2026-04-03 · 7 min read
The multi-bearer question: why choosing VSAT or Starlink is the wrong frame
The maritime connectivity market spent three years debating VSAT versus Starlink. Meanwhile, every serious operator quietly built a stack that uses both.
This is not a story about technology convergence. It is a story about failure modes. The operators who moved fastest to multi-bearer were not chasing throughput — they were responding to specific events that made the cost of single-bearer dependence visible.
What single-bearer dependence actually costs
On 24 July 2025, a software fault in Starlink's core routing infrastructure triggered a 2.5-hour global outage. An estimated 75,000 vessels lost their primary internet connection simultaneously. Documented effects included loss of emergency monitoring access, fleet communications, telemedicine services, and weather forecast feeds. The vessels that noticed were the ones with nothing else to switch to.
Four months earlier, a hacktivist group claimed to have compromised Fanava Group, an Iranian VSAT service provider. The claim, partially corroborated by AIS tracking data showing Iranian fleet vessels going dark in the affected period, described 116 ships across two major shipping companies losing communications simultaneously — AIS dark, voice down, ship-to-shore offline. Not because their terminals failed. Because their provider's infrastructure did.
These are not outliers. They are illustrations of a structural property: a single bearer, however reliable under normal conditions, concentrates risk at the provider level. And provider-level failures do not affect vessels one at a time.
Each bearer has a specific failure mode. The ship operates in all of them simultaneously.
The coverage map and the operating map are not the same
Starlink's coverage map looks close to global. Its operating permissions do not.
In December 2025, Chinese maritime authorities issued the first documented penalty to a foreign vessel for operating Starlink in territorial waters. Enforcement is expected to intensify. China is not an edge case in any fleet manager's route planning. Add the UAE, Taiwan, South Africa, and Saudi Arabia, where Starlink maritime operation is similarly restricted, and the coverage gap is not a firmware problem. No satellite constellation has ever been immune to the political map.
VSAT faces its own version of the same problem. Beam contention is worst in exactly the waters where traffic is heaviest — the approaches to Rotterdam, the English Channel, the Straits of Malacca. The vessel is present. The satellite is overhead. The capacity is not there.
What operators who have solved this actually run
SIEM Shipping's ro-ro fleet runs the clearest published version of where the industry has arrived: two Starlink terminals for primary offshore bandwidth, two 5G antennas activating automatically within 85 kilometres of shore, LTE for tertiary redundancy, and a single Iridium Certus terminal as the final fallback — covering any geography, any regulatory environment, any outage.
Maersk's "One SATCOM" programme across approximately 550 vessels uses SD-WAN to load-balance across Ku-band VSAT, Ka-band Fleet Xpress, Starlink LEO, and 4G simultaneously. The stated uptime target for cloud application support is 99.5 percent. That figure is not achievable on any single bearer. It is a consequence of the architecture.
By May 2024, Marlink had deployed 3,200 hybrid LEO installations across managed multi-bearer networks. Speedcast runs GEO, LEO, 5G, L-band, LTE, microwave, and fibre as a single software-automated network across P&O Maritime's 59 vessels. The pattern is the same everywhere: not one system, but four, managed as one.
What each layer is actually for
LEO (Starlink and equivalents) — the bandwidth layer — In good conditions — open ocean, stable coverage, no regulatory restriction — LEO delivers throughput at a cost that has no precedent in maritime. That is its value. It cannot guarantee a committed minimum rate in most contracts, it is subject to restriction in multiple jurisdictions, and July 2025 demonstrated it is capable of a global outage. None of that diminishes what it does well. It does mean it cannot carry the full operational weight alone.
GEO VSAT — the committed floor — A Committed Information Rate on a geostationary link is a guaranteed minimum that LEO cannot match in most current contracts. Beam contention degrades it in congested corridors and 600-millisecond latency rules it out for real-time applications. But as the layer that provides a defined floor regardless of LEO availability, it fills a role that LEO has not replaced.
LTE and 5G — the nearshore layer — Within 20 to 50 nautical miles of the coast — covering most port approach work, cargo coordination, and communications with agents and authorities — terrestrial cellular delivers the lowest latency of any bearer. Vessels that run their entire port rotation on satellite are leaving both performance and cost on the table.
L-band (Iridium Certus) — the backstop — Iridium's constellation covers every ocean and every polar route with no dependency on regional ground infrastructure. Certus 700 delivers up to 704 kilobits per second — sufficient for GMDSS distress communications, position reporting, critical voice, and out-of-band management when every other bearer has failed. Recognised as a GMDSS service in late 2024. At its data rates, it is not a productivity tool. It is the layer that keeps a vessel reachable when nothing else will.
The decision that actually matters
Fleet managers who have worked through this do not frame it as VSAT versus Starlink. They frame it as an architecture question: four layers, managed as one network, with an orchestration policy that makes bearer selection automatic and invisible to the crew.
The decision that requires attention is not which system to run. It is who controls the failover policy, what triggers a switch, whether the fleet manager can see what each bearer is doing in real time, and whether any single provider has the visibility to manage the stack honestly on your behalf — or whether that visibility needs to sit with you.
The VSAT-versus-Starlink debate was always a vendor-framing of an operational question. The operational question is: when your primary bearer fails — and it will — what happens next, and who decided?
Orbit measures what your SLA should.
Multi-bearer visibility, incident governance, and monthly SLA packs — independent of your bearer providers.
Book a call