---
title: "Shipbuilding Engineering Change: A Worked Case | Veldarium"
canonical: https://www.veldarium.com/case
html: https://www.veldarium.com/case
description: "One synthetic change followed through the whole system: what it reaches, the seats that must act, the named person who decides, what reaches the floor, and why a machine is not yet capacity."
last_updated: 2026-10-12
representation: registry-derived Markdown
---

# The Worked Case

> One synthetic change followed through the whole system: what it reaches, the seats that must act, the named person who decides, what reaches the floor, and why a machine is not yet capacity.

A change has entered the vessel program. Synthetic, on a concept vessel. Every record, cell, figure and person on this page is a teaching fixture; Veldarium operates no yard, crew or machine. What is real is the logic that computes each answer, and it runs in Veldarium’s code.

Program state for VS-001 on 2026-08-20, mission to vessel:

- **Mission and requirement** (VS-001): Concept profile only
- **Design** (Drawing DR-118): Route revised to Rev D
- **Configuration** (Rev C governs): Rev D pending release
- **Engineering change** (CFG-029): Entered the program
- **Material** (MAT-441 · Seawater spool set kit): 2 of 9 spools held for Rev D
- **Work package** (WP-019): Not ready
- **Production** (3 workcells): 3 of 3 may not start
- **Inspection** (INS-124): Re-issued at Rev D on release
- **Accepted physical state** (DIM-B30-019): 0 of 2 observations accepted
- **Vessel** (Block B30): 3 items wait downstream

The questions the case answers, in order:

1. What changed?
2. What does it affect, and what not?
3. Who needs to know?
4. What evidence governs?
5. Who has authority to decide?
6. What work remains safe to release, and what must stop?
7. What happens after that decision?

## Something changed

CFG-029: Route revised Rev C → Rev D on the seawater cooling spool set. 3 of 3 workcells could start before the release, and 0 can after it.

## What it reaches

- Known affected: WP-019 · Seawater cooling spool set (work package). Assessed (EV-IMP-019): The heat exchanger inlet nozzle moves, so the route the package installs moves with it. It is re-released at Rev D before anyone starts.
- Known affected: MAT-441 · Seawater spool set kit (material). Assessed (EV-IMP-441): 2 of 9 spools re-specified. The closure spool shortens and its neighbour re-angles; the other spools keep their definitions. The re-specified items are held until they clear against Rev D.
- Known not affected: WP-014 · Machinery foundation structure (work package). Assessed (EV-IMP-014): The change moves a nozzle on the heat exchanger, not the exchanger: its foundation, seating and bolting are unchanged, so no structural weld moves.
- Impact not established: WP-030 · Adjacent zone electrical routing (work package). No impact assessment is on record. It is governed by DR-118, which the change re-specifies, and in the same zone as WP-019, so the impact is not established, and it stays unresolved rather than assumed clear.

## Who needs to know

- **Production supervisor.** Can my crew begin WP-019 this morning? The system can establish: Not ready. WP-019 may not proceed. Known open: configuration and material. It will not guess: Whether the crew can work around an open gate. A known open gate is not waived by assertion, and a gate nobody can establish is never read as a pass. A person decides: M. Okafor holds the release, and under this verdict the one action open is to hold.
- **Chief engineer.** What did Rev D reach, and what did it not? The system can establish: CFG-029 reaches 4 objects by topology, 3 of them assessed, and 1 package waits downstream of the work it affects. It will not guess: Whether the adjacent zone electrical routing is affected. No impact assessment is on record, so it stays not established instead of unaffected. A person decides: Disposition of CFG-029: what proceeds, what is re-released at Rev D, and what is withdrawn. Owned by engineering / configuration control.
- **Production planner.** Which gate holds WP-019, and what moves it? The system can establish: 6 gates evaluated. Configuration binds, owned by engineering / configuration control. It will not guess: A date. The engine says which gate binds and who owns it; it does not forecast when receipt or release will happen. A person decides: Whether to resequence the crew onto other released work while the gate is open.
- **Quality and inspection.** What will it take to accept the dimensional hold point on WP-019? The system can establish: 0 of 2 required observations accepted on file, from independent sources. DIM-B30-019 (dimensional acceptance) requires 2 accepted observations of laser-scan or machine-vision; 0 are on record. The evidence present may be entirely valid and still not be enough. The work has not run, so there is nothing yet to measure: this holds acceptance, never the start. It will not guess: To let one valid reading stand in for two, or let an instrument accept its own result. A person decides: Disposition of the hold point. The inspector accepts; the scan only measures.
- **Supply chain.** Is the kit for WP-019 on hand and accepted at the governing revision? The system can establish: 7 of 9 spools stand at Rev D; 2 are held, because Rev D re-specifies them. The review board dispositions those: SP-019-08 is remade; SP-019-09, the closure spool, is trimmed and re-tested. It will not guess: To clear a held item against the superseded revision, or to treat a purchase order as material on hand. A person decides: Whether to expedite the remade spool or hold the kit. A substitute is an engineering change, not a supply decision.
- **Program leadership.** What does this change do to block B30 and the capacity it needs? The system can establish: Do not release WP-019. Governed readiness and capability qualification are blocking, whatever the other domains report. Physical evidence holds acceptance of the work once it has run, not its start. 3 items wait downstream. WP-019 queues for pre-outfit (pipe & spool): qualified capacity exists and is committed. It will not guess: A delivery date or a cost. Nothing in the case measures either, so neither is computed. A person decides: Priority and schedule risk across the block. Release stays with the named release authority.
- **Supplier.** Is my source qualified for the work WP-019 is written in? The system can establish: Candidate source A (synthetic) is at first article / pilot for REQ-ALLOY-SEAM. One good part is not sustained capability at rate. Reaching approved source requires: Qualification closed out; approval recorded against the capability class and revision effectivity. It will not guess: To count one good part as sustained capability. A person decides: Approval of the source, owned by supplier quality, with welding engineering.

## What reasoning determined, and who decides

- What changed: Route revised Rev C → Rev D on the seawater cooling spool set.
- The program state it reasons over: After Rev D is released, WP-019 re-released at it and every item accepted.
- Which consequence paths to evaluate: 5 domains: governed readiness, finite qualified capacity, capability qualification, physical evidence, nonconformance.
- What deserves attention: 1 of the 3 domains that decide the start blocks release: capability qualification. Physical evidence holds acceptance, not the start.
- What evidence is missing: 2 accepted observations at the dimensional hold point, from independent sources, once the work has run.
- Where uncertainty prevents an answer: WP-030 · Adjacent zone electrical routing: impact not established, and left open.
- What to propose: Hold. It may not release.
- Who must decide: M. Okafor, Production release authority · B30.

VeldMind proposes; a named human authorizes; VeldBuild records. Neither system releases work.

## The floor

- CFG-029 is open and 2 of 9 spools are held for Rev D: Not ready, configuration binds. Reaches the floor: no. Two gates are known open. The work cannot start, and the system says which gate binds first.
- Rev D is released and WP-019 re-released at it; the 2 spools are still held: Not ready, material binds. Reaches the floor: no. Closing the change does not release the work. It moves which gate binds.
- Rev D is released, WP-019 re-released at it, and every item is accepted: Ready. Hold: capability qualification blocks the start. Reaches the floor: no. Every gate is established and the package reads ready. Ready is not authorized: capability qualification holds the release, because no weld procedure is qualified for the spool alloy. Acceptance evidence comes once the work has run, and it holds acceptance, not the start.

## From release to delivery

- **Engineering release:** Released work packages, a material list and cutting nests, each naming the revision it was released at.
- **Material procurement:** Accepted plate, profiles and equipment, each traceable to its heat, its certificate and the revision that ordered it.
- **Fabrication:** Cut, marked and formed parts and pipe spools, each traceable to its certificate and the revision it was nested from.
- **Assembly (the worked case):** Structural blocks closed to their neighbours’ interfaces, welds inspected, each with its dimensional record, and outfitted while open: foundations, seats and pipe hangers before the coat, pipe, ducting and trays after it.
- **Integration:** A closed hull, its machinery aligned and its systems joined across every block joint.
- **Inspection:** An accepted hull record: tightness and pressure tests passed, weld inspection closed and each hold point signed by quality and the class surveyor.
- **Launch:** A hull afloat at the outfitting quay, with no ingress, its draughts and trim recorded against the design.
- **Commissioning:** Systems proven in harbor and at sea, with signed trial records and a list of the defects still open.
- **Delivery:** A delivered vessel: the signed protocol of delivery and acceptance, its certificates, flag registration and the as-built record.

Nameplate is not capacity. For the synthetic plasma cutting cell (effective capacity = staffed hours × qualified rate × availability × first-pass yield): Staffed hours in the window: 160; Nameplate (160 h × 15 plate parts per hour (qualified rate)): 2400; After availability (× 0.85 (−360)): 2040; Accepted first time (× 0.98 first-pass yield (−40.8)): 1999.2 plate parts, against demand of 1600. Only an asset that is released counts as capacity.

## The industrial base

Where can this vessel actually be built? Unknown: no facility record meets the publication standard yet. In the synthetic capacity model, panel line weld binds first, then structural weld (robotic + manual). What any yard depends on:

- **Yards:** Berths, cranes and covered space decide what can be built at once.
- **Technical authority:** Nothing is released or accepted until an engineering authority says so.
- **Long-lead equipment:** Propulsion and power arrive in sequence or erection waits.
- **Suppliers:** A part from an unqualified source cannot be accepted, whatever its quality.
- **Qualified processes:** A machine without a qualified procedure produces nothing that counts.
- **Tooling and fixtures:** Work waits on the fixture, not only on the crew.
- **Material:** Plate at the wrong grade, or without a certificate that matches it, cannot be cut, whatever the tonnage in stock.
- **Qualified labor:** Certified trades take years to train and cannot be hired from demand alone.
- **Facilities and utilities:** Power, extraction and layout bound what a machine can do.
- **Inspection:** Acceptance gates every stage, not only the last.
- **Transport:** An oversized load with no route, or a missed heavy-lift slot, holds every stage waiting on it.

## Why a yard would pay for this (Target)

Every mechanism is a target: shown on synthetic records, measured on none.

- Less time rebuilding the current state by hand from several systems. Measured by: Time from a question to an answer the reviewer accepts, against a baseline agreed before anything is connected.
- Earlier sight of what a change reaches, and of what nobody has assessed. Measured by: Affected objects found before work started against those found after, on reviewed changes.
- Fewer packages released while a condition is still open. Measured by: How often released work was later found blocked, before and after.
- The binding constraint and its owner named, instead of a status color. Measured by: Whether the named owner was the person who actually cleared the gate.
- Evidence that stays attached to the decision it supported. Measured by: Whether every source cited in a decision survives an independent review.
- Clear ownership of each consequential decision, recorded where the next person can check it. Measured by: Whether the release authority acts on the answers without authority moving off the people who hold it.

The exchange a design partner would make is on https://www.veldarium.com/work-with-us.

## Contact

To discuss a first VeldBuild deployment, or the company: [chris@veldarium.com](mailto:chris@veldarium.com). Name the release decision your team rebuilds by hand. Do not send confidential information in the first message.

## Sitemap

- [Full site map](https://www.veldarium.com/sitemap.md)
- [Machine context](https://www.veldarium.com/llms.txt)
