HMNZS Manawanui

[David Newton]

Looks like a report has been released into the sinking of that RNZN survey ship.

Turns out the problem was the ship was on autopilot and no one noticed or checked during the wrecking.

[Poohbah]

Someone hook a generator up to my Dad's niche in the columbarium at Fort Rosecrans, you oughta be able to generate a few hundred megawatts...

[kdahm]

The Senior Chief will have a few words to say. Many may even be gentle enough to be posted here.

[Micael]

Oh boy.

ONE FOR THE BOOK
NZ Navy ship accidentally left on autopilot before sinking, inquiry finds

New Zealand Navy chief Garin Golding admitted a “series of human errors” was to blame for the HMNZS Manawanui " crashing into a coral reef before sinking.

The inquiry’s first report, released Friday, showed they believed a problem with the ship’s thruster was stopping the vessel from changing course when all they had to do was “lean over to the panel and check whether it was on autopilot.”

Separate disciplinary proceedings will be launched involving three unnamed crew members involved in the incident.

Neither the person piloting the Manawanui, their supervisor, nor the captain, thought to check whether the autopilot was still on. All are under investigation and may face further disciplinary action.

[Nightwatch2]

Unbelievable

That’s about the most printable thing I can say

What an incredible bit of incompetence

About the only positive thing I can say is that they were on a mission to map the sea floor and they did find it…

[gtg947h]

... wow.

"Mode awareness" was a theme of a couple investigations I conducted at work over the last couple weeks. Fortunately none of them ended with bent metal or injuries...

[Micael]

Details from interim court of inquiry report into HMNZS Manawanui incident released

HMNZS Manawanui grounded and sunk off the coast of Samoa as a result of human error, the interim Court of Inquiry report shows.
NZDF Kiwi
29 NOVEMBER, 2024

The Royal New Zealand Navy ship grounded on a reef on the southern side of Samoa, on Saturday 5 October 2024 while conducting survey operations.

Chief of Navy Rear Admiral Garin Golding stood up a Court of Inquiry following the incident in order to understand the facts of what occurred.

“The direct cause of the grounding has been determined as a series of human errors which meant the ship’s autopilot was not disengaged when it should have been,” Rear Admiral Golding said.

“The crew did not realise Manawanui remained in autopilot and, as a consequence, mistakenly believed its failure to respond to direction changes was the result of a thruster control failure.

“Having mistakenly assessed a thruster control failure, standard procedures should have prompted ship’s crew to check that the ship was under manual control rather than in autopilot. This check did not occur. Remaining in autopilot resulted in the ship maintaining a course toward land, until grounding and eventually stranding.”


FILE IMAGE OF THE BRIDGE OF HMNZS MANAWANUI SHOWING LOCATION OF THE AUTOPILOT BUTTON AND THRUSTER CONTROLS.


TRACK OF HMNZS MANAWANUI BEFORE GROUNDING AND SINKING.

Rear Admiral Golding said why it happened and what would come next in terms of lessons learned were still being worked on as part of the wider Court of Inquiry, which was expected to be completed in the first quarter of next year.

Rear Admiral Golding said given human error was identified as the cause, a separate disciplinary process would need to be commenced once the Court of Inquiry had concluded.

As well as the primary cause, there were also a number of identified contributing factors leading to the ship’s grounding, which will be considered in more detail during phase two of the Court of Inquiry.

“To provide some immediate assurance, we have conducted a series of audits in the Fleet and looked to implement initial lessons identified from the interim report around training, risk management, and improving relevant orders, instructions and procedures” he said.

Rear Admiral Golding said he was grateful for the support from the Transport Accident Investigation Committee, the Royal Australian Navy, and Royal New Zealand Air Force for their support to the Court of Inquiry investigation, which used evidence collected through interviews with the crew, expert witnesses and data from the ship’s recordings to find the primary cause.

The New Zealand Defence Force remains thankful for Samoa’s ongoing support regarding the situation, and is committed to working alongside Samoan authorities on the response, he said.

“I want to reassure the public of New Zealand that we will learn from this situation and that it is on me, as the Chief of Navy, to earn back your trust.

“We have incredible people within our Defence Force who answer the call to serve every day. Much of their work is risky and involves decisions that can mean life or death. No one turns up to work to have a bad day.

“In this situation, we thankfully did not lose any lives but lives have been affected nonetheless, and we continue to support and work closely with those who were onboard Manawanui on that day.”

Timeline:

On Saturday 5 October 2024, HMNZS Manawanui was conducting survey operations on the southern side of Apia, Samoa in a strong breeze of up to 25 knots and moderate swell.
The survey was conducted in a box-shaped area, running east to west in survey lanes that start on the outside, working inwards.
At about 6.15pm, the ship’s crew attempted a routine turn to starboard, initially to a course of 340 degrees, within the survey area, as part of a turn. The crew attempted to turn off the 340 degree course to starboard towards an easterly course but the ship did not respond as intended.
Shortly after, Manawanui left the approved survey area, and in an effort to stop the ship, the crew conducted further actions that they believed should have resulted in the ship essentially braking.
Manawanui did not slow or stop, and instead the ship started to accelerate towards the reef, grounding for the first time at or about 6.17pm at a speed of more than 10 knots.
The ship then travelled around 635 metres (400 yards) before becoming stranded, grounding multiple times along the way.
Full control of Manawanui’s propulsion system was not regained until 10 minutes later, at 6.27pm, when the ship’s autopilot was disengaged. The inability to turn the ship to an easterly direction from the 340 degree course and stop the ship is attributed to the ship being in autopilot mode.
Unsuccessful attempts were then made to manoeuvre the ship off the reef.
Manawanui was brought to emergency stations after the grounding, and searches were conducted to check for damage.
No damage or flooding was detected inside the ship. However, stability assessments made after the grounding indicated Manawanui was no longer stable.
At approximately 6.46pm, about 30 minutes after the initial grounding, the decision was made to abandon ship.
The timeliness of the decision to abandon ship and to keep Manawanui’s generators running contributed to the successful abandonment process and likely prevented serious injuries or death.
The ship suffered a series of catastrophic fires after being abandoned, before capsizing and sinking on the morning of Sunday 6 October.

[MikeKozlowski]

...Follow up question: I would have thought that any steering/throttle input at all should have shut down the autopilot and immediately returned manual control to the helm. Why would this not have been the case?

Mike

[David Newton]

...Follow up question: I would have thought that any steering/throttle input at all should have shut down the autopilot and immediately returned manual control to the helm. Why would this not have been the case?

Mike

Idiots designed the system?

[Micael]

...Follow up question: I would have thought that any steering/throttle input at all should have shut down the autopilot and immediately returned manual control to the helm. Why would this not have been the case?

Mike

It depends on the steering system in question. There’s even a small number of aircraft models where it will only shut off if you press the dedicated disconnect switch. As I understand it the rationale in those cases is that it could be more dangerous if you inadvertantly disconnect the autopilot by bumping a control, and don’t notice it. Whereas in a case where you want to go from autopilot to manual control and forget to press the switch you should realize that pretty quickly if the craft doesn’t behave as expected, since you’re paying more attention then as you intend to control it manually.

[Drunknsubmrnr]

Something similar lead to a Canadian helicopter crash a few years ago. It’s not just NZ.

[Poohbah]

...Follow up question: I would have thought that any steering/throttle input at all should have shut down the autopilot and immediately returned manual control to the helm. Why would this not have been the case?

Mike

It depends on the steering system in question. There’s even a small number of aircraft models where it will only shut off if you press the dedicated disconnect switch. As I understand it the rationale in those cases is that it could be more dangerous if you inadvertantly disconnect the autopilot by bumping a control, and don’t notice it. Whereas in a case where you want to go from autopilot to manual control and forget to press the switch you should realize that pretty quickly if the craft doesn’t behave as expected, since you’re paying more attention then as you intend to control it manually.

There should be an indicator at the helm station itself showing the autopilot status. But I'm weird that way

[Johnnie Lyle]

...Follow up question: I would have thought that any steering/throttle input at all should have shut down the autopilot and immediately returned manual control to the helm. Why would this not have been the case?

Mike

It depends on the steering system in question. There’s even a small number of aircraft models where it will only shut off if you press the dedicated disconnect switch. As I understand it the rationale in those cases is that it could be more dangerous if you inadvertantly disconnect the autopilot by bumping a control, and don’t notice it. Whereas in a case where you want to go from autopilot to manual control and forget to press the switch you should realize that pretty quickly if the craft doesn’t behave as expected, since you’re paying more attention then as you intend to control it manually.

There should be an indicator at the helm station itself showing the autopilot status. But I'm weird that way

A blatantly obvious one to boot.

[Poohbah]

It depends on the steering system in question. There’s even a small number of aircraft models where it will only shut off if you press the dedicated disconnect switch. As I understand it the rationale in those cases is that it could be more dangerous if you inadvertantly disconnect the autopilot by bumping a control, and don’t notice it. Whereas in a case where you want to go from autopilot to manual control and forget to press the switch you should realize that pretty quickly if the craft doesn’t behave as expected, since you’re paying more attention then as you intend to control it manually.

There should be an indicator at the helm station itself showing the autopilot status. But I'm weird that way

A blatantly obvious one to boot.

Complete with the Alarm Tone of Shame if you move the controls without disengaging the autopilot first.

[Johnnie Lyle]

There should be an indicator at the helm station itself showing the autopilot status. But I'm weird that way

A blatantly obvious one to boot.

Complete with the Alarm Tone of Shame if you move the controls without disengaging the autopilot first.

“ATTENTION DUMBASS!” or “HELLO STUPID?”

Unfortunately, it’s the same problem as trying to bearproof the trash cans in Tahoe. Except with panic making people even less able to think clearly.

[Poohbah]

A blatantly obvious one to boot.

Complete with the Alarm Tone of Shame if you move the controls without disengaging the autopilot first.

“ATTENTION DUMBASS!” or “HELLO STUPID?”

Unfortunately, it’s the same problem as trying to bearproof the trash cans in Tahoe. Except with panic making people even less able to think clearly.

A calm but firm voice says, "TURN OFF THE AUTOPILOT IF YOU INTEND TO MANEUVER" while the On/Off switch flashes.

That should be enough.

[Micael]

Complete with the Alarm Tone of Shame if you move the controls without disengaging the autopilot first.

“ATTENTION DUMBASS!” or “HELLO STUPID?”

Unfortunately, it’s the same problem as trying to bearproof the trash cans in Tahoe. Except with panic making people even less able to think clearly.

A calm but firm voice says, "TURN OFF THE AUTOPILOT IF YOU INTEND TO MANEUVER" while the On/Off switch flashes.

That should be enough.

You might in some instances then run into one of two problems that have featured in some aircraft accidents.

1. There’s multiple cautions going off at once and:
1a. They interfere with one another so that the crew misses one or more of them.
1b. If given an internal prioritity order the caution is silenced by another caution which has been given higher priority.

2. The crew is deep enough in the stress cone that their minds inadvertantly filter out the cautions because they’re so focused on a pressing issue that they can’t take it in. This seems to more be an issue with oral cautions and less so with visual ones. The visual ones doesn’t seem to be filtered out by the brain as easily.

So yes, as described it could help in some situations. In others it might not. I would probably be inclined to advocate for a very clear flashing warning light.

[Craiglxviii]

“ATTENTION DUMBASS!” or “HELLO STUPID?”

Unfortunately, it’s the same problem as trying to bearproof the trash cans in Tahoe. Except with panic making people even less able to think clearly.

A calm but firm voice says, "TURN OFF THE AUTOPILOT IF YOU INTEND TO MANEUVER" while the On/Off switch flashes.

That should be enough.

You might in some instances then run into one of two problems that have featured in some aircraft accidents.

1. There’s multiple cautions going off at once and:
1a. They interfere with one another so that the crew misses one or more of them.
1b. If given an internal prioritity order the caution is silenced by another caution which has been given higher priority.

2. The crew is deep enough in the stress cone that their minds inadvertantly filter out the cautions because they’re so focused on a pressing issue that they can’t take it in. This seems to more be an issue with oral cautions and less so with visual ones. The visual ones doesn’t seem to be filtered out by the brain as easily.

So yes, as described it could help in some situations. In others it might not. I would probably be inclined to advocate for a very clear flashing warning light.

The first human sense to be dulled in a high-stress situation is hearing. It’s much easier for us to blank out eg someone screaming an order, than it is something we can see happening.

A visual and/ or tactile warning is far more effective.

[kdahm]

This really gets into the design philosophy of the ship or airplane. What is the proper behavior if the craft if the controls are moved when the autopilot is on? If the assumption is that the autopilot or computer is more likely to be correct, then the control systems ignore the input and continue with what they were doing. If the assumption is that the pilot/helmsman is generally correct, then the craft follows the inputs given while the autopilot may or may not attempt to correct during or after the control input, all the while an alarm is indicating that the autopilot is on.

In my opinion, if the helmsman gives a right rudder and medium back control input, I damn well better see the rudder move right and the engine go to medium back. There should not be a hunt to turn the autopilot off first. I'd like to see a force feedback to the helmsman in the direction the autopilot wants to steer the boat, but the helmsman should be able to override without excessive force.

If the autopilot is on with a ship this size, there should be a projector putting letters at least 20cm tall across the top of the bridge windows "AUTOPILOT ON", especially if it disables the helm control. There should be a Big Red Button at each station where the ship can be controlled that disables the autopilot when pressed and triggers a notice. If the Navy wants to be sure it's not pressed accidentally, fit a molly guard to the button.

Anyone designing the layout of a helm station must be qualified to stand a helm watch. Anyone designing the overall bridge layout of a navy ship must be qualified to stand at least a JOOD watch.

[Pdf27]

Two observations:

1. Ship and aircraft autopilot systems have very different requirements. Things happen at least one and more usually two orders of magnitude faster in an aircraft, leading to much more severe cognitive overload.
2. <snark> Unlike most previous grounding or collision incidents of this type, they were at least aware of the object before they hit it! </snark>