📚 Series Maritime Cybersecurity Learning Roadmap | PART 4 Lesson 17 — 2 of 4 in this PART · Course Index →

SIEM: Turning Logs into Security Intelligence

What logs are · collect, parse, normalise, enrich · detection rules, correlation and UEBA · choosing priority logs · the logs IACS UR E26 asks ships to keep.

Captain Paul
Captain Paul
Maritime Cybersecurity · IACS UR E26/E27
November 2026
🎯 Lesson Objective

After this lesson you can explain what a log is and why it matters, describe how a SIEM collects, parses, normalises, enriches and stores log data, distinguish single-event, threshold, correlation, threat-intelligence and behaviour-based detections, explain what a use-case library is and why alert tuning matters, and identify the records that IACS UR E26 requires ships to produce.

Maritime Cybersecurity Learning Roadmap

After a collision or grounding, investigators go first to the records: the deck log, the engine log, the ECDIS track and the VDR. Without records there is only memory and opinion. In cybersecurity the records are logs, and the system that turns millions of them into something a human can act on is the SIEM.

In Lesson 16 the SIEM appeared as the centre of the SOC's tool set. This lesson opens it up: how logs get in, how they are made comparable, how rules decide what is worth an analyst's attention, and how the investigation that follows uses the same data.

Everything here applies to any SIEM product. Lessons 18 and 19 then show how two widely used platforms — Microsoft Sentinel and Splunk — put these ideas into practice.

1. What Logs Are — and Why They Matter

A log is a time-stamped record of something that happened on a system: a user logged on, a program started, a firewall blocked a connection, a setting was changed. Every computer, network device and cloud service can write logs, and most do so by default in some form.

Logs are to cybersecurity what the deck log, engine log and VDR are to a casualty investigation: the only reliable record of what happened, in what order. NIST's long-standing Guide to Computer Security Log Management (SP 800-92) was written to help organisations understand the need for sound log management, because logs that are never collected, are deleted too early or cannot be trusted are worthless when an incident happens.

Windows shows how specific logs can be. Microsoft documents security event 4625 — “An account failed to log on” — as logged for any logon failure, on the computer where the attempt was made. The event records, among other fields, the account name, the logon type (2 = interactive at the keyboard, 3 = from the network, 10 = remote desktop) and the source network address. One failed logon means nothing; fifty in a minute from one address means something.

Why a SIEM is needed: A single server can write thousands of events an hour, and a fleet has hundreds of devices. No one can read that by hand. A SIEM — Security Information and Event Management system — collects it all in one place, makes it searchable and watches it automatically.

2. The Log Pipeline

From raw log to usable data
1
Collect

Logs reach the SIEM through an agent installed on the source, through syslog (a standard way for network devices to send log lines), or through an API connection to a cloud service. Microsoft Sentinel, for example, lists Common Event Format (CEF), Syslog and REST API as ways to connect sources that have no ready-made connector.

2
Parse

Each product writes logs in its own format. Parsing splits a raw line into named fields — time, user, source address, action — so they can be searched.

3
Normalise

Different products call the same thing by different names (“src_ip”, “SourceAddress”, “client”). Normalisation maps them to one common schema, so one rule can work across many sources. Sentinel calls its approach the Advanced Security Information Model (ASIM).

4
Enrich

Add context: which ship or office the device belongs to, how critical it is, whether an IP address appears in threat intelligence.

5
Store and retain

Keep data searchable for investigations and long enough for regulators and audits. Recent data is usually kept in fast, more expensive storage, older data in cheaper archive storage.

Time matters: If the clocks on the ship's devices are wrong, events land in the wrong order and an investigation can draw the wrong conclusion. Consistent time synchronisation is a quiet but essential part of any logging design.

3. Detection — Rules, Correlation and Analytics

Once data is in the SIEM, it is watched by detection rules — saved searches that run automatically and raise an alert when they find a match.

Ways a SIEM detects
Single-event rules Alert on one event that should never happen, such as a new administrator account being created on a bridge workstation.
Threshold rules Alert when something happens too often in a time window — for example many 4625 failed logons for one account within ten minutes, a sign of password guessing.
Correlation rules Alert on a combination across sources and time: failed logons followed by a successful one from the same address, then a new remote session. NIST CSF 2.0 lists this as outcome DE.AE-03 — information is correlated from multiple sources.
Threat intelligence matching Alert when traffic or e-mail involves an address, domain or file already known to be malicious (CSF DE.AE-07).
Behaviour analytics (UEBA) User and Entity Behaviour Analytics learns what is normal for each user or device and flags unusual behaviour even when no rule describes it — such as an account active at an hour it has never been used.

A SOC keeps its rules in a use-case library: for each rule, the threat it addresses, the ATT&CK technique it maps to (Lesson 15), the log sources it needs, the expected false alarms and what the analyst should do when it fires.

Alert fatigue: A SIEM that raises hundreds of meaningless alerts a day trains analysts to stop looking — exactly like a bridge alarm that sounds so often that the crew silences it without reading it. Tuning rules, adding known-good exceptions and retiring noisy rules is continuous work, not a one-off task.

4. Choosing What to Log First

Sending every log to a SIEM is expensive and slow, and on a ship it is often impossible over the satellite link. The real skill is choosing which logs matter most.

The CISA / ASD guidance package on SIEM and SOAR implementation (May 2025) includes a dedicated practitioner guide, Priority Logs for SIEM Ingestion, written to help organisations decide which logs to send to a SIEM first. A practical approach is to start with the sources that show who got in and what they changed, then grow:

A starting order for a shipping company (teaching example)
Identity Sign-in logs for e-mail and cloud accounts, domain controller logs ashore.
Perimeter and remote access Firewall logs at the office and on each ship; logs of every remote access session to ship systems.
Endpoints Security events from office PCs and from Windows-based workstations on board, or alerts from EDR.
Ship networks and OT Alarms and records from network monitoring on board, switch and device management activity.

This order is the author's teaching suggestion, not a ranking taken from the CISA/ASD guide; read the guide itself for its recommended priorities.

5. From Alert to Investigation

When a rule fires, the work moves from the machine to the analyst:

Investigation workflow
1
Triage

Is it real? Check the raw events behind the alert, look up the user, device and addresses involved, and either close it with a reason or escalate.

2
Scope

Pivot: search for everything the same account, device or address did before and after. Build a timeline.

3
Decide and contain

Agree the response — reset a password, block an address, isolate a computer, or call the vessel — following the incident response plan.

4
Record and learn

Document the case and ask: what rule or log source would have caught this sooner?

⚓ Maritime Connection — The Logs IACS UR E26 Asks Ships to Keep

IACS UR E26 Rev.1 does not mention SIEM products, but several of its requirements produce exactly the records a SIEM is built to watch:

Remote access logging (4.2.6). Remote access to systems in scope must provide a logging function that records all remote access events and keeps them long enough for offline review, for example after a cyber incident has been detected.

Network monitoring records (4.3.1). Monitoring must include monitoring of network connections, monitoring and recording of device management activities, protection against connection of unauthorised devices and a bandwidth alarm. During commissioning, the test procedures include demonstrating that audit records are generated.

Inspection of audit records. In the operation phase the shipowner must describe in its cyber security and resilience programme how anomalies are detected, including inspection of security audit records.

Practical design questions. Ship OT systems often write logs in proprietary formats, or none at all, so parsing them may need custom work and help from the equipment supplier. Over a limited satellite link, a common design is to collect and filter logs on board and send only the most important records ashore, keeping the full set on board for later investigation. Whatever design is chosen, it should be agreed with the system suppliers, since E26 requires any intrusion detection to be passive and supplier-qualified.

✅ Key Takeaways — Lesson 17
  • Logs are the time-stamped record of what happened on systems; without them there is no detection and no investigation.
  • A SIEM collects logs, parses and normalises them into common fields, enriches them with context, stores them and runs detection rules.
  • Detections range from single-event and threshold rules to multi-source correlation, threat-intelligence matching and behaviour analytics (UEBA).
  • A use-case library links each rule to a threat, an ATT&CK technique, its data sources and an analyst procedure; tuning against alert fatigue never stops.
  • Choosing which logs to send first matters more than collecting everything — CISA and ASD published dedicated guidance on priority logs in May 2025.
  • IACS UR E26 requires remote access logging, recording of device management activity and inspection of audit records — natural inputs for SIEM-based monitoring.
📚 PART 4 — Security Operations
Current LessonLesson 17 — SIEM: Turning Logs into Security Intelligence
Captain Paul
Captain Paul
Maritime Cybersecurity · IACS UR E26/E27 · ShipPaulJobs

Captain Paul writes the Maritime Cybersecurity Learning Roadmap on ShipPaulJobs — a step-by-step course that takes readers from security fundamentals to the cyber resilience rules that now apply to ships.

← Course Index LinkedIn

⚓ Join the ShipPaulJobs Community

Join →
Share

Comments

Top Ranked · Last 30 Days

Popular Posts