The Steel watch

What this watch is

Turning drawings into hulls, and hulls into a process that repeats.

The first hull was built by hand, by people learning as they went, and it shows. It floats, it works, it earns. It is also a first article in every sense: the sequence was discovered rather than designed, and almost nothing about how it was built would survive contact with a tenth unit.

That is the work. Not making the second hull better — the second will be better no matter who builds it. Making the tenth one cheap, and being able to show that the answer came from a system rather than from someone working weekends.

Every load-bearing fitting on a Thalor hull should carry a documented rating with a calculation and a physical test behind it. Establishing that as ordinary practice, so it happens on every unit without anyone remembering to ask, is on this watch.

What you are inheriting

Honestly: one hull, one welder engaged per job, and no jigs.

There is a complete engineering record for the first unit — principal particulars, load conditions, stability, structural calculations, a testing program, and a findings register with seven entries. It is hand-auditable and every input is in it. It is not PE-stamped and it is not classed.

There is a shop. Its full capability is not yet documented; whether certain machining operations happen in-house or get bought is an open question, and the fabrication plan currently carries both routes because nobody has answered it.

There are no fixtures, no cut lists that survived the build, no hours-by-operation data, and no cost curve — only a total. Building the instruments that produce those is the first quarter of this job.

The numbers you own

A watch means a set of numbers that are yours, that you publish weekly, and that other people plan against.

Cost per hull. Not the first-article cost, which is history. The replicable cost, with the learning curve shown and the non-learning floor named separately.

Hours by operation. The only thing that turns a learning curve from an assertion into evidence. Nobody has ever measured this here.

Days from steel on the ground to hull in the water.

Maintenance and repair hours per hull-year in service. The number nobody collects until it hurts.

Occupancy, price, and demand belong to other watches. If you find yourself owning them, something has gone wrong with how we divided the work.

The problems

These are real and open. We are not testing whether you get them right — several have no known right answer. We want to know whether you have opinions and where they come from.

1. The first hull came in materially lighter than its design estimate — enough that the difference is a finding, not a rounding error. Where does an error like that most plausibly live in a weight estimate for a small welded steel hull, and how would you narrow it down with the vessel already in the water?

2. Analysis found the governing scantling before launch. The fix was adding transverse structure while the hull was still open, rather than during a later shipyard period. On the second unit: do you change the design so the added members are unnecessary, or keep them and change nothing? Argue the weight, cost, and labor either way.

3. Three independent watertight compartments, each of which must hold an air test on its own. Describe a welding sequence in which tightness is a by-product of how the hull goes together rather than something discovered at test and reworked.

4. Name the first fixture you would build. Not the most useful one eventually — the first one, when there is one welder, no jig plate, and the next hull needs to start.

5. Retrofitting the first hull is hard: it is afloat, and some spaces cannot be reached without confined-space entry. What provisions go into the next hull while it is open and flat, and how do you decide what earns the weight?

6. The tenth hull has to cost what the second one promised. What do you measure during the second build so that promise is falsifiable rather than aspirational?

Your first ninety days

By day 14 you have read the engineering record and the findings register, and you have been aboard.

By day 30 you have produced your own build plan for the next hull — the sequence you would actually use — with a written list of what you would change from the first one and why. It will disagree with what exists. That is the point of hiring you.

By day 60 the first fixture exists, and a real build is producing hours by operation for the first time in this company's history.

By day 90 you are publishing the Steel report weekly, and you can state what the tenth hull costs with the evidence behind the number.

Why this might be the wrong job for you

You want a crew. There isn't one. You would be building it, and for a while you will be swinging the grinder yourself.

You want the drawing to be right before you cut. It won't be. You will be correcting it in steel and writing down what you found.

You want a fully equipped shop. Some of what you need does not exist here yet, and part of your job is deciding what to buy versus what to send out.

You would rather build one beautiful hull than ten identical ones. That is a real and honorable preference and it is the wrong one for this watch. This job is about the tenth.

You do not want your mistakes written down. Ours are, with names attached. Yours would be too.

One problem before you write

Use whatever tools you use at work. We do.

The arithmetic is not the test. A model will do that part in seconds and we know it. What we are reading is which answer you believe, what you would do next, and what you noticed that nobody told you to look for. That part is still yours.

Our first hull is twenty feet by ten, flat-bottomed, and floating in seawater.

During design we estimated its lightship weight at fifteen thousand pounds. Afloat and empty, it draws about ten inches.

Those two facts disagree.

Tell us by how much. Tell us which one you believe. Tell us how you would settle it.

Half a page. Longer is not better.