Blister defects traced to one curing press. And a cure spec with too little margin.
A tyre manufacturer traced a rise in sidewall blisters to a curing press running at the edge of a recently revised cure spec. The run proposed a scoped hold and an SOP step, and the quality manager approved both.
- Customer
- A tyre manufacturer
- Scope
- On-premises · one plant
- Published
What was in the way
Sidewall blisters are found late, at final inspection or worse, in the field. By the time a trend is visible, weeks of production have passed through the same presses. The plant knew blisters were rising in one size range. It did not know why.
The evidence was split across the defect tracking system, the press curing logs, the compound mixing records and a folder of cure specs and SOPs with revision histories. Each team had a theory and the data to support it, but nobody had all of it in one view.
The plant also wanted any hold to be justified and scoped precisely. A hold that is too wide costs output. One that is too narrow lets defects ship.
How Agent Hive was set up
- 01
Connect the plant records
Defect records, press curing logs, mixing records and the document store for cure specs and SOPs were connected read-only. Barcode genealogy links a cured tyre back to its green tyre and compound batch. - 02
Find where defects cluster
The agent joined defects to curing history and found the concentration on two presses in one bay, then compared their cure cycles against the bands in the current spec. - 03
Read the spec history
Cure spec revisions were read as documents, not just numbers. The recent revision that shortened the cycle was linked to the start of the defect rise, with page references. - 04
Scope the hold with genealogy
Lot genealogy gave the exact set of tyres cured on the suspect press since the revision, split by what was still on site and what had shipped. - 05
Propose; the owner approves
The run proposed the hold, a temporary SOP step and a spec review. The plant quality manager approved them, with one edit to the scope, and the SOP change was versioned with its evidence.
What changed
- From question to approved hold
- 1 day
- was about two weeks
- Presses implicated
- 2
- out of the full bay
- Fewer blisters on that size
- 62%
- six weeks after the change
- SOP step added
- 1
- versioned, owner approved
The press, the cycle and the spec
The investigation started from the defect records: blisters on the sidewall, concentrated in one size range over three weeks. The agent joined those records to the curing log through the green tyre barcode and found that most affected tyres had been cured on two presses in the same bay. On one of them, the cure cycle showed platen temperature sitting at the low edge of its band for the early part of the cycle.
Curing press cycle vs spec band. One cure on press T2-B08 to CURE-PCR-SN2 v4: close, shaping at about 0.9 bar, HP steam at about 16 bar, nitrogen at about 24 bar, exhaust, open, over 13.5 minutes, each against its spec band (illustrative values). Every supply stays inside its band; state of cure ends at 0.97, above the 0.90 limit, and the tyre goes on to final finish.
Being inside the band was the clue, not the alibi. The cure spec had been revised a month earlier to shorten the cycle for that size, and the revision narrowed the margin on the early temperature ramp. Tyres cured on a press that ran cool were still technically in spec, but with too little margin for the compound in use. The agent laid the spec revision, the press history and the defect timeline side by side, with each claim linked to its record.
Scoping the hold
A hold is only useful if it is the right size. Using lot genealogy, the agent traced which compound batches, green tyres and cured tyres had passed through the suspect press since the spec revision, and which had already shipped to the warehouse.
Lot genealogy and traceability. Genealogy of 27 records in 5 steps: Materials, Components, Green tyres · TBM-07, Cured tyres · curing, Dispatch. T-3302/17 (uniformity reject · blister) is traced back along its made-from links: T-3302/17 ← G-3302/17 ← TR-0418 ← CB-1187. Source: CB-1187. Made from the same source and held too under SOP-QA-001: G-3302/15, T-3302/15, G-3302/16, T-3302/16, PAL-62, G-3302/18, T-3302/18.
A proposal, then an owner
The output was a proposal, not an action: hold the scoped tyres for inspection, add a temporary SOP step requiring a temperature check on that press at cycle start, and send the spec revision back to process engineering for review. The plant quality manager approved the hold and the SOP step, and edited the scope to include one extra shift. The change to the SOP was versioned with a link to the investigation that justified it.
Every press was in spec on paper. Seeing the spec revision next to the defect timeline is what changed the conversation.