8-Things Earth Observation 🌍 🛰 : Q2 2022
Analysing some key developments in EO industry from the last quarter
Hey! It’s been a while! Welcome to “8-Things EO” in which I take a deeper look at some developments in EO from the last quarter - lazily categorised into 8 sections.
Disclaimer: Please don’t treat this as an exhaustive list of all things that happened in EO, but rather an analysis of where the industry is going from a highly opinionated and happily independent consultant.
It was an exciting 3 months for EO: from conferences like ESA’s Living Planet Symposium (that I attended and tweeted about) to major announcements including the first multi-billion dollar contract from the NRO for commercial EO data (that I also tweeted about).
Without any further, let’s get to 8-things EO!
2 Interesting Trends
1. Sovereign Earth Observation 🚩
This is anything but new in EO, and, erm, in world affairs, but I was wrongly under the impression that NewSpace EO would somehow push this narrative away, as we look at global datasets shared with everyone, everywhere. But it seems like my naive assumption that countries would move away from hardware satellite procurement to software data procurement (data-as-a-service) is quite wrong. Given the geopolitical situation we are in, I believe I am becoming more convinced that countries would prefer investing in sovereign EO capabilities (satellites built locally and data processed locally) as opposed to procuring data from potentially foreign providers who have assets in orbit - either due to legal requirements (security) or due to sovereign reasons (jobs & know-how). Some examples from the past quarter:
Iceye is planning to launch a US-only “parallel” constellation of satellites - built in the US specifically for the DOD.
NaraSpace Technology, a South Korean startup raised ~$8M for launching a constellation of EO satellites and building an EO platform for local use.
Australian EO startup LatConnect is collaborating with the local player Gilmour Space to build and launch a hyperspectral imaging sat constellation.
Quad nations (Australia, India, Japan and US) launched a satellite-based maritime monitoring initiative - specifically to track illegal fishing, and then, one week later, China launched something similar with BRICS.
You have probably heard too many times that satellites are becoming easier to launch, but more importantly, the emergence of space-as-a-service (SpaaS) business models could interest some countries as they think about their national (civil & military) EO strategies. In any case, I expect to see more locally built EO constellations with a sovereign narrative being announced & funded over the next few years - with climate and national security being the two focus areas. Some might leverage on SpaaS, but my money is on nations choosing to build their own EO satellites - even if it happens to be one with no major technological advancement or differentiation from others.
2. Satellite-as-a-service 🛰️
More “as-a-service” business models in EO - I am going to shamelessly call this one SataaS. However, what Iceye announced should probably be called SAR-as-a-service. Although Iceye has officially sold its SAR satellites to fellow EO player MDA and the Brazilian Air Force before, it was interesting to see them announce this as one of their main offerings. What this translates to is basically what one of their executives said:
“We’ll sell the data if that’s what you want. We’ll sell a satellite, we’ll sell the full system to you, or we’ll sell you a solution if you want to do something specific for a specific use case.” - Jerry Welsh, CEO, Iceye US
Some countries that would want sovereign EO capabilities might be interested in SAR-as-a-service, but I am curious to find out if there will be interest from the commercial sector? Perhaps an insurance company - maybe, just maybe one of them that Iceye is already working with? Watch this space!
More Satellite-as-a-Service Examples?
In other news, Belgian EO startup Aerospacelab might be going in the same direction after they announced their new “megafactory” could dish out 500 EO satellites every year. Surely, they are hoping that the SataaS model catches on to justify this investment?
Oh, while we are on the subject, this might be an extreme form of the SataaS model, but seems like one doesn’t have to order a satellite and wait a year before launching, apparently one can just directly buy an operational EO satellite in orbit. Smart!
2 Often-Forgotten Things
3. The Importance of Anchor Customers and Early Adopters 💰
We understand the role of the US National Reconnaissance Office (NRO) as EO’s biggest anchor customer - we saw an example of that with the big announcement - but is this just the case in the defence sector? Will it work in the commercial sector?
Unlike the NRO, much of the commercial sector (agriculture, insurance, energy etc.) do not (yet) have large in-house remote sensing teams and data analysts capable of using & fusing different types of EO data - meaning they don't have: i) an understanding of the capabilities of EO data and ii) willingness to pay for EO data.
But, we did see some developments in the commercial sector in the last quarter showing that the trend is changing with more commercial companies becoming anchor customers and early adopters of EO - whether they are buying data or they are asking the companies to build products for them is yet to be confirmed.
Orbital Sidekick (OSK), which is building a hyperspectral imaging satellite constellation, was selected by Energy Transfer, an oil & gas company to monitor its assets (thousands of miles of pipelines).
Satellite Vu, which is building an infrared imaging satellite constellation signed a partnership with Landmark Information Group, a real-estate company to collaborate on testing products related to thermal efficiency of buildings.
Satlantis, which is building a multispectral imaging satellite constellation to monitor greenhouse gas emissions received a strategic investment from Encino Environmental Holdings, an environmental consultancy owned by BP.
What does this mean? Companies outside the defence sector are starting to understand the business case for EO, especially when the EO solution is verticalised. What is interesting is that these anchor customers and early adopters may chose to adopt their own EO strategy - that may set the tone for the future of EO, which is:
A “space strategy” and acquire EO satellites (or) A “data strategy” and acquire EO data (or) A “software strategy” and acquire EO products.
In the meantime, shall we take a moment to celebrate the opening up of RADARSAT-2 archives by MDA for use by an anchor customer, Global Fishing Watch? We need more of these anchor customers with specific use cases for solving important problems using EO, that might justify opening up the archives of EO data. More on this in my short Twitter thread!
4. The Need for *Scalable* EO Solutions 📶
So, we have some large companies becoming early adopters of our favourite EO technologies. Awesome - but I believe this is where the challenge starts and the real work begins.
As an industry, EO is stuck in what I call the “demonstration” phase - finding more use cases, building more algorithms and selling a number of pilots projects. At some point, we need to start thinking about the "deployment" phase i.e. how do we scale the pilots and integrate them into large-scale projects.
It’s a Software Problem
Fortunately for the remote sensing folks, I don’t think this is a remote sensing problem. This is a software problem. Having seen enough impressive pilots from different EO startups over the years, I am not fascinated by new use cases anymore (ok, I am lying - every now and then I come across a new application of EO and I am mind-blown!).
What I would like to see more is essentially how do we scale this amazing use case that we piloted in one specific location to the entire state, entire country and perhaps, the entire world. Monitoring one pipeline over a selected area might be easy, but monitoring thousands of miles of pipelines is certainly not.
And for that, we need more software engineers and data scientists in EO - who have nothing to do with remote sensing. Perhaps, even the very same ones that have built the Ubers, Instagrams and Tinders of the world that scale across countries, devices, languages and payment systems.
Bottomline: EO has a software problem and we need more software thinking in EO - after all, I have always said and believed that EO is just another type of data, that happens to come from space - data that we can build useful software with.
As if to prove my point on this one, Satellite Applications Catapult published a brilliant report on what needs to happen for EO to be used widely within the public sector. The recommendations aren’t surprising: solve the problems of accessibility, affordability, interoperability, reliability and scalability. Some highlights from the report below in a Twitter thread, by yours truly:
2 Business Buzzwords
5. Vertical Integration ⏪ ⏩
Now, let’s talk about my favourite things: business buzzwords, starting with vertical integration, of which there are two types. Let’s use examples from this quarter.
Forward Vertical Integration
This is what we are seeing from Planet, which announced the launch of “planetary variables” - moving from space-only vertical integration i.e. owning the build and operation of their satellites to full-blown vertical integration i.e. owning the development of products based on data from satellites.
Essentially, Planet is attempting to own and control business activities that are ahead of the value chain of its industry, which until recently was restricted to space/EO, but with this announcement moving down to specific industries - agriculture, in this case.
Backward Vertical Integration
A good example of this is The Husqvarna Group, a Swedish forestry equipment manufacturer, with multi-billion euros in revenue, that recently launched Husqvarna Intellion, a vegetation management solution powered by satellite data.
Interesting because even though there are several EO-based readymade products for vegetation management from Overstory, LiveEO, AiDash etc., the company decided to build their own in-house solution, aimed at Husqvarna's customers which are mostly utility companies and forestry companies, but choosing to smartly partner with existing players as and when necessary.
I wrote about backward vertical integration last time with examples from Palantir, ExxonMobil and AXA. And I don’t expect this trend to stop!
Note: “Vertical integration” not be confused with “verticalisation” - the latter is more about companies choosing to build products for specific use cases as opposed to those who offer multi-purpose data across sectors. So, basically like the development we saw from ConstellR, an infrared EO startup focused on the agriculture market acquiring ScanWorld, an hyperspectral EO startup, also focused on agriculture.
6. Go-To-Market Strategy 🎯
The second buzzword: go-to-market (GTM) strategy. I like this one because very few want to talk about it, but without a proper GTM strategy, EO companies with goals to offer subscription-based products will end up becoming project-based consultancies.
The GTM for EO Data Companies
We discussed the GTM for EO data companies in topic #3 above - which is to look for anchor customers willing to acquire data-as-a-service mostly through long-term contracts and occasionally subscription-based contracts. Satellogic deserves a mention for innovating on its GTM strategies - just in the last few months they have:
Partnered with Orbital Insight (OI) to bring their data on OI’s platform
Collaborated with Astraea to distribute EO data to Ukraine
Partnered with Palantir to demonstrate their edge computing AI on orbit
I admit I find their approach confusing sometimes, but then I tell myself we are so early in the development of EO that nothing has been figured out yet - there is no right or wrong!
The GTM for EO Platforms
I have a lot of things to say about EO platforms - and I did in this Twitter thread, especially about the need for verticalised, domain-specific platforms instead of generic, all-purpose platforms. But, we did see few interesting developments in Q2:
One of my favourite EO startups, Earth Blox raised seed money for their no-code, cloud-based EO processing platform.
Unexpectedly, a space tech satellite manufacturing startup launched a new EO platform aimed at enterprises to access and process EO data.
And finally, of course, we saw Google attempting to make Google Earth Engine (GEE) mainstream by releasing GEE for use by enterprises and governments.
The GTM for EO platforms is pretty simple - acquire as many enterprise/government customers as possible, ideally with as many users as possible. So, not very different from a SaaS-model. But, platforms are just tools, they don’t build business cases.
The GTM for EO Products
And finally, the most interesting EO segment simply due to the variety of applications available, but oddly enough, in my opinion, with only two scalable GTM strategies.
Based on partnerships and core integrations with the large enterprise software companies (SAP, Oracle, IBM, Salesforce etc.) or the market leaders within the specific verticals —> this is the GTM announcement we saw from Cloud to Street, a flood mapping startup.
Based on strategic partnerships with the large consulting companies of the world (BCG, Aon, PwC, Deloitte, Accenture etc.) —> this is the GTM announcement we saw from Jupiter Intelligence.
Of course, some customers may choose to directly acquire the EO products from the EO companies, but I don’t see how they can scale to large subscription-based contracts with high Annual Recurring Revenues (ARR) without this strategy. I might be wrong, but I fail to see how a company that is already paying hundreds of thousands of $ for either an enterprise software product or for consulting will want to pay more for an EO product.
That Other Thing
7. Weather ⛈️
Let’s discuss that thing we take for granted and yes, let’s be honest, that thing we talk about to fill awkward silences during conversations - the weather. I may have left Tomorrow.io, but it doesn’t mean I am going to stop analysing the state of the weather industry (no arguments - it is part of EO).
The last quarter had some interesting developments you might have missed:
Tomorrow.io officially announced addition of microwave sounders (passive sensors) to its planned weather satellite constellation of radars (active sensors).
Spire received a multi-million dollar subcontract from the DOD, making this the largest numerical weather prediction model deal to date.
A weather modeling company raised seed funding to improve long-term seasonal forecasting, which if successful, would transform many industries.
I have been surprised at how little traction weather gets within the space industry, considering it is the perfect example of space technology touching our lives everyday. Yet, it doesn’t get a mention in surveys about space, but, you know what, I am not surprised as “the amount of media coverage a segment of the space industry receives is inversely proportional to its overall revenue share to the overall size of the space industry.”
With respect to weather, I expect that to change as climate change (and ergo, weather) becomes the main talking point politically, economically and socially - we are starting to see that with US providing $3 billion to NOAA over the next 5 years, of which almost $1 billion is allocated for “climate data and services” - guess which part of the commercial space industry will be contributing for most of that in the future?
Anyway, if you are unsure what all of this means and would like a primer on what’s going on with respect to weather and space, I will be publishing my piece on ‘demystifying weather from space’ in August.
1 Personal Wish
8. The Need for Holistic Strategy in Earth Observation 🌍
Global crises such as the pandemic, climate change and the war in Ukraine may be creating new demand for Earth observation, but as the EO industry continues to grow, I just have one personal wish.
To start looking at solving problems through EO with an objective and holistic lens.
Because, at the end, it is not about a specific type of sensor (hyperspectral vs infrared vs SAR vs lidar), not about a specific medium (satellite vs aerial vs in-situ), not about EO vs weather vs GNSS and certainly not jargons such as geospatial or spatial or location intelligence.
I believe that an objective, holistic view of EO would enable us to design products with customer experiences that are more product-driven and less technology-based. It will allow us to "solve problems with our technology" and not "match solutions to our technology." And, maybe, a holistic, objective view might transform the EO industry to become more open-minded and collaborative with shared learnings.
But, here is the problem. Nobody, and I mean nobody is incentivised to be holistic and objective today, except maybe a few small EO service providers, who work on a project basis to solve problems. Every EO company is incentivised to push their own tech (and rightfully so), but how can we build an efficient product for the customers without having an objective, holistic view of the entire market?
This brings me to the relevant development from this quarter that might help solve this problem: EY launched satellite image processing service. I wrote about this announcement and my thoughts on the need for objective, independent EO consulting in this Twitter thread. As I wrote, I don’t think that’s the right approach.
Will consulting companies entering the EO market be the answer to what I described above? I am not fully sold, but I promise there is more to come from me, on this one!
If this was forwarded to you, subscribe below:
PS. EO, both the technology and the startups, are starting to get some kind of mainstream coverage, like here, here and here. But, we need to communicate more and better. Which is why I am currently realigning my focus for TerraWatch Space, the podcast and the newsletter. Stay tuned to hear more!