Maritime AI & Data Foundations Series : Part 2/4. How Ship Sensor Data Flows — From Onboard to Shore

Maritime AI & Data Foundations Part 1: What Is the SFI Code Part 2: Ship Sensor Data Flow
📡 Ship Data Flow AIS · ECDIS · IAS · VSAT NMEA · Protocols Maritime 4.0
PART 2 OF 4 — Maritime AI & Data Foundations

How Ship Sensor Data Flows — From Onboard to Shore

AIS, ECDIS, and IAS Data Architecture in the Connected Maritime World · ~12 min read

Captain Ethan
Captain Phul
Maritime 4.0 · AI, Data & Cyber Security  ·  linkedin.com/in/shipjobs

In Part 1, we established that SFI codes are the classification backbone that gives every onboard system a standardized address. But classification alone is not enough. Once data is generated on a vessel, it needs to travel — through layers of onboard networks, across satellite links, and into shore-based analytics platforms.

Understanding how sensor data moves from ship to shore is essential for anyone building maritime data pipelines, auditing OT cybersecurity, or developing fleet intelligence platforms. This is the architecture that makes maritime AI possible.

Key Terms
NMEA 0183 — ASCII serial data protocol standard for marine electronics (IEC 61162-1)
NMEA 2000 — CAN-bus-based binary protocol for vessel networks (IEC 61162-3)
AIS — Automatic Identification System (VHF transponder, ITU-R M.1371)
ECDIS — Electronic Chart Display and Information System (S-57/S-101 ENCs)
IAS — Integrated Automation System (onboard OT monitoring & control)
GNSS — Global Navigation Satellite System (GPS, GLONASS, Galileo, BeiDou)
VSAT — Very Small Aperture Terminal (Ku/Ka-band satellite internet)
S-AIS — Satellite AIS (AIS detection from LEO/MEO satellites)
ENC — Electronic Navigational Chart (S-57 current standard, S-101 next gen)
OPC-UA — Open Platform Communications Unified Architecture (modern IAS protocol)
TDMA — Time Division Multiple Access (AIS channel-sharing method)
VTS — Vessel Traffic Service (coastal maritime traffic management)



Section Ⅰ — What Data Does a Modern Ship Generate?

Navigation · Machinery · Safety · Environmental · Administrative

A modern commercial vessel is continuously generating data across five broad categories. Each category has different update frequencies, formats, and data owners.

Category
Data Sources
Update Frequency
🧭 Navigation
GPS/GNSS, gyrocompass, speed log, AIS, radar (ARPA), ECDIS
Continuous / 1–10 sec
⚙️ Machinery
Main engine, aux. generators, PMS, fuel flow meters, temperature/pressure sensors
1–60 seconds (IAS historian)
🔒 Safety
Fire detection, bilge alarms, flooding sensors, stability monitoring system
Event-driven + continuous
🌊 Environmental
Anemometer (wind speed/direction), barometer, sea temperature, BWTS sensors, EGCS analyzer
1–5 minutes
📋 Administrative
Crew hours (STCW), cargo manifests, port logs, maintenance records (CMMS)
Manual / per event

Section Ⅱ — Onboard Data Architecture: 4 Layers

From raw sensor signal to shore-bound data packet

Ship data does not travel from a sensor directly to shore. It passes through four distinct layers, each with its own protocols, network boundaries, and cybersecurity considerations.

Layer 1 Sensor & Instrument Layer GPS · Gyrocompass · Flow meters · Pressure/Temp sensors
Protocol: Analog signals, RS-232/RS-422, NMEA 0183 sentences
Layer 2 Subsystem Layer ECDIS · AIS transponder · IAS nodes · Radar (ARPA)
Protocol: NMEA 2000, Modbus, proprietary subsystem interfaces
Layer 3 Vessel Network Layer OT Network (IAS/NMEA) · IT Network (crew/admin) · Firewall/DMZ separation
Protocol: Ethernet/IP, OPC-UA (integration), VLAN segmentation
Layer 4 Shore Communication Gateway VSAT · Inmarsat · Starlink · Iridium · 4G/LTE (port)
Protocol: HTTPS/VPN tunnel, encrypted data packets → Shore systems

Section Ⅲ — AIS Deep Dive: The Most Widely Used Maritime Data Stream

VHF Broadcast · TDMA · Message Types · Reception Paths

AIS is the only onboard system that broadcasts vessel data publicly — making it the backbone of global maritime domain awareness and the most accessible data source for maritime AI applications.

📡 How AIS Transmits
Broadcasts on two dedicated VHF channels simultaneously:
Ch 87B — 161.975 MHz
Ch 88B — 162.025 MHz
Uses TDMA (Time Division Multiple Access) so vessels share channel time without interference. Class A mandatory for SOLAS vessels ≥300 GT on international voyages.
⏱ Transmission Intervals
>23 kts / maneuvering2 sec
Under way (slower)10 sec
At anchor3 min
Static data (vessel name, dimensions, destination): every 6 minutes
📥 Reception Paths
🚢 Vessel-to-vessel — direct VHF (range ~20–40 nm)
🗼 VTS stations — coastal receivers feed national databases
🛰 Satellite AIS (S-AIS) — LEO satellites detect AIS globally (providers: Spire, Orbcomm, exactEarth)
What an AIS Message Contains (Class A — Dynamic Data)
📍 Position (Lat/Lon)
🔢 MMSI (vessel ID)
🧭 Course over ground (COG)
⚡ Speed over ground (SOG)
🔄 Rate of turn (ROT)
🏷 Navigation status
📛 Vessel name & call sign
📐 Vessel dimensions (L × B)
🚢 Vessel type (cargo, tanker…)
⚓ Destination port
🕒 Estimated time of arrival (ETA)
📦 Type of cargo (hazmat flag)

Section Ⅳ — Communication Pathways: Ship to Shore

VSAT · Inmarsat · Starlink · Iridium · 4G/LTE

The channel a vessel uses to send data to shore determines the volume, latency, and reliability of what reaches shore-based analytics platforms. Modern ships typically operate with a primary link and at least one backup.

1 VSAT — Very Small Aperture Terminal Ku / Ka band Primary link
⚙️ How It Works
Dish antenna on vessel communicates with GEO satellites (~36,000 km altitude). Ku-band widely deployed; Ka-band offers higher throughput.
📶 Bandwidth
Typical: 5–20 Mbps (down/up symmetric in many maritime contracts). Coverage: near-global except polar regions.
📦 Data Use Cases
Fleet management data, voyage reporting, ECDIS ENC updates, crew internet, video surveillance upload.
2 Inmarsat Fleet Xpress L-band + Ka-band Reliable backup
⚙️ How It Works
L-band provides reliable low-bandwidth safety communications (GMDSS). Ka-band (Global Xpress) provides broadband layer on top.
📶 Bandwidth
Fleet Xpress Ka-band: up to 50 Mbps down / 5 Mbps up. L-band fallback: narrowband safety channel always available.
📦 Data Use Cases
GMDSS distress communications, operational data transfer, remote diagnostics. Pole-to-pole coverage advantage over VSAT.
3 Starlink Maritime Ka-band LEO High bandwidth / newer
⚙️ How It Works
LEO constellation (<600 km altitude) provides low-latency, high-throughput service. Phased array antenna self-orients toward satellites.
📶 Bandwidth
Typical: 250–350 Mbps down / 20–60 Mbps up (varies with conditions and offshore distance).
📦 Data Use Cases
High-volume sensor data offload, real-time video, crew welfare internet. Increasingly adopted as primary link for new vessels.
4 Iridium Certus L-band LEO True polar coverage
⚙️ How It Works
66-satellite LEO constellation covers 100% of Earth including poles. Cross-linked satellites relay data without ground station dependency.
📶 Bandwidth
Certus 700: up to 704 kbps down / 352 kbps up. Low throughput, but extremely reliable and the only service with true pole-to-pole coverage.
📦 Data Use Cases
Critical backup for distress communications, GMDSS fallback, essential operational data in polar/remote routes where other links fail.

Section Ⅴ — Protocols & Formats You Need to Know

NMEA 0183 · NMEA 2000 · Modbus · OPC-UA
NMEA 0183 IEC 61162-1 Navigation & AIS Sentences
ASCII-based serial protocol. Each sentence starts with $, a two-letter talker ID, and a three-letter sentence type. Transmitted at 4800 baud (standard) or 38400 baud (high-speed AIS).
Common use: point-to-point connections between bridge instruments
$GPGGA — GPS fix (position, altitude, fix quality)
$GPRMC — Recommended minimum GPS data
$AIVDM — AIS message from other vessel
$AIVDO — AIS message from own vessel
NMEA 2000 IEC 61162-3 Vessel Network Bus
CAN-bus-based binary protocol that supports multi-drop networks — multiple devices on a single backbone cable. Replaces multiple point-to-point NMEA 0183 connections. Supports plug-and-play device connectivity. Data rate: 250 kbps. Widely used for newer bridge instrument integration and engine room monitoring.
Modbus RTU/TCP Machinery OT Layer
Widely used for IAS-to-machinery communication. Simple master/slave architecture. RTU over serial (RS-485); TCP over Ethernet. Found in PMS, main engine controllers, and auxiliary system interfaces.
OPC-UA Modern Integration Layer
Increasingly adopted as the integration layer between IAS and shore systems. Platform-independent, secure, and supports semantic data modeling. Enables standardized machine-to-machine communication for fleet performance platforms.

Section Ⅵ — Cybersecurity Risks in the Data Flow

AIS Spoofing · GNSS Spoofing · VSAT Gateway Exposure
📡 AIS Spoofing
AIS protocol is unauthenticated by design. Fake AIS transmitters can inject false vessel positions, create ghost targets, or mask a real vessel. Data flows from AIS to ECDIS, VTS, and fleet monitoring platforms — corrupting all downstream analytics.
🛰 GNSS Spoofing
Falsified GPS signals affect ECDIS position display, AIS broadcast, and voyage data recorder (VDR) logs simultaneously. GPS spoofing incidents in the Black Sea and Middle East have been documented as real operational events.
🌐 VSAT Gateway
The ship's satellite router is the IT/OT network boundary. Misconfigured gateways allow lateral movement from crew internet into vessel OT networks. IACS UR E26 requires secure and monitored shore communication channels for CBS.
⚓ Captain Ethan's Take

"Most maritime data professionals I work with understand the shore-side platform — the dashboards, the APIs, the ML models. Very few understand what happens before the data reaches the API. The satellite gateway is where maritime data architecture and cybersecurity converge. If that boundary is misconfigured, every decision downstream — from maintenance scheduling to regulatory reporting — is built on data that could be compromised."

Section Ⅶ — Career Relevance: Why Data Flow Architecture Matters

How this knowledge applies across maritime data roles
Data Engineer

Building a maritime data pipeline without understanding the 4-layer onboard architecture leads to incomplete ingestion. AIS data from S-AIS providers has different latency and coverage characteristics than coastal VTS data — knowing the source determines how you handle gaps and duplicates.

Predictive Maintenance

IAS historian data logged via Modbus at 1-second intervals is fundamentally different from daily condition reports transmitted over VSAT. Feature engineering for ML models requires understanding the update frequency and protocol of each sensor stream — otherwise you introduce phantom anomalies from transmission gaps, not equipment faults.

Cybersecurity Analyst

IACS UR E26 CBS inventory and network segmentation requirements map directly to Layer 2 and Layer 3 of the onboard architecture. Auditing whether an OT network is truly isolated from the crew IT network — and whether the VSAT gateway enforces that boundary — requires a working model of how data physically flows through the vessel.

Platform Developer

Integrating NMEA 0183 sentences from a legacy bridge system alongside OPC-UA feeds from a modern IAS requires protocol translation at the ingestion layer. Fleet performance platforms that support multi-source sensor fusion need engineers who understand both what each protocol delivers and what it cannot reliably transmit under bandwidth-constrained satellite links.

⚓ Captain's Take — Key Takeaways

Ship data does not arrive at a shore analytics platform by accident. It is the product of four interconnected layers — sensor, subsystem, vessel network, and satellite gateway — each with its own protocols, failure modes, and attack surfaces. Understanding this architecture is what separates a maritime data engineer from a generic data engineer.

Ship data is generated across 5 categories (navigation, machinery, safety, environmental, administrative) with update frequencies ranging from 2 seconds (AIS) to manual entry (cargo manifests).
AIS broadcasts on VHF Ch 87B (161.975 MHz) and Ch 88B (162.025 MHz) using TDMA. Class A intervals: 2 seconds (>23 kts), 10 seconds (normal), 3 minutes (at anchor).
NMEA 0183 (IEC 61162-1) handles point-to-point serial data between instruments. NMEA 2000 (IEC 61162-3) handles multi-drop vessel network buses. Both remain core to ship data infrastructure.
Satellite connectivity defines the volume and latency of data reaching shore: VSAT (5–20 Mbps), Inmarsat Fleet Xpress (50/5 Mbps), Starlink Maritime (250–350 Mbps), Iridium Certus 700 (704 kbps backup).
AIS spoofing, GNSS spoofing, and VSAT gateway misconfiguration are the three primary cybersecurity risks in the ship-to-shore data path — each affecting the integrity of downstream analytics and compliance reporting.
📌 Series Navigation — Maritime AI & Data Foundations
Part 1: What Is the SFI Code — And Why Every Maritime AI & Data Career Starts Here
Part 2 (this article): How Ship Sensor Data Flows — From Onboard to Shore
Part 3: Python for Maritime Engineers — 5 Real Use Cases with Ship Data
Part 4: From SFI to Smart Ship — How IACS UR E26 CBS Inventory Works in Practice
#ShipDataFlow #AIS #ECDIS #NMEA #VSAT #MaritimeData #MaritimeAI #IACSURe26 #OTSecurity #Maritime4.0
📚 Related Standards & References
1
ITU-R M.1371 — Technical Characteristics for an Automatic Identification System (AIS)
ITU · Current edition: M.1371-5 · Defines AIS VHF frequencies, TDMA protocol, and message structure · itu.int
2
NMEA 0183 Standard — National Marine Electronics Association
NMEA · IEC equivalent: IEC 61162-1 · ASCII serial data protocol for marine electronics · nmea.org
3
IHO S-57 — Transfer Standard for Digital Hydrographic Data (ENC format)
IHO · Current ECDIS chart format · S-101 (S-100 framework) is the next-generation standard · iho.int
4
IACS UR E26 — Cyber Resilience of Ships
IACS · Requires secure shore communication channels and OT/IT network segmentation for CBS · iacs.org.uk
5
ShipPaulJobs — IACS UR E26/E27 Resource Library
ShipPaulJobs · Download IACS, BIMCO, NIST, IEC 62443 compliance PDFs · shippauljobs.com
Captain Ethan
Captain Paul 
Maritime 4.0 · AI, Data & Cyber Security ·
Maritime Intelligence Platform · Cyber · AI · Data
shippauljobs.com

Comments