Netflix and Earth Observation: Pondering the Strange Similarities
Netflix and Earth Observation: Pondering the Strange Similarities
This is not expected to be a professional write-up, but rather the raw ramblings of an Earth observation enthusiast.
Increasing Search Costs in Video Streaming
So, you decide to sign up for a video streaming service provider because you finally discover you have a couple more hours per day, because you do not have to commute to work during the confinement. Or, you are already a subscriber to a few (or all) of them, so you decide to watch something to spend the extra time you did not think you would have. These are the options (see figure above) that you are faced with (depending on where you live, this could be more or fewer). You think to yourself:
These video streaming companies were supposed to make my life easier, aggregating content on to a single platform, and reduce my overall search costs. Are they?
Video Streaming and Earth Observation
I was working in the software industry for a few years, including a couple of years at Amazon (whose value proposition is to reduce the search costs for products) after which I decided to take the plunge into space sector. As I began this unexpected journey, I developed a keen interest for Earth Observation (EO), the domain of space that deals with satellites that monitor the Earth and generate a lot of data leading to some insights that I did not associate with space before. I was amazed and found this to be revolutionary. I did not know you can monitor changes on Earth over time and visualise that from space. I did not know you can use satellite images from the night to calculate economic growth models based on “night lights”. I did not know you can use data from satellites and count the cars in parking lots to provide input to consumer spending models. This timing of my interest into EO also coincided with the growing EO sector, with hundreds of satellites being launched every year, meaning petabytes of data was being collected every day. So, I am not entirely surprised that we see more satellite imagery around us than ever before in the news media, even from Donald Trump.
So, I wanted to spread the word. I became an Earth Observation evangelist, if you will. I gave talks to students in schools and universities in India, France, Mexico and Australia. I participated in conferences and talked about on how satellite could be useful and encouraged people to “think out of the (box)atmosphere”. Thankfully, my daily job involved doing a similar type of evangelisation for the project I have been involved in, called “Insights from Space.” As a result, I got a lot of interest from people from the non-EO community who were surprised to know about all this and wanted to know more. And they asked me some simple questions, which I still have trouble giving short and simple answers to. They ask me:
So, how can I find and access satellite data? Do you recommend a tool I can use to perform some analysis on satellite data? How much does it cost?
The answer for all those questions unfortunately is, “it depends.” It is not because I have put on my consultant hat, but it really depends — what problem are you trying to solve? What is the spatial resolution do you need? What is the frequency of imagery you want (hourly, daily or every few days)? Do you want data downloaded from satellites real-time or do you want to use archived imagery? And finally, what is your budget — if you want to use free data, there are a few ways of doing that and if you want to buy satellite data, there are a bunch of companies that provide different types of data to solve different problems. Let us assume that you have patiently listened to my responses and understood the nuances. Now you are ready to go access some data and start building those applications. Alas! How do you choose?
The following satellite data providers (see figure below) is what you are faced with. Here is where I go back to my video streaming analogy. Similar to the options there, each of these providers below have their own differentiating content. Netflix offers ‘Lost in Space’, Amazon offers ‘The Expanse’, Disney offers ‘Star Wars’ while HBO offers ‘Game of Thrones.’ Similarly, Maxar offers 30 cm resolution, Planet offers daily frequency, Iceye offers radar data while Copernicus is free. As with the former case, each of the satellite data providers have great content i.e. data. Each of them offer different types and configurations of data which are all essential components for solving various industrial, policy-based and environmental use cases.
However, there is one major difference. Video streaming does not suffer from a lack of awareness, most digital consumers are aware of these brands. Earth Observation is still in its early stages. Another analogy: Earth Observation is where the internet was in the early 2000s — you had this technology called web browsers and you had options of Internet Explorer, Netscape, Amaya, Safari, Opera and Mozilla Firefox. Very few people knew what they were for and how to use them.
The applications, complexities and nuisances of EO data are very well-known within the bubble of the EO sector and its traditional users such as the military and governmental institutions and partially for the agriculture and insurance sectors. Outside this, however, search costs become enormously high due to the complexity of the type of data offered, the unstandardised business models (per sq.km or volume-based or yearly subscriptions) as well as the said lack of awareness.
I can think of a few approaches that will play a role in the uptake the of EO data. As an EO evangelist, I do want more people to discover the potential of this data, either directly or indirectly.
Aggregators / Marketplaces:
Marketplaces aggregate content from different satellite data providers and act as a single point of access for satellite data. Although an “Amazon for EO data” does not exist yet, some start-ups are taking this path — Airbus’ Up42, Astraea, SkyWatch, Arlula, Maxar’s GBDX & Google Earth Engine, to name a few. However, the lack of interest from some satellite data providers in the marketplace model coupled with the complexity in sorting out the already complex satellite imagery business models, makes this approach challenging. Still, I remain very enthusiastic about this direct approach. I hope, sometime in the future, there will be an Android-like framework for EO data, where one can get access to different types satellite data (and possibly other types of data), develop various sorts of applications and release them into an EO app-store (like a Google Play).
2. Insights-as-a-Service (IaaS):
An indirect approach to increasing EO data uptake is to filter out the complexity and avoid search costs for customers by just offering insights from satellite data (raw numbers downloadable as tables. This is exactly what companies such as Orbital Insight, Descartes Labs, SpaceKnow and Ursa Space Systems offer. They extract data from satellites, combine that with other sources of data and offer answers to some key questions. Some companies such as BlackSky are taking this route from the scratch, and many other data providers are also jumping on this bandwagon. This approach simplifies the business model (almost SaaS-like) and does not require interested parties to fully understand what spatial resolution, spectral resolution and revisit time mean. However, different parties require different answers and the challenge is in identifying these key scalable use cases.
3. Consultancies / Advisory Firms:
Or, perhaps, the answer is somewhere in the middle. Perhaps, interested parties do not need to know much about EO data, they just want answers to questions. EO data could be just one of the types of data used to build models that could help in answering the big questions for organisations. As far as the consultancies and advisory firms justify these costs, the customers are happy and they do not have to get into the complexities I described above. Organisations might just choose to outsource this part to consultancies and advisory firms, and trust them in doing the right thing for them. Would this approach increase the use of EO data exponentially? Probably not, but at least it could be used in circumstances that fit well, with the right approach.
As it happens in any industry, I expect some consolidation in the next few years, which should bring down the choices available for end customers. Maybe there will not be so many choices in front of us, both for video streaming and for EO. Maybe there will be an aggregator service (a single-point-solution) for video streaming. In the case of EO, perhaps the business model and value proposition will become easy enough to explain in one slide and the uptake will rapidly improve. OR, it could also be the case that none of this is a big deal. People will continue to subscribe to all video streaming services and maybe the use of EO data will grow organically, by continuing status quo and things will be figured out. As for now, I think I want to watch some comedy, so I am going to go search for interesting shows on Netflix, Amazon, Disney, HBO etc.
Until next time!