Sceye’s 30-day roundtrip and SoftBank’s Japan trial establish meaningful flight, payload and telecom-integration milestones. They do not establish commercial scale, durable economics, spectrum readiness or a 6G network.
By Land Offset PAI
Sept. 12, 2026
TOKYO — Sceye’s Service Test 1, or ST1, completed a nearly 30,000-kilometer roundtrip between New Mexico and Japan, carrying a cellular payload that SoftBank used for a limited communications trial over Japan. The mission is a material engineering and integration result: an unmanned lighter-than-air platform crossed the Pacific, remained in a defined operating area, connected ordinary smartphones through SoftBank’s core network, and supported tests involving drones, emergency messaging and onboard computing. 2 3
The result is narrower than the language of a new global communications layer. It is a pre-commercial trial, not evidence of a commercial service with published capacity, coverage density, availability, price or spectrum-coordination performance. SoftBank says it will use the results to improve operations and communications quality with the aim of commercializing HAPS services in Japan from 2027 onward. 2
What ST1 demonstrated
Sceye launched ST1 from New Mexico on Aug. 9. It reached Japanese airspace after approximately 13 days and more than 15,000 kilometers of stratospheric flight. SoftBank said the platform then operated over a test area off Cape Muroto in Kochi Prefecture, chosen for disaster-response scenarios, and maintained position within a radius as small as five kilometers for an extended period. 2
The two companies report that a 4G LTE-equivalent base station on the platform connected through SoftBank’s core network to support voice calls, text messages, internet applications, video calls and video streaming on smartphones. The trial also tested earthquake and tsunami emergency alerts, an emergency maritime call, control and video transmission for a drone, and an onboard server. SoftBank reports an average 68-millisecond round-trip response time for the onboard-processing test, which it says was more than 40% lower than the internet-cloud comparison used in that test. 2
After the Japan test, ST1 returned to the United States. Sceye says the full mission lasted about 30 days, traveled nearly 30,000 kilometers and ended with a planned termination and descent near its New Mexico headquarters. Sceye also says the craft operated at roughly 16.5 to 17 kilometers altitude and in Japanese-managed airspace for more than seven days. Those latter endurance, altitude and time-in-airspace details come from Sceye’s release; SoftBank’s announcement independently confirms the cross-Pacific flight, extended operation around Muroto and the network trials. 3
| Claim | Public record supports | Public record does not establish |
|---|---|---|
| Flight endurance | ST1 completed a New Mexico–Japan–United States roundtrip over about 30 days. 3 | Months-long station keeping required for a sustained commercial network, fleet availability or maintenance cycles. |
| Phone connectivity | SoftBank tested calls, messages, data and video on smartphones through its core network. 2 | Capacity per user, total simultaneous users, cell-edge performance, handover behavior, service availability or retail economics. |
| Disaster use | The companies tested emergency alerts and an emergency call in a disaster-oriented test area. 2 | That a platform has met public-safety certification, can be launched on demand after a disaster, or can replace a damaged terrestrial network at scale. |
| Edge computing | An onboard-server test reported a 68-ms average round trip under the companies’ comparison. 2 | General latency or throughput superiority across a commercial network, or a demonstrated operational artificial-intelligence service. |
| Network scale | Sceye has designed its SceyeCELL payload for wide-area direct-to-device service. 6 | Sceye’s claim that one platform can cover an area equivalent to 500 terrestrial towers; no comparable delivered-coverage or capacity result was published from ST1. |
The crucial distinction: a network test is not a network business
This mission crosses several practical thresholds that had to be crossed eventually. The platform’s passage to Japan showed long-range flight endurance. The Muroto operation tested a combined air vehicle, radio payload, core-network connection and ground-device path. The direct-to-device design also matters because it aims to use ordinary cellular devices rather than specialized satellite terminals. 2 5
But a commercial system must answer different questions. It must show how many platforms are needed, how long each stays on station, how operators rotate and recover them, what capacity is available under peak demand, how service degrades in difficult weather and urban terrain, how the network manages interference, and how costs compare with terrestrial expansion and low-Earth-orbit satellite capacity. The available ST1 releases do not provide those measurements.
IEEE Spectrum made the same distinction before the mission. It reported that a viable HAPS network needs reliable links, stable throughput, interference management and eventually inter-platform coordination, while outside researchers said the aerospace portion still needed work. 5 A separate industry analysis identifies spectrum allocation, spectral efficiency, station keeping, operational staffing and cost as core commercialization hurdles. It also notes that Google’s Loon and Facebook’s Aquila did not establish durable commercial models, even though they advanced the underlying technology. 7
SoftBank’s own earlier test illustrates why measured results cannot be transferred casually from one platform to another. In 2025, the company flew a six-cell payload on a light aircraft at 3,000 meters and reported average downlink throughput of approximately 33 Mbps 15 kilometers from the aircraft’s circling path. That was a useful payload and coverage test, but it was not a Sceye platform flight at stratospheric altitude. 8 ST1’s public releases do not publish a corresponding throughput, user-count or coverage measurement.
“HAPS” remains a technology label, not a regulatory conclusion
The reporting terminology also deserves precision. Sceye and SoftBank call ST1 a high-altitude platform system, or HAPS. The International Telecommunication Union’s Radio Regulations backgrounder defines HAPS as radio stations on objects 20 to 50 kilometers above Earth at a specified nominal fixed point. Sceye says ST1 operated at approximately 16.5 to 17 kilometers during its mission. 1 3
That does not invalidate the test or settle the platform’s legal classification in Japanese airspace. It does mean that the product label should not be mistaken for proof that every spectrum, aviation and cross-border coordination condition for an ITU-defined commercial HAPS system has been resolved. The ITU identifies protection of incumbent satellite, aviation and weather services, as well as cross-border interference coordination, as central issues for HAPS-based mobile systems. 1
A milestone with four unresolved tests
ST1 moves the question from whether Sceye can connect a trial payload after a trans-Pacific deployment to whether it can produce a service. The next public evidence should answer four specific questions.
| Required next proof | Useful public measure |
|---|---|
| Persistent operations | Time on station across repeated vehicles, weather conditions, recovery cycles and maintenance intervals. |
| Commercial radio performance | Coverage footprint, usable throughput, concurrent devices, handover behavior, availability and interference results. |
| System economics | Platform, launch, operations and replacement costs compared with terrestrial buildout and satellite alternatives. |
| Regulatory readiness | The applicable airspace approvals, spectrum authorizations, coordination arrangements and public-safety operating conditions. |
The source LinkedIn post accurately identified a genuine technological milestone, but it treated the result as though it had already resolved the economic and operating questions of a “third layer” of global infrastructure. 4 It has not. What the public record supports is more valuable and more limited: a 30-day airship mission and integrated Japan telecom trial that make Sceye and SoftBank credible candidates to test the proposition further.
References
- International Telecommunication Union, “HAPS – High-altitude platform systems.” 1
- SoftBank Corp., “Achieves Japan’s First HAPS Trial Service in Preparation for Commercialization,” Sept. 2, 2026. 2
- Sceye, “Stratospheric Platform Re-enters United States After Record-Breaking Roundtrip Flight,” Sept. 9, 2026. 3
- Keith King, “The stratosphere is emerging as a third layer of global communications infrastructure,” LinkedIn, Sept. 2026. 4
- Rahul Rao, “Sceye Is Testing Out Its Stratospheric Cell Tower,” IEEE Spectrum, Mar. 28, 2026. 5
- Sceye, “Sceye Unveils SceyeCELL,” Feb. 23, 2026. 6
- Frank Rayal, “Navigating the Future of HAPS: Challenges, Investments, and Innovations,” June 23, 2025. 7
- SoftBank Corp., “Develops High-capacity 6-cell Capable HAPS Payload,” Sept. 18, 2025. 8