In this episode, we welcome back Lauren Guy, CEO and founder of ASTERRA, a groundbreaking company using L band and synthetic aperture radar (SAR) for commercial purposes. Lauren shares his journey as a geophysicist and discusses the innovative applications of L band in detecting water leakages, soil moisture, and even minerals from space. Dive deep into the technical, commercial, and environmental aspects of SAR technology and learn about the future potential of this exciting field.

Key Topics Covered:

1. Introduction to Asterra: Overview of ASTERRA’s mission and Lauren Guy’s background as a geophysicist. The unique use of L band and SAR for commercial applications.
2. Understanding L Band and Synthetic Aperture Radar (SAR): Explanation of the electromagnetic spectrum and how L band fits in. Advantages of L band, including its ability to penetrate the ground.
3. Technical Insights into SAR: Detailed discussion on polarizations, signal processing, and the electrical properties of materials detected by SAR. Comparison between L band and other bands like X and C band.
4. Applications and Benefits of L Band: Real-world examples of how Asterra uses L band for water leak detection and soil moisture mapping. Discussion on the environmental and commercial impact of these applications.
5. Challenges and Limitations: Addressing issues such as noise interference from cell phones and radars. Limitations in resolution and the complexities of SAR technology.
6. Case Studies and Success Stories: Success stories including the detection of 118,000 water leakages worldwide and the discovery of significant lithium deposits.
7. Business Strategies and Market Penetration: Insights into Asterra’s business model, customer education, and market challenges. Strategies for overcoming barriers and building trust with clients.
8. Future Aspirations and Technological Developments: Plans for launching their own satellites to ensure reliable data sources. The role of AI in enhancing SAR capabilities and improving detection accuracy.
9. Entrepreneurial Advice for Remote Sensing Practitioners: Lauren’s advice for remote sensing scientists and entrepreneurs in the industry. The importance of data feedback loops and the journey from a 20% to an 86% success rate in detections.

Guest Information: Lauren Guy, CTO and founder of ASTERRA. Connect with Lauren on https://www.linkedin.com/in/lauren-guy-asterra/

Company Information: ASTERRA. Learn more about ASTERRA’s innovative solutions at https://asterra.io/

Episode Highlights:

• “We can find water leakages from space and distinguish treated water from other types of water based on their dielectric properties.”
• “ASTERRA has verified, dug, and fixed 118,000 leakages across 65 countries using L band SAR technology.”
• “Our success rate has increased from 20% to around 86% thanks to the integration of AI and continuous data feedback.”

Recommended Listening

• Finding Water Leaks From Space
• Introduction To Synthetic Aperture Radar-SAR
• Flood Monitoring From Space (using SAR)

In Conversation

What Is L Band?

Daniel: You’re the CTO and co-founder of ASTERRA, and this episode is here to answer why people should care about L band. Let’s start with the basic idea of bands.

Lauren: I’m Lauren Guy, a geophysicist from Israel. I founded ASTERRA around 11 years ago, in 2013. ASTERRA is quite unique because we use L band and synthetic aperture radar for commercial uses, which you don’t usually see. During my studies I worked on SAR projects — the most interesting was using SAR to find indications of water on other planets, specifically Mars, and that’s how I got to know SAR and this whole industry. To take a step back: most remote sensing information is in the optical light — very short wavelengths, around nanometres. SAR works on a much longer wavelength, in the microwave spectrum — wavelengths of centimetres, tens of centimetres, even metres. Within microwave there are many bands — X, C, P, Ku — divided by wavelength. We use L band, the longest wavelength currently usable with synthetic aperture radar, around 30 centimetres, a frequency of about 1.3 gigahertz. Its main capability is that it can penetrate the ground all the way from space.

Why L Band Penetrates the Ground

Daniel: Can the other bands penetrate the ground?

Lauren: Every band has its own benefits, but no — shorter wavelengths like C band or X band can’t. There’s a rule of thumb that you can penetrate the ground about 5 to 10 times the wavelength. L band’s 30-centimetre wavelength gets you 1.5 to 3 metres of penetration; X and C band wavelengths are just a centimetre or two, so you only penetrate around 10 centimetres. Actual ground penetration is only available with L band.

Daniel: What are X and C band typically good for?

Lauren: The most common SAR band is X band, well used mostly in the defense industry. SAR can penetrate the atmosphere and it’s an active satellite — you’re not relying on sunlight — so you can illuminate Earth and see any location, day or night, in any weather. X band has great resolution, a few centimetres or sub-metre, which isn’t the case with L band, because L band has much lower energy. There are private X-band constellations, like Umbra, providing great images — mostly for defense, unfortunately, because I’m a big advocate of using this technology for environmental and commercial solutions.

Polarizations and Electrical Properties

Daniel: Back to L band — what else can it do?

Lauren: You need to think about it differently. With optical, light comes back into our eyes and we interpret what we see. SAR is actually measuring electrical properties — it’s signal processing; every wave coming back has a value representing the electrical properties of the material it interacted with. SAR is also divided by the polarizations of the images. In most cases it’s a single polarization, the easiest to achieve; at the other end of the spectrum is polarimetric SAR, using different polarizations to derive more information. Every material on Earth — or below it, since we can penetrate the ground — has different dielectric properties, so if you know in advance the dielectric property of what you’re looking for, you can detect it and have a great target detection tool.

Daniel: Could you explain polarizations?

Lauren: We all see light in a single polarization. Think of a wave going up and down; sometimes we have glasses that polarize the light and we see different colours — it’s a kind of filter that changes the polarization. If you use four polarizations, you get four different signals back from one target at any given time — it’s like a cheat code to get more information from one image. And it’s done on the sensor at the time of imaging: the antenna sends a pulse as a horizontal signal, then very quickly one in the vertical, and receives them in different ways.

The Trade-offs: Resolution and Noise

Daniel: It’s never all positive — what are the disadvantages?

Lauren: Because of the low energy, L band doesn’t have the best resolution. The main satellites — ALOS-2 from the Japanese space agency JAXA, and SAOCOM from the Argentinian agency CONAE — have a spatial resolution of around 10 metres. Any remote sensing specialist will laugh at that. But that’s the trade-off: you get poor spatial resolution but great radiometric resolution — how well you can differentiate between targets within those 10 metres. The other shortcoming is a very high entry point — it’s not intuitive like optical, it’s a niche, heavy on mathematics and physics, so it’s hard to find employees. But it’s rewarding: visible light is only 0.003% of the spectrum — we’re basically blind — while the microwave and radio spectrum is 15%, so there’s so much around us we can see.

Daniel: I remember from our last conversation there was something about noise and cell towers.

Lauren: L band is around 1.3 gigahertz — the same frequency as cell phones, mostly the older 2G and 3G generations. So if you image an urban environment, you get a big white cloud of noise from all the cell phones in the city. That’s why L-band research is usually in the middle of nowhere — deserts, agriculture. Overcoming that noise is one of the biggest obstacles for commercial urban solutions. Honestly, 90% of our IP is cleaning noise — we’re a noise-cleaning company. We want to make London appear like it’s on Mars, with no cell phones, and then get the information we need from the soil. That 90% of noise removal is the same for any solution, so to scale up we only need to change the other 10% — great for a startup. We use polarimetric SAR to identify the source of the noise, which can also come from airport radars, high-rise buildings, even vegetation.

Plumber from Space: Finding Water Leaks

Daniel: Your kids call you a “plumber from space.” What do you mean by that?

Lauren: Most SAR applications are defense, and especially in Israel you tend to end up in defense industries — I didn’t want that; I want to help the environment. I was amazed to learn water leakage is a huge global issue: 30% of the water taken from reservoirs or desalination plants is lost in the pipes before it reaches your house — up to 50% in the UK, 40% in the US. And 20% of a Western country’s energy goes to pushing water, so 7% of a country’s energy is just going into waste. During my studies I worked on finding liquids on Mars, so I figured we could do the same on Earth. If I know the dielectric properties of treated water versus other water, I can find it under the ground from space. Up to now we’ve verified, dug, and fixed 118,000 leakages, working with 65 countries — one algorithm that works everywhere. After 11 years, we’re still the only commercial L-band company out there, which is crazy, but I think that will change as more L-band satellites launch, including one from NASA.

Soil Moisture and Mineral Detection

Daniel: Last time we spoke you mentioned mineral detection. Are you doing that now?

Lauren: We ran a contest in the company where every employee suggested an idea, and one we chose was mineral detection — specifically lithium, because lithium is the future for EVs and has great electrical properties, which is why it’s used in batteries. If you Google ASTERRA and lithium, you’ll see we found big lithium reservoirs in the US. It was so successful we spun that technology off into a separate company, because it’s such a different industry and business case. Every mineral has a certain dielectric property, so if you know what you’re looking for, yes — you can detect gold, copper, and so on too.

Daniel: What about soil moisture mapping — is that the same as finding leaking drinking water?

Lauren: It’s actually much easier. Finding a leak is looking for a very specific, tiny target — a needle in a haystack, in a 10-metre pixel, in a chaotic urban environment. With soil moisture you’re measuring any kind of moisture. Unlike most soil-moisture solutions, which are indices derived from thermal bands, we build from the bottom up: we take soil samples around the world and task the satellites to image at the same time, so we can predict the actual percentage of soil moisture. It matters because soil moisture is the biggest enemy of infrastructure — 75% of infrastructure issues around bridges, rail tracks, and highways are caused by soil moisture that isn’t drained correctly. We can monitor huge areas and pinpoint utilities to the exact spots with high soil moisture near key infrastructure. We’re not only saving water — we save lives.

The Business of L Band: Actionable Insights

Daniel: Let’s talk about the business side. You mentioned you don’t work with general indices — you build your solutions from the ground up.

Lauren: The market is so used to indices and the vegetation-index type of thing. When we come to a client, especially in government or utilities, most don’t know anything about remote sensing — they may have bad past experiences — so we have to relearn them. Indices are usually taken into other models as general information; what we give is very actionable, specific information: go to that location and fix it, based on our information and no other. That’s a change in how they work, but if the client sees the benefits, they’re on board immediately. We deliver it as SaaS — we’re not a project-oriented company. We have an application called EO Discover; every client logs in, sees their tasks for the next few days, goes and fixes them, and reports back.

Daniel: How did you convince people to work that way?

Lauren: When I launched the company 10 years ago, I was sure everyone would be standing in line because it’s so cool. I learned the opposite — people don’t care that you bring the solution from space or use fancy words; they just want a working solution, especially in old-fashioned markets like utilities and governments. Sometimes people looked at us as charlatans. The best selling tool we had: as Israelis we can be not very PC, so we’d analyze a city in advance, then show the utility manager all his leakages on a map. He’d say, “Who told you about that? Why are you looking at my city?” It’s a punch to the face — if you can frighten the other side, you can sell in an instant.

Closing the Loop, Satellites, and Advice

Daniel: Tell me about that feedback loop.

Lauren: When the client reports back — there was a leak, or there wasn’t — that gets into our system, and our AI learns from it. We close the loop: delivering the insight and collecting back what actually happened in the field, which you don’t see a lot of in remote sensing. When we launched, our success rate was only 20% — two out of ten locations correct, which sounds awful to a scientist, but it was still twice better than what they had. Today we’re around 86 to 90%. We had a glass ceiling around 50 to 60% with geophysics alone, because of SAR’s chaotic nature; once we built an AI department four years ago and incorporated the feedback, we broke through. My advice to others in remote sensing: don’t be afraid to close the loop and request the information back.

Daniel: Would you ever launch your own satellites?

Lauren: We use satellites belonging to space agencies, which aren’t commercially oriented — sometimes I ask for an image and they just say no. We have aspirations to launch our own; we’ve been in deep discussion with very big names, and hopefully we’ll have something to say in the next 12 months. We’d aim for a constellation of three or four satellites to get an image every day, which would let us move to mission-critical solutions — we’ve had a patent for finding gas leakages for six years, but we can’t sell it yet because a gas leak needs to be found today, not next week.

Daniel: Any advice for remote sensing scientists and entrepreneurs?

Lauren: Find a market that isn’t necessarily using remote sensing today but is technological enough to see the benefits — like insurance, which uses a lot of resources for underwriting — rather than a red ocean of rivals like agriculture. And as scientists we want the perfect solution all the time — we started at a 20% success rate, and I almost decided not to commercialize it because, as a scientist, that’s just not a good number. But you can start pushing to commercialization even when you’re not perfect. Perfection comes as you get more data, and data is key.