans at it.
Daniel Bailey: So we’re just at the very beginning of scratching the surface of what this new wave is, and that’s what Space 2.0 is to me all about: how do we better leverage cloud architectures, AI/ML techniques, multimodal data fusion techniques, to unlock better insights about the planet and make better decisions?
Where Are We on the Hype Cycle?
Daniel: I’m not sure if you’re familiar with the Gartner hype cycle, but this sounds like a model we could apply to this situation to help better understand it. The hype cycle says that there’s a trigger in technology — something new is available — and I think that’s probably what we’re talking about here when we talk about machine learning and cloud architecture and the increased availability of satellites. If you follow the curve, straight away you get the peak of inflated expectations, and then we drop down to something called the trough of disillusionment, up the slope of enlightenment, and the plateau of productivity. Where do you think we are on this hype cycle?
Daniel Bailey: It’s interesting — I come at it from a very different perspective, I come from the AI/ML space. The whole satellite industry — it was only about 25 years ago, in 1984, where the US Commercial Space Act was put into place that allowed organizations to launch satellites into space for the purpose of remote sensing the planet and make it commercially available. So that’s only 25 years ago. We forget how short a time has actually passed. And it was in the 70s where we had the AI winter, where, to your point, it was a new shiny technique — we had more computer processing, we had things called neural networks and other things of that nature, and everybody got excited that we can mimic the brain and get to artificial intelligence. And then there was a big backlash, because we were at that peak of inflated expectations.
Daniel Bailey: I think there are some parallels to Space 2.0 and leveraging this new data source. There’s been a flood of venture capital and other money come into the space, a lot of startups like ourselves working in it, and a lot of the businesses and organizations are grappling with the business cases, the use cases. Where’s the justification to leverage this data? Does it really make a lot of sense to count cars in Walmart parking lots, or can I get better data that’s actually cash register receipts, or other ways to get that data that is more effective? And so we are at risk, I think right now, of falling into that trough of disillusionment.
Daniel: It was really interesting to have you put that perspective on it in terms of time. When this thing started — 25, 30 years ago — that’s a long time ago. We’re talking about the incredibly early days of the internet, long before smartphones. That seems like a long time in the tech world, so it’s understandable that people had some different kinds of expectations back then, and it’s understandable that people look back and go, wow, maybe we should have come further. So if we stay with this idea of the hype cycle, you think we’re perhaps falling down to the trough of disillusionment. Is that because people are struggling to find applications for this? Is there anyone who has found an application? Can we point at any one or two particular industries and say, okay, they have found a use case, it’s working over there?
Daniel Bailey: The most clear use case has been, and continues to be, Department of Defense and intelligence use cases. They have been leveraging this and are the leaders in cutting edge technologies for earth observation. But it’s really about now, how do you commercialize? Because it’s only been really commercially available for roughly 25 years. Insurance and finance and agriculture and others have went a long time without leveraging this data and have been just fine. Now the question is, where are those use cases?
Daniel Bailey: It’s really interesting to me — I think the timing and the time to adoption is being accelerated by a lot of the events that we see going on with climate change and a changing planet. This is a unique opportunity: the technology is available, the ability to launch satellites, the ability to measure the planet is available, to unlock some of those insights and to find those real use cases that will propel the adoption of this technology. And hopefully maybe we won’t spend as much time in the trough of disillusionment as perhaps other technologies have in the past.
Finding the Use Case
Daniel: Interestingly enough, we both did this — we both thought about the military aspect of this, we both thought about how governments are using it. But if you think about the name, what we’ve been calling this technology for a long time is earth observation. Observing the earth. Over the last few years it’s become more and more apparent that we’re facing some pretty big challenges in terms of what’s happening with our earth, and yet it doesn’t seem that that’s the go-to use case for this technology. Do you think that’ll change, or is it always going to be industry and profit driven?
Daniel Bailey: I think it’ll change. At least I hope it’ll change. That’s part of the reason why we created this company — to hopefully do our part to propel change. At Astraea we believe it’s the right of everybody to understand their planet and how it’s being used, and we’re trying to create better accessibility to this data and to leverage it for a wide set of use cases, but it is all focused on sustainability.
Daniel Bailey: I’m encouraged by the fact that one of the big thought leaders, Larry Fink, BlackRock CEO — one of the largest asset managers in the world — recently came out in his annual letter and said that the climate crisis will reshape finance. We’ve been fortunate to see that there’s a big movement in financial markets to focus on sustainable finance: what does that mean, and evaluating companies on their practices and having sustainable operations. One way to do that is to use earth observation data to evaluate how their operations are going, and to evaluate things like stranded assets as we transition from a coal-based energy environment to a renewable energy based environment.
Daniel: Right at the start you talked about your company Astraea and you said this is why we started the company. It seems to me that there are lots of companies out there that are interested in this field, trying to help other people and organizations learn what they can do with this data — they want a piece of the pie, essentially. How is what Astraea is doing different from these other players?
Daniel Bailey: I try to focus on a “yes, and” mentality. I don’t really look at it as competition as much. I think there are worthy competitors out there, but there’s no end to the problems that we face as a society and that we’re facing on a planetary scale. So we focused on creating an ecosystem and really trying to help overcome the fragmentation that we’re seeing in the space — all the different data providers, data sources. It’s big data, and there’s the fragmentation of the skill set that you need to access the data, to turn the data into information, to make that available to other folks that aren’t going to write in Python or use an AI/ML technique, that may want it on a map or in a chart or a graph.
Daniel Bailey: So there are a couple of different dimensions to this. There’s the skills dimension, and so at Astraea we focus on providing a number of different platform products that increase accessibility for a number of different user types. If you just want to look at an image and see the change from today and five years ago over your city, you can do that. If you want to get more advanced, we provide you the platform and the scalable compute assets to do that. We’re really focused on the business-to-business enablement — I think that’s where we’re a little bit different. We’re not trying to be the be-all and end-all solution provider. We want to enable specialists in the fields, all the different agronomists and utilities and energy players that could use this data. How can we help them incorporate this into their offerings and to the end customer that they serve?
Daniel: If a business comes to you — and I guess it really depends on what industry they’re involved in — what’s your approach to them? Is earth observation data always the right answer?
Daniel Bailey: No. I think that’s one of the things — as technologists we’re often guilty of creating a hammer and looking for a nail. That’s kind of our predisposition as technologists. Earth observation is cool, and it’s a lot of big data, and AI/ML is cool, but it’s a technology, and we need to be business driven. What are the problems, and what’s the best way to solve that problem?
Daniel Bailey: And that’s why, when I think about it — we are on a geography podcast, and that’s fair — when we think about earth observation we typically equate that to satellite. But there are a number of different platforms and increasingly other data streams that are observing the earth, including from your cell phone. There are a number of interesting technologies going on, crowdsourcing platforms where people are taking pictures that then can be stitched together and fused with satellite data and lidar data and create awesome 3D maps. So we can fall in love with the technology, but we’ve got to fall in love with the business problems and what’s the best way to solve that.
Daniel: I completely agree, but I could see how it’s confusing for people. It’s exciting and confusing. On the one hand we have what seems like an incredible opportunity, all this new stuff coming onto the market, and history has taught us that the first people to grab these new tools and find a use case for them can have some huge benefits. If we think about the early days of the internet — the first people to make websites, the first people that realized they could buy AdWords on Google — they had a massive advantage. So I could understand people wanting to grab these tools and make use of them as soon as possible. So it’s really interesting to hear you say it’s not always the right answer. When is it the right answer?
Daniel Bailey: The shiny example that we’re seeing, and this is why to me it’s exciting — you think about global development and sustainable development goals from the UN, and so understanding across borders economic indicators and other things. That is a purpose-built use case for this data, just because of the large geographical region that data can be collected on from the earth observation satellite platforms.
Daniel Bailey: If you’re a more localized business you’re like, well, that’s great, those are big governmental use cases, and you’ve got the DoD use cases, got it, that’s interesting — but what about me, how does this impact me? I think that we’re seeing more and more use cases on infrastructure monitoring, right-of-way monitoring, in the property insurance space, understanding what’s on homes — are solar panels going on homes. There are some very interesting use cases that are much more localized that this data is increasingly being used for.
The Training Data Bottleneck
Daniel: I think monitoring in general is a really big category — just being able to watch things change, see how they’re developing and try to identify trends based on that. But I can see a massive amount of barriers too. What about identifying meaningful change? That must be something that’s difficult. To say a pixel’s moved from red to green over the last three or four days, we can all imagine that and do that. But meaningful change is something different. What does the timeline for this look like — is this something that we’re going to realize in the next five years, 10 years, 15 years, or are we going to be waiting another 30 years?
Daniel Bailey: That’s a great question. If I could answer that we’d both be rich. It’s a super challenging space. Without getting too techy on the AI/ML case — you’re absolutely right, change and understanding the change is of utmost interest to a lot of folks. There are a lot of barriers to that in leveraging this data. Cars in parking lots is not going to be the same algorithm as coral reef health, monitoring that across the entire Caribbean. Detecting that change and the required processing of the data and the algorithm is very different from one another. There are components that are shared, but it’s different.
Daniel Bailey: The barrier specifically in teaching a machine is you still have to have that labelled training data. You have to have not just examples of the change you care about, but of other changes, so that the computer can start to discern. When you think about kids, it’s not that much different. I have a couple of young kids, and at first they start to say everything’s a cat — everything that has a tail and ears and is a little bit fuzzy is a cat. And then over time you start saying, no, that’s a dog and that’s a cat, and they start to understand the difference between those two species. That’s equivalent when we’re talking about AI/ML.
Daniel Bailey: This is what makes this really challenging about applying it to earth observation data: there are just so many variations. The planet’s a big place, and there are so many different types of change and underlying land cover types, that we’re going to be busy for quite a while to get to a fully automated capability to detect all types of change, of all variations, for all use cases. It’s going to take a lot of dedication to creating those labelled training datasets. There’s some great work going on from non-profit groups like Radiant Earth, that has an ML Hub focused on creating quintessential training datasets and benchmark datasets, so we can understand how these algorithms are doing across a number of different use cases. And then there are a number of interesting open source projects to help with the labelling of that data, and increasingly companies — social good companies and others — that are focused on providing those labellers. It kind of started with Amazon Turk, which was the early indication of that in the AI/ML space, but doing this on geospatial and earth observation data, there’s uniqueness to it, and so we’re increasingly seeing more specialized companies and toolsets to help us create those labelled training datasets.
Daniel: I’ve been recently talking to other thought leaders in the space, and it seems to me that the general consensus is that skilled professionals in this field are going to move from doing these kinds of analysis on images themselves towards helping create training sets for these algorithms. Would you agree with that?
Daniel Bailey: It has to happen. There is a lot of interesting work to try to create synthetic datasets and use other techniques, which are really interesting — these kinds of physics-based synthetic datasets. There’s a lot of academic research going into that, and that’s an effort to try to shortcut the manual process of putting people against creating these labelled training datasets. But we’ve got a ways to go there before we can only rely on those kinds of synthetic data processes.
Ground Stations and Algorithms On Board
Daniel: We mentioned earlier that earth observation is not just data collected from satellite platforms — it can be from a lot of different kinds of platforms. Aerial platforms, drones, cell phones, and increasingly perhaps also cars, with cars being equipped with cameras and moving around the world. Is there any one of those platforms, or perhaps one I haven’t mentioned, which you think has a whole bunch of really exciting capabilities or promise at the moment?
Daniel Bailey: Interesting question. As a technologist I see the value of all of them. I think satellites, as a class of data, continue to be exciting. There are a number of impending launches of new sensor types — hyperspectral, more synthetic aperture radar, lidar, RF, other sensor types that are going to help us understand methane emissions and a lot of these other big drivers of climate change. So that’s super exciting.
Daniel Bailey: One of my good friends recently started a company called Pixel8, and they focus on taking cell phone data — where you take pictures — and conflating it with satellite derived imagery to create these high fidelity 3D maps. I think that’s a really interesting use case too. Globally everybody’s walking around with a cell phone in their pocket, and those are very powerful machines that have a concept of position and the ability to observe the earth. These crowdsource platforms that are pulling that in — it’s really, how do we handle that data, and then how do we fuse it with other data, and orthorectify it, and all the processing steps we have to do to make sense of it?
Daniel Bailey: We are seeing a lot of drone — more and more demand for drone data is coming. But once again, being a young industry, in earth observation we’re still grappling with having standards and standards-based approaches, so that it makes it easier to discover these data and easier for these data to work with each other. I think continuing to push on that’s going to be important, so that we can make use of all these exciting new data sources coming online.
Daniel: That observation really speaks to that idea of market fragmentation that we’re seeing at the moment. There are so many players out there. In some ways I feel like a lot of them are going to have to die off, and we’re going to have to establish some clear winners, and those winners are going to step up with their standards and say, okay, this is what earth observation data looks like from now on. And even though that sounds really bad for the industry, I think in some ways it’s going to simplify things a lot.
Daniel Bailey: Absolutely. I think we’re already seeing it — there’s definitely some consolidation going on. What we haven’t seen yet, though, is — we see consolidation within one of those kinds of data streams. So in the satellite space we’re seeing some consolidation. One of the big newsworthy events was when DigitalGlobe and MDA merged and created Maxar. But that was still within that kind of satellite space. What’s interesting is the players like Microsoft, and AWS announcing last year that they were going to outfit most of their data centers with satellite ground stations. So I think you’re right, but it’s going to be interesting to see how it plays out from this state of earth observation where we have technology companies that are working across all these platforms.
Daniel: Could you just walk the listeners through what you mean by these ground stations? Are they satellite launch pads, places where data can be downloaded from satellites? What does that mean?
Daniel Bailey: It’s cheaper than ever to get to space, it’s cheaper to launch. You’ve got SpaceX putting up 14,000 communication satellites and they’re launching no matter what, and there are people doing rideshares. But those ground stations are the critical piece, because those ground stations spread across the planet are the opportunity for, when the satellite comes within range, that it can transmit the data that it’s collecting back down to the earth to be processed and leveraged. And so having these ground stations co-located with data storage and compute assets, that we more and more do in a cloud native environment, reduces that time to operationalizing it, to getting it and making decisions on it as an organization.
Daniel: That’s really interesting. That’s a piece of the satellite infrastructure that I don’t often think about. I tend to focus on the rockets that are launching the satellites, and not think about how much it takes to get that data back down to earth. And of course it makes perfect sense to be downloading that data at the same place where the data is potentially going to be processed, instead of having to move those huge volumes of data over to another storage center to process it over there.
Daniel Bailey: No problem. I think the next future, as it gets launched, is more and more — as we find these high volume algorithms like detecting methane, those algorithms are actually being put on the sensors themselves. So instead of transmitting down the raw data that then has to be processed, we’re starting to see some applications where we’re actually applying the algorithm on board on the satellite itself and transmitting down the answer. That’s really where the space is going — how do we speed up the time from the collection of the measurement, the pixel if you will, to the insight and to drive decisions?
Daniel: I like that idea of taking the algorithm to the data as much as possible, as opposed to trying to bring all the data over to the algorithm. Why not? I guess it depends on the compute power that you have on board these different platforms, but when you say it like that it makes perfect sense.
Daniel Bailey: It’s exciting times right now. Those types of applications are really being focused on million dollar problems. I think though that the applicability of earth observation data, and what it’s going to take to leverage it to solve some of our big global problems — it’s really going to take us continuing to create better accessibility, so that we can solve for folks the ten thousand dollar problems without that amount of resources, because it’s pretty expensive still.
Daniel: In terms of access to data, one of the things I thought about was, well, that’s another way of condensing the data. And that must make things cheaper and faster, and hopefully will mean we’ll get more penetration out into the market because it’ll be more accessible.
Daniel Bailey: It absolutely is, it’s exciting. My AI/ML side of me says I still want some of the raw data coming down, so hopefully we don’t just do that — because there are a lot of opportunities, and that’s been the whole excitement around machine learning: to use data that was collected for different purposes and unlock hidden insights in it, and ask questions that we didn’t collect it for to begin with. That’s exciting too.
Data Fusion and Death by a Thousand Portals
Daniel: I’ve just got a couple more questions. The first one is back to an idea we talked about earlier — data fusion. Primarily when I think about earth observation, when I think about remotely sensed data, I tend to think of one channel of data, one data stream. But I think that perhaps the opportunity might be in fusing data with lots of different kinds of streams. Even if we stopped collecting data now, could this actually be the real gold rush — fusing data instead of collecting more of it?
Daniel Bailey: It’s interesting, I haven’t thought about it in those terms before as far as it being the end game, but I definitely think there’s a gold rush there. We’re seeing some applications of that, and we at Astraea have done it, where we’ve been fusing passive optical data streams — you can think pictures — with some of these active sensors, the synthetic aperture radar. That’s been viable because synthetic aperture radar, while it can be noisy, has the ability to penetrate clouds. And traditionally in earth observation, clouds are the bane of existence of any remote sensing scientist. So being able to fuse those two data sources together — we’ve seen huge improvements in our modelling to be able to classify crops and other applications that we’ve worked on. So I think there is a lot of opportunity to fuse data. It’s just now, literally in the last six to nine months, that we’re really starting to see some compelling use cases. But you’re right — more and more data streams do have a chance that, as technologists, we can get distracted by all the cool shiny objects coming, as opposed to staying focused and really finding that problem that that technology answers.
Daniel: It’s interesting, we keep coming back to: yes, we’ve got all these shiny objects, we could do this and that, but it still really feels like we’re both missing that use case. We’ve talked about a few different use cases, but it feels like everyone’s still fumbling in the dark trying to find one. They’ve got the tool — where’s the problem? I’ve got a hammer, where’s the nail?
Daniel: My final question: we often talk about being overwhelmed by these huge volumes of data, and for me I always think of earth observation data. But as a geospatial professional — I work as a consultant in my day job — I don’t often feel like I’m overwhelmed by my choice of building footprint data, for example, or the different road datasets I could use. And earth observation data is often the basis for a lot of this kind of stuff. Traditionally it would be digitized and then make its way through the system down to people like me. When are people like me, at the bottom of the data chain if you will, going to see a flood, a wave of data coming down towards us?
Daniel Bailey: It’s interesting. We’re starting to see more — there are a lot more open projects, collaborations that I think will start to produce more and more data. I’m always encouraged by times where — Microsoft just recently released their building footprints dataset. There’s plenty of data out there, to your point. The question is that a lot of the big use cases have been larger geographies, and I guess the question is how do we enable folks that are focused on delivering localized solutions to access more of the data — one, discover the data that is out there, and then know that they can incorporate it. Because I think there’s actually more data out there than perhaps you’re aware of and have used in your traditional day job. But in this fragmented marketplace it’s almost like death by a thousand portals. It’s really hard to find the data that you can use and to use it together.
Daniel Bailey: That’s one of our products at Astraea — our Earth OnDemand is focused on that data aggregation, data discoverability, helping pull from all these different portals and putting it into a place where you can discover it and then go get it from the different portals. So I think there’s actually a lot more data out there than typically is being used. But there’s a cost that comes to that — you have a deadline, you have something to do, and you know the data sources and you trust those data sources. So how do we increase the visibility and data discoverability? I think that’s a big challenge and opportunity.
Daniel: Would you mind providing me with a link to that source you mentioned? I think that’d be really interesting for our listeners to be able to follow up on and check out that data discoverability tool that you’ve built.
Daniel Bailey: Sure. You can go to earthondemand.org. It’s a free tool that you can use — it doesn’t require any big sophistication, no coding. It’s simply interacting with the map, drawing your area of interest on the map and the time that you care about, and then you can start to see what free image resources are available to you. Then you can look at those — we do the processing right there on the fly to let you see a natural color representation — and then provide you capabilities to get the underlying scientific data to incorporate.
Daniel: That sounds really interesting, I will definitely include a link to that in the show notes. Daniel, this has been a fascinating conversation. I really appreciate you coming along and telling the listeners a little bit about earth observation. Where can we go to reach out to you if we have questions?
Daniel Bailey: You can check out our website, astraea.earth, and then of course we’re on Twitter and LinkedIn — @AstraeaInc is our Twitter handle, so you can follow us there, and we have a LinkedIn company page that we connect with as well.
Daniel: I’ll be sure to link those up in the show notes. Thank you so much for your time, I really appreciate it.
