Crew internet on ships is a separate, metered network that lets seafarers video call home and stream, carried over the same satellite antenna as ship business but on its own VLAN. Built properly it helps retention and answers the newer welfare expectations, while navigation and engine reporting traffic keeps its priority.
Why connectivity now decides where officers sign on
Ask a crewing manager where they lose good officers and connectivity comes up fast. If two owners offer the same wage and rotation, and one lets a second engineer put his daughter on a video call on a Sunday afternoon, he takes that ship. Seafarer connectivity is part of what a crewing department sells now.
Owners we work with put the cost of replacing a senior officer in the tens of thousands of dollars, once agency fees, travel, overlap handovers and a slow first two months are counted. Your own crewing department will have a better figure than any published average. A workable crew allocation on a properly specified maritime VSAT service costs a few hundred dollars a month.
What do the welfare rules actually require?
The 2022 amendments to the Maritime Labour Convention entered into force on 23 December 2024 and brought social connectivity into the text. Shipowners are expected to provide access to ship to shore communications including internet where reasonably practicable, at a charge to the seafarer that is reasonable. That reads as guidance rather than a technical standard. Flag and port state inspectors are reported to be raising the subject, though accounts differ by region, so confirm the position with your flag.
Nobody is going to detain a ship over slow crew WiFi. The harder question arrives when your crew pay 20 dollars a gigabyte and somebody asks how that figure was set. Keep the tariff and the supplier invoice with the welfare policy and that conversation stays short. This section was last reviewed in September 2026, and regulatory detail ages faster than network design, so check the ILO amendment text and your flag circulars for anything newer.
How much data does one crew member actually use?
Consumption depends almost entirely on whether video calling works. On a ship with only a geostationary link, where calls stutter, usage looks reassuringly low because people give up trying. Give the same crew low latency and per person use climbs sharply within a month or two, often by a multiple.
| Activity | Typical data | What it really needs |
|---|---|---|
| Messaging and voice notes | 10 to 30 MB per hour | Works on any link |
| Voice call over IP | About 30 MB per hour | Latency under 400 ms, steady jitter |
| Video call home | 300 to 500 MB per hour | Steady latency and jitter |
| Streaming at 720p | 0.9 to 1.4 GB per hour | Sustained throughput |
| Streaming at 1080p | 2 to 3 GB per hour | Throughput, the biggest cost driver |
| Social feed with autoplay | 300 MB to 1 GB per hour | Throughput, and nobody notices |
| Phone and app updates | 1 to 5 GB per event | Arrives as a spike |
Turning per person figures into a monthly commit
A workable planning figure is 5 to 15 GB per person per month on a vessel with good connectivity, and the top two or three users will reach 40 GB or more. The band on its own puts a crew of 20 between 100 and 300 GB, before those heavy users are counted. Add them back and the sum is three at 40 GB with the other seventeen at 10 to 15 GB, which comes to 290 to 375 GB, so the commit belongs near 375 GB. Coastal cellular and berth WiFi absorb some of that close in, so plan around sea passage and check the beams over your trading area on the satellite coverage maps.
Why crew and operational networks stay separate
Every crew phone that comes up the gangway is an unmanaged device with an unknown patch state, and a few of them are carrying malware, which is true of any population of personal devices ashore or afloat.
On the other side of the ship sit ECDIS, the engine alarm and monitoring system, ballast and cargo control, and the VDR. IMO Resolution MSC.428(98) put cyber risk management inside the safety management system. IACS unified requirements E26 and E27 apply to vessels contracted for construction on or after 1 July 2024, and on those ships class will expect a network diagram. Crew devices sharing a broadcast domain with operational technology will attract a comment there. Older tonnage picks the obligation up through the safety management system, so confirm your position with your society.
What the segmentation looks like in practice
- Separate VLANs for operational technology, ship business IT, crew, visitors and management.
- A firewall between all of them that denies by default, with named exceptions written down.
- Crew SSID mapped only to the crew VLAN, with client isolation on so one crew phone cannot see another.
- Separate DHCP scopes and DNS resolvers per VLAN.
- No management path from the crew VLAN to any switch, router, antenna control unit or firewall, including read only access.
Network segmentation is the part of the design a surveyor can actually see, so draw it before buying anything. On new builds and larger tonnage, many owners now run a physically separate switch fabric for operational systems. Ships in service usually put segmentation, shaping, link bonding and failover in one router platform, which is how we deploy Peplink and SpeedFusion hardware under a satellite SD WAN policy.
Should crew bandwidth be capped or shaped?
Most arguments about crew internet on ships end up here. A hard cap is administratively simple and operationally poor. It wastes capacity at 03:00 when nobody is awake, and it gives the crew the version they remember, where the link works fine for eleven days and then stops dead.
A hierarchy does the job better. Operational and ship business classes get a guaranteed minimum and top priority. Crew traffic takes what is left, with a floor of around 20 percent so it never starves and a ceiling of 70 or 80 percent when the ship side is quiet. Inside the crew class, a fair queue per device stops one game update flattening the mess room.
Fair use per person still matters, and it works better as a priority change than as a wall. Once a user passes the monthly allowance that device drops down the priority order, so they keep messaging home and stop winning contention. Give voice and video latency priority even for users over allowance, because a ten minute call home costs about 60 MB and buys more goodwill than any other 60 MB aboard.
How do you charge crew for it, and who pays?
| Model | Cost to the owner | How crew react | Admin burden aboard |
|---|---|---|---|
| Owner pays, free to crew | A few hundred dollars per vessel monthly at present | Best by a distance | Lowest, no cash handling |
| Free allowance plus prepaid vouchers | Base cost, with crew funding their own extra data | Good, heavy users fund themselves | Moderate, someone sells vouchers |
| Fully prepaid, crew pay for all data | Close to nothing | Poor, harder to defend each year | Moderate to high |
| Managed crew calling platform | Per user licence plus the data | Depends on their tariff | Low aboard, managed ashore |
Those figures move with constellation, region, commitment and contract term, so treat the table as a shape. Captive portals with prepaid vouchers are still common on managed tonnage. They work, though the cost runs past the bandwidth, because somebody aboard becomes the shop, usually the purser, handling cash or slop chest deductions and complaints about codes that failed. Budget an hour a week of that, and name the job in the policy so the captive portal has an owner.
Ownership matters as much as the model. The technical department buys the bandwidth and the master carries the consequences at sea, so put one name against the policy and review it yearly.
What happens to crew internet when the primary link fails?
Few specifications cover it, and it is where crew goodwill gets destroyed. Say the primary is a low earth orbit service and the backup a geostationary Ku band link at a fraction of the capacity. On failover the crew class has to collapse automatically, to zero or to a trickle for messaging, or the backup fills with somebody's film while the chief engineer sends a condition report ashore. Define that degraded state in the routing policy so it needs no human decision, and tell the crew in advance.
Two design choices make the outage less painful. The first is a genuine hybrid where the backup carries real capacity, well past the token 2 Mbps older specifications settled for, which is the point of our hybrid satellite link designs. The second is an Iridium or Thuraya handset kept charged for the welfare call that cannot wait.
One item belongs in the written policy, where the next master can find it. The master must be able to push the crew network down or off without phoning shore first. That is rare, and it does happen, for a medical case needing a clear channel to telemedicine or for a security incident. Give the bridge one labelled action and make sure the watchkeeper knows it.
Content filtering and logging
Malware domains and phishing get blocked, along with anything clearly illegal. Peer to peer file sharing goes too, for the bandwidth and the copyright letters it attracts. Blocking social media is trivial and usually a mistake, since it is most of the reason the network exists. Shape it instead, and tell people that autoplay video can burn 800 MB in an hour.
Logging is where an owner quietly creates a problem for itself, so record volume per device for billing and fault finding and leave browsing content alone. If your company is established in the EU or the UK, or you process crew data there, GDPR or UK GDPR is likely to apply. Crew nationality on its own settles nothing, so take your own advice. A browsing history you never needed is liability without benefit. Publish the policy in the crew mess, in the working language of the ship.
Why low orbit constellations changed crew welfare
Round trip latency is the reason. A geostationary satellite sits about 35,786 km out, which puts round trip delay near 550 to 700 ms in practice. That delay is invisible to a mail client and ruinous to a video call, because it stacks with jitter until both ends talk over each other. Low earth orbit constellations typically deliver figures in the tens of milliseconds, the lower orbits at the fast end, which for a call home feels close to a connection ashore.
| Option | Round trip latency | Realistic download | Video calls home | Where it fits |
|---|---|---|---|---|
| GEO Ku band | 550 to 700 ms | 2 to 20 Mbps by plan and antenna | Poor | Global fallback, ship business |
| GEO Ka band and VHTS | 550 to 700 ms | 10 to 50 Mbps | Poor to fair | Regional, cheaper per GB |
| Low earth orbit | Roughly 30 to 70 ms, by constellation | 40 to 200 Mbps | Good | Primary crew and welfare traffic |
| Medium earth orbit | Roughly 120 to 180 ms | Tens of Mbps upwards | Fair to good | Regional capacity, larger sites |
| LEO with GEO backup | Follows the active link | As above | Good | Most managed vessels |
Pricing shifted along with the latency. Geostationary capacity has historically sold in the hundreds of dollars per Mbps per month, which is why crew allowances used to be counted in megabytes. Newer service tiers put throughput in the tens of Mbps inside a welfare budget. Compare OneWeb maritime service against a Starlink and Peplink bonded setup for your trading pattern, and treat availability and licensing as things to confirm before committing.
A checklist before you sign anything
- Size the band and the heavy users separately. Twenty crew is seventeen people on 10 to 15 GB plus three on 40 GB, so commit near 375 GB.
- Draw the network. If operational systems and crew devices share a VLAN, redesign before buying bandwidth.
- Write the priority table, with the crew floor and ceiling in it.
- Decide the degraded state on failover in writing, before the first outage.
- Choose the charging model and name the person aboard who administers vouchers.
- Set fair use as a priority drop, never a hard cutoff, keeping voice and video priority for everyone.
- Fix the filtering and logging policy, then publish it where the crew can read it.
- Confirm the master can prioritise operational traffic without calling shore.
- Ask your provider what the crew network looks like in class documentation.
Turning this into a specification
Crew internet on ships gets specified once and then lived with for years, so send us the crew list and the trading area, with any monthly data figures you have. We will check the route against where each constellation is available and licensed, then come back with a segmented maritime VSAT design and a shaping policy the master can sign. If the fleet wants a conversation first, talk to our maritime team.
