Five connected modules turn fragmented factory data into defect detection, yield prediction, equipment intelligence, root-cause evidence, and test traceability.
Rulnix Vision analyzes high-resolution inspection imagery to detect, classify, and localize defects across the wafer surface — then clusters them into spatial patterns engineers can investigate immediately.
Particle contamination, pattern deformation, edge damage, alignment error, and unknown anomalies.
Automatically clusters die-level defects into wafer-scale patterns — rings, edges, streaks, or clusters.
Tracks defect-class frequency across lots, tools, and shifts to surface emerging quality risks early.
Rulnix Yield scores wafer lots by risk, predicts production loss, and measures whether corrective action actually improved output — turning yield management into a closed loop.
Actual yield vs. target, last 30 days
Plasma Etcher 04 — before / after maintenance
Illustrative comparison — actual results vary by fab and corrective action.
Prioritized by predicted yield impact
| Lot ID | Predicted Yield | Expected Yield | Risk | Est. Loss | Confidence | Primary Risk Factor |
|---|---|---|---|---|---|---|
| B-241 | 83.7% | 94.1% | High | $127,000 | 89% | Contact failures — Etching Step 87 |
| C-108 | 89.2% | 94.6% | Medium | $64,000 | 81% | Particle contamination — Deposition |
| A-317 | 91.5% | 95.0% | Low | $21,000 | 76% | Minor alignment drift — Lithography |
| D-092 | 85.9% | 93.8% | Medium | $88,000 | 84% | Vacuum recovery delay — Etch |
Rulnix Equipment continuously scores machine health from vacuum, pressure, vibration, and cycle-time telemetry — surfacing degrading tools days before an unplanned failure.
Predicted issue: Vacuum-pump degradation
Inspect during the next scheduled maintenance window. Detected changes include increased pump vibration, longer vacuum recovery time, declining pressure stability, and a rise in minor alarms.
Rulnix CauseGraph connects wafer lots, machines, process steps, defect patterns, maintenance events, and test failures into one traceable knowledge graph — so engineers can see exactly why something happened.
Every relationship in this graph is generated from source manufacturing records — inspection logs, MES events, and equipment historian data — with full lineage back to the originating system for audit and review.
Rulnix Test Intelligence clusters electrical-test failures and traces them back through packaging, process, tooling, and inspection data to their shared manufacturing characteristics.
Select a category to trace it back to origin