Engineering Change Note Tracker

What this tool does

CM8-275 is a working register for engineering change notes — changes to a product, a part, an assembly or a drawing. You record what is changing and why, classify it, decide what happens to stock already made, track the approvals, and record the date the change became real in production. It then shows what is stuck, what that costs in frozen stock, and how far effectivity slips.

Everything runs inside this single file — no account, no upload, no network request — so drawing numbers, customer part references and supplier changes never leave this computer.

Why change control exists

A product that works is a set of matched decisions: this material, that tolerance, this supplier, that assembly sequence, these test limits. Most of those reasons were never written down. Change one without the others catching up and the product rarely fails immediately — it fails later, in the field, in a batch nobody can identify. A supplier moves a casting to a different foundry and mentions it on the phone. Each is reasonable alone, and each quietly breaks an assumption somewhere else. Six months later the warranty claims start, and the only question that matters — which parts have the change and which do not — has no answer, because there is no change note, no effectivity date and no stock decision on record. That is how a good part becomes a recall.

Change control does not stop changes. It makes them visible, deliberate and traceable.

The ECN lifecycle this tool models

• Raised — a problem or an opportunity written down against a specific part: a request, not a decision. Anybody may raise one; the register records who did.
• In review — the change is assessed: what it affects, what must be requalified, what it does to tooling, supply, stock in hand and parts already at customers. The class is decided here, and most of the real work happens here.
• Approved — the people who carry the consequences have said yes, in writing, with dates. An approved change nobody schedules will happen accidentally.
• Implementing — documents reissued, tooling updated, suppliers notified, stock dispositioned, people retrained. The longest stage, and the one nobody resources.
• Effective — the change is live. Parts made now have it.
• Rejected — it will not proceed. Keep the row: it stops the same idea returning every year.

The important word is effectivity. Approval is a signature; effectivity is the moment the change is real on the floor. A register that tracks approvals and stops there changes nothing, so the tool will not mark a change effective without an actual effectivity date.

Minor, major and critical

Classification drives everything else. These classes are generic; if your industry or customer contract imposes its own scheme, use theirs.

Minor. Form, fit and function unchanged, and the part interchanges completely with the one before it in both directions — a tidied work instruction, a tolerance relaxed to what the process already holds. It still needs its documents reissued and a revision, but no requalification, and old and new stock can be mixed.

Major. Form, fit or function moves. The new part is not freely interchangeable, or it behaves measurably differently. Expect requalification: a capability study on a new process, a first-article inspection at a new supplier, a re-run of a functional test. It almost always creates a stock question, because somewhere there will be both versions.

Critical. Safety, regulatory conformity or a customer's written approval is in play — a guard interlock, a pressure-containing wall thickness, a restricted substance, a customer-designated special characteristic. The tool will not save one without its approvals recorded. Expect a first-article submission and, where the contract says so, the customer's signature before a changed part ships. Effectivity traceability matters most here.

The stock disposition question everyone forgets

Every change to a part that already exists creates a question about the parts that already exist. There are five honest answers:

• No stock affected — nothing made yet, or nothing in hand is touched.
• Use existing stock first — run the old version out, then switch. Cheapest, and only valid if the old parts are genuinely acceptable. It pushes effectivity out to whenever stock runs down, which is why "use first" changes drift.
• Rework existing stock — bring the old parts up to the new revision. Costs labour, and needs its own instruction and inspection.
• Scrap existing stock — the honest answer when the old parts cannot be made right.
• Segregate pending decision — quarantined while somebody decides.

That last one is why this register carries a money column. Segregation is reasonable for a few days and expensive as a habit: the stock is frozen and its value sits on the balance sheet doing nothing while the change note ages. The stock value at risk tile totals the stock tied to every open change note with a disposition. Enter a figure even if it is rough — "we'll decide later" is the most expensive answer on the list, and a number beside it is the only way to see that.

Document sets that must move together

A change is not one document. Most touch a set that has to move as a unit: the drawing, which defines the part; the work instruction or routing, which is how it is made; the control plan, which says what is checked and to what limits; the bill of materials; plus tooling files, packing specifications and test procedures.

A change that updates the drawing but not the work instruction has failed: the drawing says one thing, the person building it follows another, and the discrepancy is found by a customer rather than by you. Treat the document list on the change note as the implementation checklist. If you cannot name them, the review is not finished.

Effectivity dates and traceability

The target effectivity date is what you planned; the actual is what happened, and the gap tells you whether your change process is predictable. A team that lands within a few days of target can promise a customer a date; one whose changes slip three weeks cannot.

That is only half the job. The other half is being able to say which parts have it — the first serial number, batch or works order carrying the change. Record it in the description or the approvals field if you have nowhere better. When a field problem appears two years from now, the first question is whether the unit was built before or after the change, and the answer has to be a fact.

Reading the reason mix

The reason donut diagnoses the organisation, not the change process. Customer requests, cost reductions and manufacturability changes are the normal traffic of a product being actively made.

A register dominated by defect corrections is saying something else: changes are arriving because the design or the supplier got it wrong first time, and you are paying for the same work twice. Heavy supplier and obsolescence traffic points upstream too — single-sourced parts, component lifecycles nobody watches. Neither is fixed by processing change notes faster; both are fixed before the change note exists.

This is not an IT change and release log

The two share a word and nothing else. CM8-121 is the IT change and release log: deployments, releases, rollbacks and emergency changes on systems and services. Its unit is a change window and its question is whether the service stayed up. This tool is product change control: its unit is a part revision, and its questions are what happens to stock already made and which serial number was first with the change.

How the figures are worked out

Two cautions. Days open on a rejected change note keeps counting, because there is no effectivity date to stop the clock — read it as the age of the row, which is why rejected rows are never flagged as ageing. And the average days to effective covers only completed changes, so a register full of stuck ones can still show a flattering average.

FAQ

Who should raise a change note? Anyone — design, quality, production, purchasing, a supplier, a customer. Restricting who may raise one reduces recorded changes, not changes.

New revision or new part number? The usual test is interchangeability. If the new part replaces the old one in every application, in both directions, it is a revision. If fitting it to an older assembly would be wrong, it needs a new part number so the two can coexist.

Can a change be effective before it is approved? No — and if it happened, record it with the real dates. That is exactly the event this register exists to make visible.

How long should a change note stay open? Set the ageing threshold on the Settings tab to what your business can sustain; 45 days is the default. The flag matters more than the number.

Saving your work

Change notes, settings and the report header are written to this browser's local storage as you type, and the toolbar shows the last save. That storage belongs to one browser on one computer — another browser, a private window or a site-data clean-up will not have it.

Treat Export .json as the real save; Import .json restores it anywhere. Export CSV gives you the register for spreadsheet work. Reset asks twice, then erases everything this tool has stored, with no undo. Treat exports as commercially confidential.

Accuracy & disclaimer

The arithmetic here is deliberately simple — date differences and sums — and the tool does it faithfully. Everything that matters sits underneath it. The register cannot tell whether a change was classified correctly, whether every affected document was found, whether the stock disposition was carried out, or whether the part that shipped really had the change in it. A row marked effective is a claim; the evidence lives in your reissued documents, inspection records and stock transactions.

Classification schemes, requalification and first-article requirements, customer notification duties and record-retention periods differ by industry, contract and country. This is an internal record-keeping and prioritisation aid, not engineering, quality or regulatory advice.