HMNZS Manawanui | 09 Materiel

Introduction

The Court finds at ¶61(q) that the ship was "assessed to have been placed in autopilot mode" at 18:13:08. Placed is a transitive verb, and it carries the bias a crew member did it, despite the act not being attributed to any crew member in the unredacted text.

The only term of reference about the equipment

TOR 33 is the only term of reference that asks directly about the ship's equipment as a cause. It asks whether the materiel state contributed to the grounding, to the loss of the ship, and to the evacuation.

TOR 33 runs to twelve paragraphs. In a 120-page report into the loss of a naval asset, two pages carry the investigation into the materiel state.

Eight ways into autopilot

In the attached diagram I set out eight ways the ship could have been placed in autopilot mode. One is solely due to the operator. Two occur at the operator/machine interface. Two are system behaviour. Three are machinery defects or faults.

Applying Heuer's analysis of competing hypotheses,¹ the inquiry as constituted could have reached only three of the eight.

What was never recovered

Also striking is the absence of any recovery or inspection of the autopilot control unit, its sensors, or the thruster control electronics. The Court made attempts to have the ship's ECPINS system recovered for further analysis (¶186). So recovery from the wreck was possible, and attempted for one system. The unredacted text records no equivalent attempt for any part of the control system.

If these hypotheses were deemed unlikely and excluded by the inquiry, Heuer would call this "satisficing" — settling on the first explanation that appears good enough rather than testing every hypothesis to find the one that best fits the evidence.

So what is my point?

The report finds that surveying was never operationally released, that no Seaworthiness Authority declaration for it existed, and that the Sea Acceptance Trials which would have proved the systems were at the correct performance standard were never completed (¶50, ¶233, ¶245(u)).

The autopilot is central to what happened. As I have worked through this case, what strikes me is that a system this close to the loss was never examined through a lens that could have ruled it out as a cause.

Despite this, the Court concluded that "the materiel state of the Ship did not contribute to the grounding or subsequent loss of the Ship" (¶234). I cannot see what would have allowed them to reach that conclusion.

1. R. J. Heuer, Psychology of Intelligence Analysis, CIA Center for the Study of Intelligence, 1999, chapter 8 — Analysis of Competing Hypotheses. Step 1 is to generate the full set of hypotheses before weighing evidence.