The evidence gap in maritime safety records

Ask a deck officer what happened on the 0200–0600 watch last Tuesday, and they'll give you the deck log entry. Ask them what happened in the hour before the radar contact on the port bow closed to 0.8 nautical miles — the query they put into the ECDIS, the conversation with the lookout, the speed reduction they considered and didn't make — and you'll get something different: a reconstruction from memory, a best attempt at a narrative that wasn't written at the time.

This is the evidence gap. It is not a failure of individual officers — it is a structural feature of how bridge records have always worked. Traditional deck logs were designed to create a narrative account of a voyage. They were not designed to create evidence.

What doesn't make it into the record

Bridge audio is rarely indexed. A watch officer might spend two minutes discussing a radar contact with a lookout — questioning the target's aspect, deciding whether the lights match the AIS picture, agreeing to alter course. That conversation is real decision-making about a real safety situation. It doesn't appear in the deck log.

ECDIS queries don't leave records. When an officer opens the route monitoring overlay, checks the CPA to a target on a converging course, and decides the situation is manageable — that's a professional judgment backed by instrument data. The ECDIS doesn't log the query; the deck log doesn't record the decision.

Alarm acknowledgements are logged by the integrated bridge system but rarely correlated with anything else. A high wind speed alarm acknowledged at 0347 tells you the alarm fired and was acknowledged. It doesn't tell you what the OOW then did — whether they logged a weather observation, notified the master, or made an entry in the deck log.

When the gap becomes critical

In normal operations, the evidence gap is manageable. Inspectors reviewing voyage records during a SIRE inspection or a Port State Control visit see the deck log, the compass deviation card, the equipment testing records. If nothing went wrong, the gap between the record and reality doesn't surface.

After an incident, the same gap becomes critical. P&I investigators, flag state inspectors, and legal counsel need to reconstruct what actually happened. They work from the VDR data — which captures bridge audio, heading, speed, rudder, and AIS — and from the deck log, which captures whatever the officer remembered to write down. The gap between those two records is where the questions live.

The cost of the gap is asymmetric: low in normal operations, potentially very high after an incident. Most operators don't discover the full extent of their evidence gap until they need to use the evidence.

Evidence-grade records: what they look like

Evidence-grade records are different from narrative records in one fundamental way: they are generated at the moment the event occurs, not reconstructed afterward.

A log entry created by voice at 0347 — timestamped, corroborated by the contemporaneous sensor record of position, heading, and weather — is evidence. A written entry made at 0600 summarising the watch is a narrative. Both have value; only one survives cross-examination.

The additional property that distinguishes evidence-grade records is integrity. A hash-chained log that cryptographically links each entry to the previous one makes any tampering detectable after the fact. Traditional paper logs can be altered; hash-chained digital logs cannot be altered without detection. For post-incident investigation and for the expectations of SIRE 2.0 inspectors, that property matters.