I guess you know as well as I do that vast opportunities await in space right now. LEO (low earth orbit) connectivity is gaining momentum, propelled by reusable infrastructure solutions that have brought the launch cost of a satellite down to somewhere between one hundred thousand and a million dollars. The new layers of LEO connectivity promise a tantalizing array of public and enterprise services – but one big question looms large. How can we overcome the remaining challenges and seize maximum value from the sector?
5G’s future is hybrid – the non-terrestrial opportunity
In this article, I’m going to explore the three key emerging areas that I think represent high value for ambitious innovators in the space ecosystem. They are the provision of enhanced connectivity, the collection of novel data and the analysis of novel data sources. I’ll unpack the challenges associated with each and identify the key technologies that have the capabilities to overcome them.
But first, let’s set the scene in this new world of satellite launch economics, where you can even book your SpaceX rocket launch online. Once in low earth orbit, satellite connectivity has the potential to deliver fast, low latency connections anywhere on the globe. The opportunities are endless: from providing highly reliable broadband for truly remote working, to collecting large amounts of data from the earth’s most out-of-the-way places.
These new horizons are also allowing a range of start-ups to launch satellites into LEO for a range of use cases. Seraphim Capital, a VC led accelerator program focused on space, recently went public with its space investment fund that backs start-ups. Companies within the fund are building solutions across a range of fascinating areas – including the collection of new types of data and novel analytics solutions for data from space.
Providing enhanced connectivity
A number of organizations are jostling to get ahead with the provision of enhanced connectivity. SpaceX and OneWeb both have established plans, but Amazon and Telesat are also entering the scene, the latter having recently secured Canadian government funding. While much of the focus of new LEO constellations is on providing broadband connectivity to rural areas, a significant amount of the potential value is likely to accrue to connectivity solutions which transport valuable commercial data. This will mean that LEO constellations will need to provide different connectivity characteristics depending on the use case.
The changes in launch infrastructure have transformed the economics of creating LEO broadband constellations, making these new connection types possible. Previously, operators were limited in the number of satellites they could launch due to the high costs. Now, they can launch significantly more satellites and – crucially – replace satellites more frequently. Low-cost satellite replacement is a powerful lever for operators to use when designing new networks.
The ability to regularly update satellite hardware, by deorbiting satellites and launching new ones, means that the system can be constantly updated in the constellation, in addition to updating the ground-based gateways and terminal devices. This allows system level improvements to take place as technology advances, allowing systems to be optimized for particular connection types.
Nevertheless, operators still face challenges in terms of cost, size, weight and power (CSWaP) constraints of terminal hardware. Currently, the terminal manufacturing costs for SpaceX’s Starlink constellation are high and ultimately passed on to consumers in a one-off charge for hardware and ongoing subscriptions. But what if terminals can be optimized to the size of a cellphone or small sensor, while still being able to connect to low earth orbit at a suitable cost point? This represents a significant opportunity for the satellite industry.
The key challenge is keeping the power level in the device suitably low, to ensure it is both safe for the user and can run on battery power for a suitable amount of time, but also powerful enough to connect to a satellite. Enhanced beamforming, as deployed in the antenna we recently designed to deliver 5G from a high-altitude platform in the stratosphere, is a potential solution to the problem.
New antenna technologies that can transmit hundreds of beams simultaneously offer the opportunity to create very narrow beams which focus the available power in a specific direction. This can allow a connection to be optimized for a particular use case without greatly increasing the overall power level of the solution.
Additional challenges for LEO satellite connectivity include tracking moving satellites and correcting for the corresponding Doppler shift in RF. Unlike geo-stationary satellites, which are fixed relative to a point on earth, LEO satellites move very quickly relative to ground-based terminals and will regularly need to handover connections between satellites and user terminals due to the rapid movement. This creates significant challenges when attempting to offer reliable low latency communications.
Constellation operators need to design the overall system in a way that creates good CSWaP trade-offs while providing reliable low latency communication links. This challenge is made even more difficult when mobile terminals have movement at both ends of the communication link. Innovation in terminal antennas may be the answer here. Small, flat panel electronically steered antennas can compensate for terminal pitch and roll, allowing terminals to track the satellites even in extreme conditions. To enable this, a combination of sophisticated RF and digital signal processing techniques, which go beyond what is currently commercially available, will be required – both active areas of R&D that we are engaged in.
Collecting novel data
In the same way that Starlink and the other mega constellations are poised to revolutionize communications, a new generation of earth observation satellites are benefitting from the lower launch costs. As new infrastructure allows an increasing number of satellites to be launched, companies are starting to consider what data can be collected from this entirely new perspective of planet earth. A significant amount of the new start-up activity is looking at what new data can be collected and what insights it can offer.
For example, Satellite Vu, a company in Seraphim’s portfolio, is providing monitoring services using high resolution thermal imagery. This is being used to monitor assets such as building efficiency and oil pipeline flows. This type of sensing would not be possible without the new perspective provided by satellites.
But significant innovation is still required in sensing technologies and connectivity solutions to ensure that data is collected at a suitable resolution and also can be transmitted in a useful timeframe (i.e., does the data need to be close to real-time?) This will require strong integration with the connectivity layer to ensure connections are appropriate for the use case. This then becomes a system engineering problem requiring expertise across a number of different areas to get right.
We’ve seen through our own work in developing telecom and satellite systems that latency is not just impacted by the distance the data has to travel. The fundamentals of the transport layer including data packet frame size, data integrity techniques and of course security protocols all come into play. Intelligent data compression and optimization techniques, along with data processing at the edge, are all innovation opportunities that can deliver enhanced system performance.
Analyzing novel data sources
As is always the case, data is only as valuable as the analysis that is applied to it and the insights that are drawn. So, we can expect artificial intelligence (AI) and machine learning (ML) technologies to become increasingly valuable in the new space ecosystem. As companies use satellites to collect more and more data from the unique vantage point of space, there is an increased opportunity to drive new and enhanced insights. Using AI and ML to fuse outputs from a variety of sensors from different satellites has the potential to offer significant value for a range of different use cases. For example, ChAI offers a commodity price forecasting service by applying AI to data such as satellite imagery alongside other data inputs. Pixxel is also proposing to use hyperspectral imaging to unearth problems which it claims current satellites are unable to address, its data platform offers near real-time insights based on the data.
Additionally, Microsoft has recently partnered with SpaceX to connect the Starlink constellation to Azure’s new Modular Datacenter to allow data to be processed. The plan is to extend this to connect to Azure edge devices, which highlights the shift towards increasingly intelligent processing taking place at the edge on smaller, less expensive devices. Understanding what data is available and then creating algorithms that can draw insights from a range of different data sets is a key challenge. But it is one that promises to offer significant value to companies which are able to collect and analyze new data.
Again, this will require a system engineering perspective across data collection to data transport and data analysis. Consideration to how the space network interacts with other network types and where data processing takes place are all key decisions for new emerging architectures which leverage the increased access to space.
The space opportunity is significant and extremely exciting. It is clear that the value of the sector will grow strongly as companies continue to evolve and leverage the new infrastructure layers to create new services and draw insights from data. It remains to be seen who will innovate quickest to derive the most value… but the race is on. Please don’t hesitate to reach out to me if you’d like to hear more about our research and discuss the technologies that are emerging. It’ll be great to continue the conversation.