> For the complete documentation index, see [llms.txt](https://intelfactor.gitbook.io/docs/llms.txt). Markdown versions of documentation pages are available by appending `.md` to page URLs; this page is available as [Markdown](https://intelfactor.gitbook.io/docs/home/closed-loop-rca/local-rca-engine-4-layers.md).

# Local RCA engine (4 layers)

How RCA hypotheses are built locally and made auditable.

IntelFactor closes the loop locally:

defect → root cause → corrective action → verified outcome

### Layer 1: Defect pattern accumulator

* Stores inspection events locally (rolling window).
* Detects anomaly spikes per defect type, per station.

### Layer 2: Process parameter correlator

* Compares anomalies against process parameters (targets, ranges, drift).
* Outputs explainable correlations.

### Layer 3: On-device explanation model

* Converts structured context into RCA hypotheses.
* Runs on anomaly triggers, not per-frame.

### Layer 4: Closed-loop action recommender

* Produces SOP-linked corrective actions.
* Captures operator feedback.
* Produces verified training triples.