“Balloons are the future of connectivity.” What sounds like a line torn straight out of a Thomas Pynchon steampunk novel is actually the core mission statement of Loon, the Google/Alphabet subsidiary that wants to loft a fleet of balloons into the stratosphere to provide global broadband connectivity.
An image posted to Loon’s website shows a transmitter covered in solar panels dangling from a translucent hot air balloon. This is one example of what is technically referred to as a High-Altitude Platform Station, or HAPS, and a visual representation of what the future of connectivity, and even the future of commercial satellite service could look like.
“Terrestrial, stratospheric, and satellite technologies are all suited to different parts of the globe and use cases. Loon believes the problem of connectivity will be solved with an all-of-the-above approach,” says Dr. Brian Barritt, technical lead at Loon and a senior software engineer for Google. “We’re proud to be adding a new layer to the connectivity ecosystem.”
This “new layer” of connectivity caught the attention of one the largest global satellite operators in the world, Intelsat.
“HAPS technologies have been — no pun intended — floating around for 20 years or so, but mainly for military applications,” says Intelsat’s Vice President of Business Development Greg Ewart, who explains that advancements in composite materials, battery technology, and solar antennas have made HAPS more commercially viable to produce, similar to the evolution of electric automobiles.
But, the question that intrigues Intelsat is — Could balloons bring its satellite services down to Earth and closer to consumers than ever before?
Thanks to the development of Software Defined Networks (SDNs), which could act as polyglots translating multiple languages in a hybrid network, a HAPS platform could be seamlessly integrated into a network with a Geostationary (GEO) satellite. Ewart adds that this integration could all happen in a back office environment, out of sight and out of mind for customers who mostly don’t care about how satellites work and just want reliable service on the devices they know and love.
For years, Intelsat has waited for a financially viable path to connect directly to consumers, without the need of satellite-specialized hardware and an array of specialized sales teams on the ground. HAPS was the most promising opportunity Intelsat had seen to accomplish this goal in years, so it watched from a distance and waited for the right time to jump in.
“We've always felt also that there is no one technology that brings everyone together from a communication standpoint,” says Ewart. “We've never been scared as a company to push the envelope with respect to new frontiers of technology. Basically, it's in our DNA.”
The International Telecommunication Union’s (ITU) definition of a HAPS system is also pretty high up on the steampunk scale. According to the ITU, HAPS systems are “radio stations” located on an object in the stratosphere, at an altitude between 20 and 50 kilometers and at a specified, nominal, fixed point relative to the Earth. HAPS can resemble balloons or solar-powered aircraft – the key is to be lightweight and power efficient to ensure long-term, continuous operation, in order to provide fixed broadband connectivity in targeted areas.
Propelled by Alphabet’s infinite resources and support, Loon was one of the first to transform HAPS from concept to prototype. To date, Loon vehicles have logged over one million flight hours and travelled over forty million kilometers in the stratosphere.
“At any given time there are dozens of Loon flight vehicles aloft worldwide,” says Barritt. “Many of these are focused on serving our commercial partners, such as in Kenya, while others are used for continued research and development purposes. We launch our flight vehicles from two launch sites — one in Nevada and one in Puerto Rico. If needed, we can launch one flight vehicle every 30 minutes utilizing our custom-built automated launching systems.”
Like Intelsat, NSR Analyst Shiva Muruganandham has also been keeping an eye on the HAPS market. He has authored several forecasts for the technology’s future. The most recent one, the fourth edition of NSR’s High Altitude Platforms study, was released in early July.
Muruganandham believes that in addition to giving satellite operators the ability to bypass specialized satellite devices and go right to consumer devices, investing in HAPS could also give GEO satellite operators a strategic position to defend against the oncoming wave of competitive services, including Low-Earth Orbit (LEO) constellation services.
“Satellites cannot exist in a vacuum. Operators have to keep an eye on adjacent technology developments and understand the new business models they might have to deal with in the future,” says Muruganandham.
Ewart has read Muruganandham’s report and admits that being able to create a clear, competitive contrast to the flashy new LEO systems would influence whether or not the operator would invest in HAPS. He explains that HAPS technology reached a maturity point that flipped the switch for Intelsat about two years ago. It was finally cheap enough to produce, tested and proven, backed by influential tech companies along the entire supply chain, and fast on its way to market. The operator was starting to identify significant business opportunities for its cellular backhaul business through HAPS.
“Some of the market segments that we serve, like smart mining or agriculture, are fairly small or regional when compared to traditional telecom markets. However, they require complex and expensive networks to serve,” says Ewert. “For instance, the capital costs of deploying a global LEO network for these kinds of markets is very expensive. They require billions of dollars upfront and a lot of work from that point on to start connecting devices, generating revenues and servicing debt. HAPS, on the other hand, is showing disruptive advantages and would allow us to address these markets with surgical precision, at a fractional of the cost of a LEO network, with MNO (Mobile Network Operator)-like performance direct to MNO user devices and efficiently deploy capital, incrementally, based on market needs and success.”
HAPS and LEO systems do share similarities. They are both in constant motion relative to the Earth, which requires a high level of orchestration of steerable beam tasking, radio resource management, and network path planning. Barritt explains that this is the reason Loon is collaborating to some degree with LEO operators like Telesat.
“However, a big difference between HAPS and LEO is that Loon flight vehicles are much closer to the Earth, which enables lower latency and the ability to communicate seamlessly with the mobile devices many people already have. For example, an early test of the Loon service in Kenya showed latency of 19 milliseconds,” says Barritt.
By integrating HAPS into its network, Ewart says Intelsat can jump right into new markets, work with specific MNOs with specialized applications and incrementally extend their cell coverage network to meet the demand, and spend less time and money than competing operators.
“In those cases, the building costs and maintenance required for LEO satellite fleets, cellular towers, and fiber, is just too much based on the revenues that they would be able pull out of the low-density markets they serve,” says Ewart. “In addition to being cheaper, a HAPS network is much faster to deploy and can be broken down and reconfigured and/or refreshed within a matter of hours or days.”
Finally, Intelsat realized that HAPS technology could help the operator realize its dream of connecting directly to consumer MNO devices. “We think that's incredibly important,” says Ewart. “Traditionally, satellite service providers always connect to an expensive specialized device — a VSAT (Very Small Aperture Terminal), a satellite phone or device modem. Going directly to the MNO device allows us an inexpensive alternative to specialized devices. It opens up a much larger addressable market where we leverage the MNO’s eco structure of low cost equipment and market access. So, not only are we complimentary, we're also able to use HAPS to extend our network beyond what we thought a traditional satellite service could do.”
NSR’s Muruganandham says that regardless of how brazen the new business models for LEO satellites and fiber services are, operators like Intelsat do not have the luxury of being able to wait for competitors to fail. With business models still dependent on lagging or declining broadcast markets and specialized broadband services in an environment ravaged by COVID-19 and travel restrictions, GEO operators need to make smart, bold moves to stay relevant. “In Intelsat’s case, it seems the strategy is to see how they can work together in a HAPS ecosystem, and benefit from a shared mission,” he says.
Enter the HAPS Alliance – a consortium of telecommunications companies that formed in February 2020 to collaborate on an interoperable HAPS system that benefits all members. Intelsat is a founding member and, as of the end of July, still the only satellite operator on a team that includes Airbus, Nokia, AT&T, T-Mobile, Airtel, and Google Loon.
HAPSMobile, a subsidiary of SoftBank and minority owner AeroVironment, co-founded the HAPS Alliance to enlist the help of industry partners to overcome a make-or-break challenge that will determine whether or not HAPS lives up to its potential – the development of a standard for interoperability.
HAPSMobile Senior Vice President Ryuji Wakikawa, who also leads SoftBank’s Advanced Technology division, firmly believes that new standards and coordinated regulatory policies across industries are needed to make HAPS solutions a reality. “With a united voice, we believe we can greatly accelerate the adoption of HAPS,” says Wakikawa. “We believe it’s also important to demonstrate to partners how HAPS, as a non-terrestrial network, complements other existing network solutions, including terrestrial networks and satellites … SoftBank also recognizes that it’s essential to build a system with interoperability, backward compatibility and non-interference with existing networks.”
There are several dimensions to interoperability: end user equipment interoperability; core network operability; and the issue of interoperability between aerospace network equipment infrastructure developed by different manufacturers, Barritt explains.
“Loon’s network design already solves most interoperability problems; we partner with existing MNOs and integrate with their core network using standard 3GPP interfaces that are commonly used to facilitate roaming between MNOs,” says Barritt. “As a result, end users are able to use their existing handsets to communicate over our aerospace network. It’s so frictionless that end users can’t easily distinguish between connection to HAPS versus a terrestrial tower.”
In addition to pooling political and regulatory expertise and resources, participating companies in the HAPS Alliance will have a unique view (and voice) into the development of common product specifications in a hypothetical HAPS network. Ewart says that for Intelsat, having access to the members themselves as technology partners is also important, as it allows them to operate more independently outside of the satellite silo of tech providers.
Yet Ewart doesn’t see a significant role in the Alliance for more satellite ground system manufacturers. “What we're fundamentally doing in the Alliance is utilizing technology more associated with traditional Mobile Network Operators,” he says. “For our work on HAPS specifically, the specialized technologies that companies like iDirect have traditionally provided to the satellite industry would be overshadowed by Nokia. In this situation, there’s no comparison to Nokia’s depth and knowledge of what we need for HAPS platforms and the off-the-shelf devices that we’re connecting. The satellite-specific manufacturers would have to invest a ton of research and development capital in order to catch up to Nokia in this space.”
Nokia is indeed positioning itself to be the ground infrastructure and technology leader of the group. “With enhanced LTE wireless networking now, and 5G connectivity next, we can use HAPS platforms to deliver major economic and societal benefits to underserved communities, particularly in remote and challenging environments,” Nokia Enterprise President Kathrin Buvac writes in the HAPS Alliance blog.
Barritt says that Loon and the HAPS Alliance share the same long-term vision — establishing a connectivity service that eliminates the boundaries between terrestrial, HAPS, and satellite systems. The consumer won’t notice a thing. “Working towards that goal, we’re participating in the Integrated Access and Backhaul (IAB) and the Non-Terrestrial Networking (NTN) standardization efforts in the 3GPP as part of 5G Release 17+,” says Barritt. “We’ve also been developing Temporospatial Software Defined Networking (TS-SDN) infrastructure to facilitate the control planes of these future hybrid networks.”
At this point, the HAPS Alliance exists mostly as a show of force, as it is still a nascent technology. Also, the Alliance isn’t the only player in the game. There are more than 40 HAPS programs at various stages of development around the world, each targeting unique markets with unique applications and facing unique challenges.
Muruganandham believes that the ceiling is very high for HAPS, “but there is still largely under-developed chemistry between the Alliance members and still quite a lot of seen and unforeseen technological challenges to overcome,” he says. “Controlling the balloons is a challenge in itself. HAPS developers are still working out how to overcome shutter speed issues and navigating Ekman wind patterns in the regions that they want to serve and need persistent flight capability.”
HAPS technology is evolving quickly, but there is consensus among alliance members that they will need better performance out of the existing batteries. The current balloon and lightweight aircraft vehicles need solid structural integrity. Unlike satellites that burn up in the atmosphere, balloons and airplanes can crash to the ground, which creates potential safety hazards.
Barritt’s Loon colleague Ken Riordan, head of product management, acknowledges the significant challenges ahead for HAPS, but adds that they can be easily overcome by combining the efforts of companies that have overcome much greater technical challenges in the past. “We recognize that for HAPS technology to reach its full potential, we will need the coordinated effort of entire industries, regulatory authorities and governments,” he says. “While HAPS is very familiar to us at Loon, we also understand that for large segments of our industries, HAPS is not yet well understood. For this reason, one of [the Alliance’s] top priorities is to educate.”
In the latest edition of NSR’s High Altitude Platforms report, Muruganandham predicts that HAPS services from balloons and pseudo-satellite platforms will generate about $4 billion in cumulative revenues by 2030. Most of this business, he says, will happen in the Americas, between a mix of commercial and government customers.
“This is obviously still a very young market, but it has been picking up a significant amount of steam since the last report we compiled,” says Muruganandham. “In the commercial markets specifically, we definitely see communications and remote sensing applications as the major drivers throughout the initial years. That said, most of the business for HAPS is still coming mostly from the scientific community, at least in the near term. But I think it comes down to seeing if these alliances actually do end up boosting innovation in the aviation and telecom sector and spearheading developments in the industry. Solutions born out of these correlations, which are uniquely suitable for HAPS deployments, will definitely be out in the market in the long term. I do think it will take quite some time for the commercial market to develop.”
Following a general dip in activity due to the COVID-19 epidemic, Muruganandham says the HAPS market will remain in recovery mode for most of the decade, with accelerated growth coming towards the end, with in-service units averaging at 8.7 percent Compound Annual Growth Rate (CAGR) through ten years.
Intelsat’s Ewart disagrees with Muruganandham’s assessment, and believes a commercial market for HAPS is closer than ten years out.
“I read the NSR report, and while I respect and value their expertise, I think they need to look a little harder at the service side of the revenues,” Ewart says. “I think the report is heavily weighted toward the manufacturing side of the platforms. The MNO marketplace is pretty much untapped right now. It’s the kind of greenfield we need to go after.” VS