We are living in an increasingly complex world. In a 2019 interview with Geospatial World, the CEO of Capella Space, Payam Banazadeh, has very aptly said that “we are far too dependent on each other to ignore the impacts of our social, economic, and political interactions.” This means that we need more oversight, which creates a sense of responsibility towards each other. I think Space is the only vantage point that provides us this constant global oversight. If you think about the impacts of climate change — increase in forest fires or flooding events — these require very fast turnaround time on decisions, so that people can start doing something about it. To some degree, the demand for real-time intelligence has always been there but the capability has not.
The new and expanded constellations that are emerging today will provide constant global coverage and faster turnaround time. Different kinds of EO data are being made more readily available, for example SAR and Radio Frequency data, which are both independent of weather conditions and time of day. This dependability is critical for near real-time (NRT) monitoring. So, then you combine this additional capacity with real-time or near-permanent communications with satellites via laser communication systems in Space, or ground stations as a service from AWS and Azure. You start looking at things like increase in cloud computing infrastructure capabilities and the ability to automate analytics using Artificial Intelligence (AI). Now, all of a sudden, NRT intelligence is more possible and accessible because of these technological advancements. The demand for NRT intelligence is not limited to the defense and security sector; we are seeing an increasing interest in commercial applications such as monitoring of vegetation along critical infrastructure before and after storms. My view is that the demand is growing as the technological capabilities make the use cases realistic and achievable.
I believe that Artificial Intelligence and Machine Learning (ML) are the fundamental technical capabilities that will unlock the downstream impact of Earth Observation for the simple reason that there is just too much data for people to go through manually and figure out what’s going on. Only a very small percentage of that data is actually being used today. Machine Learning allows people to quickly process large amounts of data to detect objects and find anomalies, and all of this can be automated. We have also started to see Machine Learning-driven super resolution algorithms helping in upscaling existing imagery, so that other ML algorithms can perform more effectively. At the same time, I think we should be careful not to have the view that AI and ML can solve all problems effectively. These technologies have a huge role to play, but we need to make sure that we understand their role and not hype it up too much, because that may only lead to disappointment.
Within Earth Observation, we are going through the process of ramping up and trying Machine Learning for many applications and finding the boundaries of where it’s working well and where it still needs to be improved. Nevertheless, it’s a massive unblocker for the downstream adoption of Earth Observation.
We are building a marketplace and developer platform that aims to make it significantly easier, both technically and commercially, to use a wide range of EO data and analytics to deliver value to your customers. What we have at UP42, from a platform perspective, is a one-stop shop for data and analytics; a very simple pay-as-you-go model where you buy credits upfront. And as you use different data sources, algorithms, computing infrastructure, etc., you burn down your credits. We try to make sure that our platform is very actionable. We see marketplaces or platforms providing information about a data source or algorithm, but they don’t allow you to get access to that data source immediately via API.
So, if you are a solutions provider and are trying to deliver solutions to an end customer, you really need this programmatic. You want to take humans out of the loop as much as possible, and be able to automate that flow of data and analytics all the way through to your customer. This is a key part of our platform — making all data and analytics workflows accessible via API. The ability for you to build end-to-end analytics workflows using algorithms from our marketplace, or even uploading your own data and algorithms, and mixing it together with what we have on the marketplace allows you to solve your customer problems using existing building blocks from some of the best geospatial companies in the world, rather than building everything yourself from scratch.
The European Union has done an excellent job in supporting the Space industry in general. A great example of that is the Copernicus Program and all the satellites that the European Space Agency (ESA) has funded, launched, and made the data available to everybody. The economic value that has been generated on top of that data by companies that are building solutions or providing consulting services is enormous. So, a large part of the EO industry has been built on this decision. In addition to that, ESA has put a lot of programs in place to fund business initiatives, thereby helping the market develop.
What’s really exciting is not necessarily what has already happened, but what’s happening now. To start with, you have an ambitious expansion of the Copernicus Program. In addition, the European Commission has just agreed on the Space Entrepreneurship Initiative under the umbrella of Horizon Europe. The program, called CASSINI, is intended to support the entire emerging Space industry from 2021 to 2027 through a series of measures including €1 billion seed and growth fund for NewSpace startups in Europe. On top of a good foundation, they are now taking another step forward and I think we are going to see some very interesting market developments over the next couple of years supported by public policies.
If developed properly, geospatial knowledge infrastructure can help in coordination…
Automation in construction has long since moved from experimental to expected. But what progress has been made, and where can benefits be most readily realized moving forward?…