Ro-ro ferry contact with a harbour breakwater
MAIB Summary of Investigations No. 1/92 · Case 2 · Published April 1992
The ferry began the turn ordered by the master. But when the helmsman eased the wheel, the rudder went hard over in the opposite direction. The autopilot was still engaged: the first response had come from an emergency override, not a completed change to manual steering.
Identification note: MAIB’s published account does not name the vessel or harbour, or give the date of the accident. We are awaiting confirmation from MAIB before adding identifying details. This review is based on the published case summary.
1 | Incident snapshot
| Item | Data |
|---|---|
| Vessel | Ro-ro ferry; name not given in the summary |
| Operation | Harbour entry at the end of a scheduled passage |
| Harbour and accident date | Not identified in the published case |
| Conditions | A dark night, calm sea and good visibility |
| Bridge team | The master, certificated officers and two seamen |
| Steering status | The officer of the watch assumed manual steering had been selected; the autopilot remained engaged |
| Outcome | Contact with the breakwater, considerable damage and withdrawal from service; no passenger or crew injuries |
Source: MAIB Case 2, Narrative and Observations 1–3, printed p. 2.
2 | Documented sequence
The summary provides a sequence, not a timed bridge transcript. No clock times or verbatim helm orders have been reconstructed here.
| Stage | What happened | Why it mattered |
|---|---|---|
| Before reaching the harbour | The helmsman took his station at the wheel. The usual change to manual steering was not made because he was unfamiliar with the system. | The officer of the watch assumed the changeover had been completed and did not check it. |
| Approach in calm conditions | No major rudder movements were needed. | The helmsman did not realise that he was not controlling the steering. |
| Turn into the harbour | The master ordered a large course change. The helmsman used full rudder and the vessel turned as intended. | The hard-over wheel position permitted emergency override of the autopilot. |
| Wheel eased from hard over | The rudder immediately went hard over in the opposite direction and the vessel returned towards its original course. | The initial turn had not established that manual steering was selected. |
| Automatic mode recognised | The sounding off-course alarm was recognised and the steering-mode error was corrected. | The vessel was already too close to the breakwater to avoid contact. |
Source: MAIB Case 2, Narrative and Observations 2–6, printed p. 2.
3 | Why the first turn was misleading
The steering arrangement described by MAIB allowed the wheel to override the autopilot when held at hard over. It did not require the autopilot to have been disengaged for that initial response to occur.
Once the wheel was eased, the automatic system applied opposite rudder. The vessel’s initial response to the helmsman was therefore not evidence that the expected manual-control mode was active.
Temporary override is not the same as a confirmed changeover.
This is the behaviour reported for this particular installation. It should not be treated as a description of every autopilot or as a general instruction for taking manual control. The case concerns automatic-to-manual steering, not a documented ECDIS track-pilot transfer.
Source: MAIB Case 2, Narrative and Observation 5, printed p. 2.
4 | What MAIB identified
The published account describes several missed safeguards, not simply an error at the wheel:
- Changeover assumed rather than checked. Shipboard practice placed the switch-over task with the helmsman, but the officer of the watch did not verify completion.
- Poor mode indication. The selector was not illuminated and there was no other indication of the active steering mode.
- Alarm identity unclear. The off-course alarm was initially taken for a navigation-light failure alarm, then an alarm from the fire-detection system.
- Preparation incomplete and late. The officer completing the arrival checklist was distracted. There was no requirement to report positively that the checks were complete, so entry continued. MAIB also found that the checklist had been started so late that even completion would probably have left insufficient room for corrective action.
The summary records that the owners took action to prevent a recurrence, without specifying those measures.
Source: MAIB Case 2, Observations 3–9 and Comment 6, printed pp. 2–3.
5 | Relevance to Helm Order Monitor
ELNAV.AI technical interpretation, separate from MAIB’s findings.
Helm Order Monitor compares recognised spoken helm orders with available bridge and steering data. This case illustrates why verification must consider the response that follows an instruction, rather than assume that a person at the wheel has effective manual control.
Two questions need to remain distinct:
| Verification question | Information needed | Limit of the conclusion |
| Does the response remain consistent with the current recognised order? | A reliable recognised instruction and correctly configured, time-aligned rudder and relevant navigation data | A discrepancy can call for verification; it does not, by itself, establish why the response is wrong. |
| Is the expected steering mode actually active? | Reliable active-mode or control-status information from the installation, or direct confirmation by the bridge team | Rudder movement alone does not prove that manual mode has been selected. |
The initial hard-over response in this case appeared correct. A meaningful verification approach must therefore consider what happens subsequently, with any new instruction replacing the previous one. It must not treat a temporary response as continuing proof of manual control.
A prompt such as NO RESPONSE: CONFIRM MANUAL CONTROL is relevant as a request to verify the control path where the expected response is absent. It is not a claim that the rudder never moved in this accident, or that this exact prompt would necessarily have been generated. Here the response first appeared correct and then reversed.
The summary does not reproduce the precise spoken orders, read-backs or sensor record needed to replay the event against HOM. We therefore do not assign an alert time, claim a warning interval gained or state that HOM would have prevented the contact. Identifying the vessel later would not, on its own, establish any of those performance claims.
6 | Operational lessons
MAIB’s comments emphasised verifying the change from automatic to manual steering and checking that manual steering works, supported by personnel familiarisation. The wider lessons were equally practical:
- Make the active steering mode clearly identifiable from the control position.
- Complete readiness checks early enough to resolve any problems before the harbour entry becomes critical.
- Require a positive report of completed safety checks; absence of a problem report is not confirmation of readiness.
- Make different alarms readily distinguishable, so the bridge team can identify the system calling for attention.
These are the lessons recorded in MAIB’s historical case summary. Its references to the 1981 regulations and the 1991 Merchant Shipping Notice are historical references, not a statement here about the legislation currently applicable to a particular vessel.
Source: MAIB Case 2, Comment 1–6, printed p. 3.
Take-home message
A response to the wheel is not proof that manual steering is engaged.
Confirm the active mode, verify the response and report readiness before the manoeuvre. HOM provides advisory support for checking recognised orders against available vessel data; it does not replace the changeover procedure or the bridge team’s responsibility to confirm effective control.
Source: Marine Accident Investigation Branch, Summary of Investigations No. 1/92, April 1992, Case 2: Contact between ro-ro ferry and harbour breakwater, printed pp. 2–3 (PDF pp. 5–6).
Read MAIB’s Case Summary (PDF)
Independent incident review. The HOM discussion is ELNAV.AI’s interpretation, not a MAIB recommendation or endorsement. This is not a report of HOM deployment aboard the ferry.
