> ## Content Index > Fetch the complete content index at: https://nexi.fund/llms.txt > Use this file to discover other available public pages before exploring further. # Sceye's Trans-Pacific Flight Puts the Stratosphere on the Network Map - URL: https://nexi.fund/stratospheric-haps-network-layer-2026/ - Published: 2026-09-11T18:30:05.000Z - Updated: 2026-09-14T14:20:30.000Z - Description: Sceye's ST1 airship flew 15,000 km from New Mexico to Japan in 13 days, then handed live mobile broadband to ordinary phones through SoftBank's core network. It marks the shift from experiment to network infrastructure, and raises the question of who certifies it. - Author: Nexi.fund Labs - Tags: Space & Expansion, #mode-1, #hook-number, #track-F 15,000 kilometers. Thirteen days. One airship, drifting from a desert in New Mexico to the airspace over Japan. Sceye, the company that built it, calls the machine a High-Altitude Platform System (HAPS): an uncrewed vehicle that parks in the stratosphere, about 20 kilometers up, and does the work of a cell tower without the tower. The craft left New Mexico on August 9 and reached Japan thirteen days later, then handed live mobile broadband to ordinary phones through SoftBank's core network. No new handset. No dish. No ground station in view. The distance is the easy headline. The harder fact is that a platform floating on helium behaved like ordinary network equipment. ๐ŸŽฏ HAPS has crossed from test flight to network integration. The hard problem is no longer altitude; it is behaving like a base station inside a carrier's core network. The economics work as a complement to satellites and towers: a persistent layer for coverage gaps, disaster response, and capacity densification, not a replacement for low Earth orbit (LEO) constellations. The first buyers are carriers and governments. SoftBank's exclusive Japan rights and AALTO's certification path matter more than any endurance record. Sceye and SoftBank announced the result on September 2\. It was the first mission of its Service Test program, and the first time the company flew a platform to Asia. 15,000 km trans-Pacific #### Sceye ST1 stratospheric crossing The ST1 airship flew the distance in 13 days before delivering mobile broadband over Japan. ยท *Sceye / SoftBank, September 2026* ## The airship that behaves like a cell tower The company builds a lighter-than-air (LTA) platform: a helium airship roughly 65 meters long that gains altitude from buoyancy and holds position with electric propulsion. Solar cells and batteries carry it through the night, which is the part that took years to prove. In 2024, Sceye became the first operator to close the power loop in the stratosphere, generating and storing enough energy to stay powered and on station through a full day-night cycle. In April 2026, its SE2 airship flew more than 6,400 miles from New Mexico to the coast of Brazil and spent 88 hours holding over a single area. ST1 put that endurance to work on a network. Over Japan, the platform carried text messages, voice calls, internet access, and video streaming using SceyeCELL, the company's stratospheric communications payload. It served unmodified mobile devices, ran edge computing onboard instead of backhauling every request to the ground, and relayed traffic with drones flying below it. The company says a full-scale platform covers an area equivalent to about 500 terrestrial towers. \~500 towers per platform #### Coverage of one full-scale HAPS Sceye's stated coverage equivalent for a full-scale stratospheric platform. ยท *Sceye, September 2026* ## Two cost curves, one coverage gap The useful comparison is HAPS against the coverage gap carriers already know how to price. A LEO satellite passes overhead for a few minutes at a time. A HAPS holds one region for weeks or months, close enough to the ground for high-capacity, low-latency links to handsets that need no modification. The carrier drew the distinction when it launched Starlink Direct in April 2026 for low- and medium-capacity coverage in mountains, remote islands, and offshore waters. HAPS aims at a different job: dense, persistent capacity over a defined footprint. An airship does not burn most of its mass reaching altitude the way a rocket does. It still has to be built, fueled, and maintained, and it serves one region at a time. The pitch to a carrier: coverage that cannot justify a new tower, sold for less than the tower would cost. As we wrote in [September](https://nexi.fund/satellite-manufacturing-race-2026/), satellite manufacturing is racing to catch constellation demand. The stratosphere is the lower, slower, and less crowded answer to a subset of the same problem. > "Flying from the US to Japan demonstrates the performance required to make the stratosphere a viable layer of infrastructure and realize the future of AI, edge computing, and 6G."โ€” Mikkel Vestergaard Frandsen, Founder and CEO, Sceye ## What still has to be certified The sequence has not been a secret. The carrier invested in the Series C round in June 2025, secured exclusive rights to provide HAPS services in Japan, and is buying the pre-commercial flight. Commercial service, not another demonstration, is the milestone the money is waiting on. Certification is the gate. AALTO, the Airbus subsidiary that owns the Zephyr platform, is working with the UK Civil Aviation Authority (CAA) toward full certification in late 2027\. Zephyr is the endurance leader: a solar-electric fixed-wing aircraft with a 25-meter wingspan that weighs about 75 kilograms, carries up to 8 kilograms of payload, and holds the record for the longest unrefueled flight at more than 64 continuous days. AALTO is targeting missions of 150 days. Chief executive Hughes Boulnois frames the value as "the persistence of a satellite with the flexibility of a drone." The constraints are real. Payload mass is measured in single-digit kilograms, so the sensor or radio you fly stays small. Station-keeping through winter, weather, and crowded airspace needs coordination across jurisdictions. A platform that loiters over one region serves that region, not a global market. None of that is fatal. All of it shapes who buys first. ### Will the stratosphere carry real traffic before it carries real revenue? ๐Ÿ”ฎ **By late 2027, at least one HAPS operator will hold civil aviation certification for commercial service, but recurring telecom revenue from stratospheric platforms will stay under $500 million a year through 2028.** Probability: 60% โ€” the certification path is in motion, while carrier procurement cycles and single-region coverage keep near-term revenue small. #### โœ… Arguments for A carrier has already paid for the flight, the equity, and the exclusive rights โ€” the commercial intent exists before certification. AALTO's payload campaign and UK CAA work give the category a concrete 2027 regulatory date rather than an open-ended promise. **Confirmation criteria:** a named carrier signs a multi-year capacity contract, not a single test. #### โŒ Arguments against A platform that serves one region at a time cannot scale like a constellation; revenue per unit stays capped by geography. Payload limits in the single-digit kilograms keep HAPS out of the heaviest sensing and communications missions. **Disconfirmation criteria:** certification slips past 2027, or the first contracts are disaster-response grants rather than commercial capacity. ๐Ÿ“Š **Key signals to track** A civil aviation certificate issued to a HAPS operator โ€” the first proof that regulators, not engineers, set the pace. A carrier signing a capacity contract measured in years rather than a one-off demonstration flight. AALTO's 2026 payload integrations turning into named commercial customers. Any platform reaching the 150-day endurance target, which would move the technology from seasonal to always-on. ### Development scenarios #### ๐ŸŸข Optimistic scenario (25%) Certification lands on schedule and two carriers sign multi-year contracts. HAPS becomes the default answer for rural and maritime coverage where towers never pencil out. **Implications:** platform builders move from grant-funded demos to recurring revenue, and a supply chain for lightweight payloads forms around them. #### ๐ŸŸก Base-case scenario (55%) Certification arrives in 2027, but the first customers are governments and disaster-response agencies. Commercial carriers buy capacity more slowly than the demonstrations suggest. **Implications:** the technology is proven and funded, yet the revenue curve lags the flight record by years. #### ๐Ÿ”ด Pessimistic scenario (20%) Airspace rules and station-keeping costs keep HAPS confined to a handful of government routes, while cheaper satellite direct-to-device service absorbs the commercial case. **Implications:** the platforms keep flying, but the business settles into a niche of surveillance and emergency coverage. ## Where this story comes from [ SoftBank Corp. Achieves Japan's First HAPS Trial Service SoftBank and its partner report the first trial HAPS service in Japan using a lighter-than-air platform, the primary account of the September 2026 connectivity demonstration. SoftBank Corp. ](https://www.softbank.jp/en/corp/news/press/sbkk/2026/20260902%5F01?ref=nexi.fund) Primary source for the flight, the network integration, and SoftBank's commercial intent. [ Sceye and SoftBank Test HAPS Connectivity and Edge Processing Over Japan Independent analysis of what the demonstration proved: non-terrestrial coverage is extending beyond satellites into a lower, persistent layer. IoT Business News ](https://iotbusinessnews.com/2026/09/03/sceye-and-softbank-test-haps-connectivity-and-edge-processing-over-japan?ref=nexi.fund) Useful for the network-architecture framing and why integration, not altitude, is the hard part. [ AALTO prepares Zephyr HAPS for commercial operations AALTO's Zephyr, the solar-electric HAPS with the longest endurance record, pushes toward UK CAA certification and a 150-day mission target. Aerospace Global News ](https://aerospaceglobalnews.com/video/aalto-zephyr-haps-commercial-operations?ref=nexi.fund) Anchors the regulatory timeline and the second credible operator in the category.