The multi-bearer question: why choosing VSAT or Starlink is the wrong frame

Every few months a new post appears claiming VSAT is dead, or that Starlink has won, or that LTE coverage has rendered offshore satellite redundant for nearshore operators. The conversation is always framed as a choice: which single system should you use?

That is the wrong question. The answer is not VSAT or Starlink. The answer is an architecture.

Why the comparison misses the point

VSAT and Starlink have genuinely different performance profiles. GEO VSAT delivers latency consistently above 600 ms, relatively predictable throughput on contracted capacity, and proven resilience on stabilised antenna systems aboard larger vessels. Starlink Maritime delivers latency typically in the 25–60 ms range under good conditions — though ocean routing paths can push this higher — significantly lower subscription costs at comparable data volumes, and rapidly expanding global coverage. Neither is universally superior.

More importantly, neither is designed to be the sole connectivity layer on a commercial vessel where connectivity failure has operational, safety, and now regulatory consequences.

The VSAT-or-Starlink framing treats connectivity as a commodity purchase. Modern vessel connectivity architecture is not a procurement decision. It is a design problem.

The real architecture: multi-bearer with a defined fallback chain

A vessel connectivity stack built for operational continuity works in three layers. The primary bearer handles the bulk of operational and crew traffic — typically VSAT for large vessels on long deep-sea routes where sea state and antenna stabilisation matter, or Starlink for smaller vessels, nearshore operators, or cost-sensitive deployments on moderate-weather routes. It carries high-bandwidth tasks: ECDIS updates, planned maintenance system synchronisation, crew welfare, video conferencing.

The secondary bearer activates on primary failure or when primary performance degrades below a defined threshold. Where primary is VSAT, secondary is often Starlink or an LTE aggregate. Where primary is Starlink, secondary is VSAT. The bearers are complementary, not competing.

L-band is the backstop. Iridium Certus or Inmarsat Fleet One sits underneath both, delivering low-bandwidth, high-reliability connectivity — sufficient for GMDSS, distress communications, and critical position reporting — when everything else fails. L-band is not the primary bearer. It is the insurance layer that should never not be there.

None of this works without an orchestration layer that monitors bearer health in real time, applies failover logic automatically, and logs what happened and when.

Why cost is not the decision

Starlink's Flat High Performance hardware — the only terminal designated for in-motion maritime use — costs approximately $2,500. Current maritime service plans are data-block based: 50 GB for $250/month at the entry tier, scaling to $1,000/month for 1 TB. A stabilised VSAT dome for a commercial vessel typically runs $5,000–15,000 in hardware, with service contracts from $500 to several thousand per month depending on capacity and provider.

The cost comparison is real but not as simple as it looks. Starlink's per-GB cost is higher on smaller data blocks; VSAT's committed capacity can work out cheaper at volume on vessels with predictable high usage. More importantly, if you are comparing the cost of Starlink alone against VSAT alone, you are still thinking about this as a single-bearer decision. The right cost comparison is: what is the total cost of a resilient multi-bearer architecture versus the operational cost of connectivity failure?

Starlink's current operational limits

Starlink Maritime's flat panel antenna performs well in moderate conditions. At significant wave height above 2–3 m, performance can degrade. In severe weather, the terminal can lose lock. This is a present operational constraint that matters on specific routes and vessel types — the flat phased array has no mechanical stabilisation.

Operators on Arctic, Antarctic, or North Atlantic routes should verify current coverage against their specific corridors. Regulatory gaps also apply: Starlink service is not permitted in all waters, and vessels transiting certain EEZs — including parts of Chinese coastal waters — may experience service interruption regardless of hardware.

What the orchestration layer actually does

SD-WAN applied to vessel connectivity manages bearer selection, traffic prioritisation, and failover according to policy rules set by the operator. OT system traffic — ECDIS, engine monitoring, planned maintenance systems — is assigned priority and routed over the most reliable available bearer, not the cheapest one. Crew welfare traffic is routed over secondary capacity during operational peaks. Large bulk transfers — software updates, log uploads — are scheduled for periods when primary bearer capacity is underutilised.

Without an orchestration layer, you have multiple bearers and no policy. You are paying for redundancy you cannot manage.

The decision that actually matters

Not VSAT or Starlink. How to design a bearer stack for your specific vessel type, route profile, and operational requirements — and how to manage it.

The vessels that will have connectivity problems in 2026 are not the ones that chose the wrong single system. They are the ones that bought multiple systems and left them unmanaged.