Recovering supply through a Tier 3 mill fire

An upstream constraint, several tiers from the line

A fire damaged the hot mill process that fed Lucid's entire aluminum supply base. The constraint sat several tiers upstream of vehicle production, above the parts it would eventually starve.

My job was to convert that upstream disruption into exact production exposure, which parts, which suppliers, which build dates, which runout windows, fast enough to act before it became a line down event. The risk was never one supplier or one part. It was a multi tier visibility problem.

The limiter was raw coil, several tiers above the parts it fed. Availability at the mill did not mean parts at the plant.

The limiter sat above the parts at risk

1. Depth of the constraintThe limiter was raw coil, several tiers above the parts that would run out. Coil at the mill still had to become gauge, width, and a stamped part before it protected a build.
2. A moving recovery windowThe recovery estimate kept sliding, 16 weeks to 20, so no single mitigation could be trusted to hold.
3. Fragmented signalUpdates arrived across calls, emails, and supplier specific files. Each of the 10 nodes reported finished goods, work in process, raw material, and shipment timing differently.
4. Technical gates outside my controlSpecialty grades tied to crash critical structural parts could not be substituted on planning authority alone. Engineering and Quality owned technical release.
5. Cost discipline under pressurePremium air freight could protect timing but could not become the default answer. Recovery cost had to be held down while production was protected.

The tracker and the dates were mine; release was not

I owned the Clear to Build tracker for the affected scope and the recurring point of contact across the supplier tiers. Mitigations here were recommendation only. I presented them in a recurring global supply management (GSM) review, where they were approved and occasionally escalated to a vice president. A standing premium freight limit in the low five figures governed the day to day calls; anything larger I escalated with the figures behind it.

From raw tonnage to a part level runout date

1. Standardized supplier reportingForced every node's update into one operating format, the least sophisticated thing I did and close to the most useful. Reporting that cannot be compared cannot be prioritized.
2. Corrected the system signalCleaned inventory, advanced shipping notice (ASN), and master data errors out of the tracker inputs. Bad data creates false confidence and aims the team at the wrong limiter.
3. Mapped raw material to partsTied grades, gauges, widths, yield assumptions, and supplier conversion timing back to part level demand, so upstream coil read as buildable coverage instead of a raw tonnage number.
4. Expanded runout visibilityMoved the conversation from "aluminum is constrained" to "this specific part runs out on this date, against this build window, unless this action lands".
5. Ran parallel recovery tracksAllocation, alternate mills, coil reallocation, substitute material, expediting, and routing, each evaluated against a physical milestone rather than a supplier promise.
6. Escalated the true limiterRouted each gap to the function that could remove it: supplier capacity, logistics, data, or engineering.

Three decisions, by kind

The signal kept tightening

The affected builds kept running through the gap. Availability at the mill did not mean parts at the plant, so the recovery was won by tightening the signal and routing each gap to whoever could remove it. Premium freight stayed the dominant recovery cost throughout.