Lot L0803-A: purity low
purity was 98.2 wt% (spec ≥ 99) , sampled at SP-1 on 2026-08-03 07:59.
The leading hypothesis, with high confidence: TI-101.PV moved low.
TI-101.PVhigh TI-101.PV (E-101) ran 5.74 degC low from 2026-08-03 05:29, 50 min upstream of the sample point; purity moved when it arrived. [E004]PT-100.PVmedium PT-100.PV (S-feed) ran 1.02 bar low from 2026-08-03 05:27, 52 min upstream of the sample point; purity moved 55 min after the expected arrival. [E005, E002]PI-103.PVmedium PI-103.PV has a flatline finding in the lot window: value frozen at 5.968 for 90 min [E006]
Hypotheses
#1 TI-101.PV high
process · score 0.97 · path E-101 → R-101 → S-product · dead time 50 min
TI-101.PV (E-101) ran 5.74 degC low from 2026-08-03 05:29, 50 min upstream of the sample point; purity moved when it arrived.
| Test | Strength | Detail |
|---|---|---|
| effect size | 1.00 | median moved -5.74 (-17.3 reference scales) |
| onset | 1.00 | step lands +0 min from the quality onset (tolerance 10 min) |
| lag correlation | 0.89 | r = +0.89 over 1440 reference points |
Explains away:
TI-104.PV: its shift of -2.88 is 100% predicted from TI-101.PV (regression on the reference lots)
#2 PT-100.PV medium
process · score 0.46 · path S-feed → E-101 → R-101 → S-product · dead time 52 min
PT-100.PV (S-feed) ran 1.02 bar low from 2026-08-03 05:27, 52 min upstream of the sample point; purity moved 55 min after the expected arrival.
| Test | Strength | Detail |
|---|---|---|
| effect size | 1.00 | median moved -1.02 (-43.3 reference scales) |
| onset | 0.00 | step lands -55 min from the quality onset (tolerance 10 min) |
| lag correlation | 0.00 | r = -0.29 over 1440 reference points, sign does not fit the deviation |
E002: FIC-102 absorbed a disturbance: FI-102.PV held, output moved 9.1
#3 PI-103.PV medium
measurement fault · score 0.45 · path R-101 → S-product · dead time 45 min
PI-103.PV has a flatline finding in the lot window: value frozen at 5.968 for 90 min
Health warnings: H0001
Ruled out
| Candidate | Why |
|---|---|
FI-102.PV | held at setpoint by FIC-102 while its output moved 9.1 (+34.0 reference scales) |
TI-104.PV | its shift of -2.88 is 100% predicted from TI-101.PV (regression on the reference lots) |
PI-103.PV | did not move in the lot |
How this was worked out
| Lot window | 2026-08-03 00:00 to 2026-08-03 08:00 at S-product |
|---|---|
| Quality onset | 2026-08-03 06:19 (AI-201.PV) |
| Regime | GRADE = 1 |
| Reference lots | L0802-A (99.6), L0801-C (99.7), L0801-B (99.7) |
| Candidates tested | 5 tags, on a 60 s grid in material time |
| Confidence | Confidence: high when the score is at least 0.4 and 0.025 ahead of the next hypothesis, medium from 0.25. Test weights: effect size 0.5, onset 0.3, lag correlation 0.3. Weights fitted on the homeostat train split (85 scenarios), thresholds chosen on the homeostat validation split (85 scenarios) (homeostat 0.1.0a3, bench plant 4144102c25e7). |
- 1 earlier lots skipped: different regime
Data health on the candidate tags
| ID | Tag | Check | Severity | From | Detail |
|---|---|---|---|---|---|
| H0001 | PI-103.PV | flatline | warning | 2026-08-03 05:59 | value frozen at 5.968 for 90 min |