A four-year-old initiative aimed at achieving communications resilience is moving slowly due to technical bottlenecks, high costs, and geopolitical factors.
BY MATTHEW FULCO
In August, Taiwan attempted a maiden flight of its first domestically developed satellite launch vehicle, in what was set to be a major milestone for the country’s nascent space industry. Shortly after liftoff, however, a steering system glitch caused the uncrewed vehicle to veer off its planned sea trajectory toward inland mountainous terrain. Operators quickly activated the emergency onboard self-destruct mechanism, then watched the rocket crash near the launch facility in southern Taiwan’s Pingtung County.
This result might not come as a surprise. Orbital launch attempts for brand-new rocket designs have a roughly 50% chance of failure during their earliest test flights before engineers iron out design flaws, historical launch tracking data show.
Yet the abortive launch highlights the challenges Taiwan faces in its ambitious bid to build a low-Earth orbit (LEO) satellite constellation for national communications resilience. Initiated by the former Tsai Ing-wen administration in December 2022, the project fuses longstanding plans to develop a space industry with a broader defense buildup linked to rising military pressure from China.
LEO satellites sit close to Earth — typically between 160 and 2,000 kilometers above the surface. They have fast speeds, minimal signal latency, and small coverage footprints. To build global internet networks with these satellites, large constellations are required.
Russia’s full-scale invasion of Ukraine in February 2022 catalyzed Taiwan’s initiative. Russia swiftly knocked out Ukraine’s telecommunications infrastructure at the war’s onset. But Kyiv overcame that loss by using SpaceX’s Starlink constellation. It was able to maintain battlefield command and control, deploy drones, and keep the civilian government connected to the outside world.
Taiwan currently relies on a network of 24 undersea cables for around 95% of its global internet and telecommunications connectivity, something it sees as a major vulnerability. If China severed those cables — a possibility in a cross-Strait conflict — Taiwan would be isolated from the global grid.
Deploying an LEO satellite constellation offers Taiwan significant benefits besides resilience in an armed conflict. It can fortify civilian networks against natural disasters such as earthquakes and typhoons that disrupt traditional terrestrial cables. Additionally, this robust digital infrastructure can accelerate the rollout of smart maritime and logistics operations.
The clock is ticking
In September last year, Taiwan Space Agency (TASA) Director General Wu Jong-shinn told the Taipei Times that “the clock is ticking” for Taiwan to build its own LEO satellite constellation. At the time, it had zero LEO satellites in orbit. “We need to speed up,” Wu said.
More than a year later, Taiwan still does not have any operational, non-experimental LEO communications satellites of its own.
“LEO constellations are very hard to make work from a technical and economic perspective,” Seoul-based space industry consultant Blaine Curcio tells TOPICS. “Taiwan is not alone in that challenge.” South Korea, for instance, is facing challenges building its “K-LEO” constellation.

Traveling through the skies at speeds of more than 27,000 km/hour, LEO constellations are arduous to manage. The satellites face constant atmospheric drag and have to use onboard thrusters to maintain their orbits. Otherwise, they will fall down to Earth and burn up.
Their extreme speed also means a user on the ground only has a line of sight to a single satellite for a few minutes. That requires the network to execute fast, perfectly timed “handoffs” that pass the internet connection to the next satellite without dropping the signal. To make this work on a global scale, the satellites must beam data to each other with high-precision lasers across great distances. Autonomous software must adjust their paths continuously to prevent high-speed collisions with space debris.
Upfront costs are massive. Billions of dollars are needed to finance large fleets before the service even launches. The satellites themselves are expensive but have short five-year lifespans, meaning companies must constantly build and launch replacements just to keep the network running.
“It’s a really big ask to say, we’re going to pull the trigger on a LEO constellation today and we expect it to be relevant in five years’ time,” Curcio says.
A crucial factor affecting Taiwan’s space program is its lack of indigenous launch capabilities. International experience shows that satellite technology and industry capabilities are built through continuous trial-and-error and in-orbit verification, and launch access is one of the key enablers, says Tzeng Chiau-ling, a senior industry analyst at the semigovernmental Market Intelligence & Consulting Institute (MIC) in Taipei. “If Taiwan lacks domestic launch capability or cannot secure enough verification opportunities, commercial timelines and product iteration will be delayed, slowing the industry’s development,” she says.
Tzeng notes that commercial launch services today are concentrated among a small number of providers. Available slots are expected to tighten in the near term as more satellite deployment programs come online. Reuters has reported that SpaceX Falcon 9 missions are already booked into 2028 and 2029.
The SpaceX factor
Elon Musk-owned SpaceX has loomed large over Taiwan’s satellite constellation plan since its inception. SpaceX’s preeminence in satellite communications — it possesses over 11,000 LEO satellites, more than all of its competitors combined — presents Taiwan with a multifaceted challenge. On the one hand, the US$2 trillion company is widely viewed as the best in the business. It boasts not just huge operational scale, but also vertical integration with SpaceX’s rocket launches.
Yet Musk has been ambivalent about entering the Taiwan market. Taiwan’s telecommunications regulations limit foreign majority ownership, typically a precondition for SpaceX to provide its services.
At the same time, Musk’s close ties to China, where Tesla has a Gigafactory and sells more than 650,000 cars a year, raise security concerns in Taiwan. In both 2022 and 2023, the outspoken mogul made public comments supportive of the ruling Chinese Communist Party’s claims over Taiwan — claims rejected by the overwhelming majority of Taiwanese people.
The ensuing fallout prompted Taiwan to explore alternative suppliers to SpaceX. That process remains ongoing. Yet the sense of urgency around the satellite constellation project has prompted Taiwan to reconsider the merits of partnering with SpaceX.
In July, Taiwan’s legislature passed amendments to the Telecommunications Management Act, easing foreign ownership and chairperson nationality restrictions for LEO satellite operators. Under the revised law, regulators can now waive the previous direct (49%) and indirect (60%) foreign ownership limits for satellite internet applicants. The strict requirement that a network operator’s chairperson must hold Taiwanese nationality can be exempted.
“Our goal is to have diversified satellite services,” opposition Kuomintang (KMT) Legislator Ko Ju-chun said when the bill was passed.
“I think ‘paving a way for SpaceX’ was the primary objective of the recent legislative activity,” says Rupert Hammond-Chambers, president of the U.S.-Taiwan Business Council (USTBC). “The to-date patchwork-quilt approach to providing Taiwan with an LEO constellation has left the island scrambling to provide a comprehensive solution for its national security interests as well as its businesses and citizens.”
Following the changes to Taiwan’s telecom law, Chunghwa Telecom is seeking to offer Starlink services. Chung-hwa Chairman Alex C.C. Chien says that his company is open to becoming Starlink’s “exclusive agent” or forming a joint venture. “We’re not ruling anything out. It depends on the other party,” he says.
Chien recommends that Starlink establish a local company and ground stations in Taiwan and that Musk’s company cooperate with local telecom operators to handle issues regarding subscriber numbers and frequency allocation. Citing the phased rollout approach that Starlink adopted in Japan and South Korea — where it partners with local telecom giants — he says, “It would be better if Taiwan could follow suit.”

As Taiwan’s largest telecom provider, Chunghwa Telecom wants to improve LEO satellite connectivity for its enterprise customers, such as airlines, shipping companies, and offshore wind farms, David Wang, an analyst at the Taipei-based market research firm TrendForce, tells TOPICS. “Compared with Eutelsat OneWeb [a European Starlink competitor that operates in Taiwan], Starlink has deployed a substantially larger number of satellites and therefore continues to hold an advantage in LEO network performance,” he says.
However, gaining access to Starlink services would not directly advance Taiwan’s goal of building its own satellite constellation, Wang notes. “Chunghwa Telecom would essentially lease satellite capacity from Starlink and obtain the right to use its communications services, rather than ownership of the constellation itself,” he says.
Mixing and matching
Determined not to rely on any single dominant satellite communications network, Taiwan has begun building out capabilities with different providers. Chunghwa Telecom, for instance, combines multiple satellite orbits. It partners with Eutelsat for LEO services and Luxembourg-based SES for medium-Earth orbit (MEO) services.
MEO differs from LEO in important ways. Because it operates at higher altitudes ranging from 2,000 to 35,000 kilometers, it allows a smaller number of satellites to cover much wider areas with moderate latency. That makes MEO the standard orbit for global navigation and positioning systems like GPS.
Chunghwa Telecom’s partnerships are intended to complement its own ST-2 geostationary [GEO satellite], creating a dedicated multi-orbit satellite communications offering, Wang says.
This multi-orbit architecture allows Chunghwa Telecom to provide customized services for enterprise clients. For example, it could work with airlines such as EVA Air and China Airlines to provide low-latency, 24/7 in-flight Wi-Fi. When an LEO satellite handover temporarily interrupts connectivity, broader MEO coverage could help maintain satellite links and improve service continuity, he says.
The same model could be applied to global maritime customers, Wang says. Historically, shipping companies relying primarily on GEO satellites have been limited to basic communications like text messaging and low-speed email. Chunghwa Telecom could support real-time vessel monitoring by incorporating low-latency LEO connectivity into a multi-orbit service, including alerts when onboard equipment deviates significantly from factory specifications. It could also provide crews with more bandwidth-intensive services such as video streaming and extended voice calls.
Next steps
Taiwan is now building indigenous technology and in-orbit verification for its satellite constellation. Six satellites are planned. The first two, focused on communications payloads, launch in 2028. The other four, aimed at building local capabilities in satellite system integration and manufacturing, will follow in 2030.
“A larger indigenous constellation is likely still under planning and discussion,” MIC’s Tzeng says.
Trendforce’s Wang notes that Taiwanese satellite component suppliers are competitive in areas such as flexible delivery schedules and high manufacturing yields for individual components. Yet they still struggle with integration of different satellite subsystems, including signal interfaces, inter-subsystem interference management, and thermal interface design.
He recommends that TASA work with Taiwanese system integrators such as Compal, Wistron, and Foxconn to establish modular satellite production lines. “Replicating and scaling this modular manufacturing model would increase satellite production capacity and help accelerate deployment of a Taiwan-owned constellation,” he says.
Space industry consultant Curcio says that Taiwan could potentially collaborate with 4-6 other countries in the region to build a pan-Asia LEO satellite constellation, referring to a suggestion TASA head Wu made in April at the Space Symposium in Colorado Springs.
“Behind closed doors in Asia there are officials saying that it might not make sense to build our own fully domestically built constellation, that it might be better to work with these allies that we really trust,” he says.
Auguring well for Taiwan’s satellite ambitions is the confirmation that Amazon Leo (formerly Project Kuiper), the Seattle-based tech giant Amazon’s LEO satellite internet network, will soon be made available in Taiwan. Talks between the Taiwanese government and Amazon have been ongoing since December 2024, but it was only in May that Taiwanese telecom operator Far EasTone announced that it had signed a distributor agreement with Amazon.
Amazon Leo’s high-speed broadband satellite services could become available in Taiwan as early as the first half of 2027, depending on the speed of the regulatory process. At a press conference in May, Far EasTone President Ching Chee said that approval will be needed from the Ministry of Digital Affairs to assign its radio frequency for the LEO satellite operations. Far EasTone is also required to obtain approvals from the National Communications Commission and Criminal Investigation Bureau to launch the satellite services.
As for the likelihood of SpaceX entering Taiwan’s market, it will require navigating complex political factors. “The recent reporting on Tesla’s consideration of compartmentalizing its China business to insulate Mr. Musk’s other businesses such as xAI and SpaceX from Chinese pressure is an important development,” USTBC’s Hammond-Chambers says, referring to a Wall Street Journal report published in late July.
Musk later dismissed the report as “absurdly fake news” in a post on his social media site X. “This has never come up in discussion ever,” he wrote.
And in another X post on August 11, Musk said, “China is awesome. I strongly encourage people to visit.” He was responding to an ARK Invest podcast that discussed whether Tesla’s large footprint in Shanghai could affect a rumored Tesla-SpaceX merger.
The SpaceX CEO has traveled to China publicly at least five times. The most recent visit came in May when he went to Beijing as part of U.S. President Donald Trump’s delegation, which also included then Apple CEO Tim Cook and Nvidia chief executive Jensen Huang.
“The primary concern in Taiwan remains that Musk’s comments on Taiwan being part of China sent a chill through the island’s political class,” Hammond-Chambers says.