# unbora/igbt-phm architecture

> Architecture diagram of the GitHub repository unbora/igbt-phm, generated by GitDiagram from its file tree, README and a sample of source files (last updated 2026-10-03). It is an AI-made overview: check details against the source.

- Interactive diagram: https://gitdiagram.com/unbora/igbt-phm
- Repository: https://github.com/unbora/igbt-phm

## Overview

This is an educational NASA IGBT prognostics-and-health-management project: learners inspect aging measurements, derive candidate features, assess degradation trends, and study remaining-life prediction. The implemented runtime prototype loads steady-state MAT records, creates windowed features with an observation-horizon proxy target, evaluates device-held-out regressors, and writes tables, a plot, and a model. README-described notebook phases provide broader exploratory and learning capabilities, but their individual source files are not in the supplied tree. The prototype’s target is not verified physical RUL.

Main entry points: [`src/rul.py`](https://github.com/unbora/igbt-phm/blob/main/src/rul.py) (Workflow Runner), [`notebooks`](https://github.com/unbora/igbt-phm/tree/main/notebooks) (PHM Notebooks), [`src/features.py`](https://github.com/unbora/igbt-phm/blob/main/src/features.py) (Window Features), [`src/data_loader.py`](https://github.com/unbora/igbt-phm/blob/main/src/data_loader.py) (Observation Loader)

## Diagram (Mermaid)

```mermaid
flowchart TD

subgraph group_learning["PHM Learning"]
  node_notebooks["PHM Notebooks"]
  node_rawdata[("Aging Dataset")]
end

subgraph group_runtime["RUL Prototype"]
  node_runner["Workflow Runner<br/>[rul.py]"]
  node_loader["Observation Loader<br/>[data_loader.py]"]
  node_features["Window Features<br/>[features.py]"]
  node_evaluation["Device Evaluation<br/>[evaluation.py]"]
end

subgraph group_outputs["Artifacts"]
  node_featurefiles["Feature Tables"]
  node_metrics["Metrics"]
  node_figures["Prediction Plot"]
  node_model["Saved Model"]
end

node_learner(("Learner"))

node_learner -->|"studies"| node_notebooks
node_learner -.->|"runs"| node_runner
node_notebooks -->|"examines"| node_rawdata
node_runner -->|"loads"| node_loader
node_loader -->|"reads MAT"| node_rawdata
node_runner -->|"builds"| node_features
node_features -->|"uses signals"| node_loader
node_runner -->|"evaluates"| node_evaluation
node_evaluation -->|"uses columns"| node_features
node_runner -->|"writes"| node_featurefiles
node_runner -->|"writes"| node_metrics
node_runner -->|"saves plot"| node_figures
node_runner -->|"saves model"| node_model
node_notebooks -.->|"studies feature stages"| node_features
node_notebooks -.->|"studies RUL readiness"| node_evaluation

click node_notebooks "https://github.com/unbora/igbt-phm/tree/main/notebooks"
click node_rawdata "https://github.com/unbora/igbt-phm/tree/main/data/raw/IGBTAgingData_04022009"
click node_runner "https://github.com/unbora/igbt-phm/blob/main/src/rul.py"
click node_loader "https://github.com/unbora/igbt-phm/blob/main/src/data_loader.py"
click node_features "https://github.com/unbora/igbt-phm/blob/main/src/features.py"
click node_evaluation "https://github.com/unbora/igbt-phm/blob/main/src/evaluation.py"
click node_featurefiles "https://github.com/unbora/igbt-phm/tree/main/data/features"
click node_metrics "https://github.com/unbora/igbt-phm/tree/main/results/metrics"
click node_figures "https://github.com/unbora/igbt-phm/tree/main/results/figures"
click node_model "https://github.com/unbora/igbt-phm/tree/main/models"

classDef toneNeutral fill:#f8fafc,stroke:#334155,stroke-width:1.5px,color:#0f172a
classDef toneBlue fill:#dbeafe,stroke:#2563eb,stroke-width:1.5px,color:#172554
classDef toneAmber fill:#fef3c7,stroke:#d97706,stroke-width:1.5px,color:#78350f
classDef toneMint fill:#dcfce7,stroke:#16a34a,stroke-width:1.5px,color:#14532d
classDef toneRose fill:#ffe4e6,stroke:#e11d48,stroke-width:1.5px,color:#881337
classDef toneIndigo fill:#e0e7ff,stroke:#4f46e5,stroke-width:1.5px,color:#312e81
classDef toneTeal fill:#ccfbf1,stroke:#0f766e,stroke-width:1.5px,color:#134e4a
class node_notebooks,node_rawdata toneBlue
class node_runner,node_loader,node_features,node_evaluation toneAmber
class node_featurefiles,node_metrics,node_figures,node_model toneMint
class node_learner toneIndigo
```

## Components

### PHM Learning

- [PHM Notebooks](https://github.com/unbora/igbt-phm/tree/main/notebooks): `notebooks`
- [Aging Dataset](https://github.com/unbora/igbt-phm/tree/main/data/raw/IGBTAgingData_04022009): `data/raw/IGBTAgingData_04022009`

### RUL Prototype

- [Workflow Runner](https://github.com/unbora/igbt-phm/blob/main/src/rul.py): `src/rul.py`
- [Observation Loader](https://github.com/unbora/igbt-phm/blob/main/src/data_loader.py): `src/data_loader.py`
- [Window Features](https://github.com/unbora/igbt-phm/blob/main/src/features.py): `src/features.py`
- [Device Evaluation](https://github.com/unbora/igbt-phm/blob/main/src/evaluation.py): `src/evaluation.py`

### Artifacts

- [Feature Tables](https://github.com/unbora/igbt-phm/tree/main/data/features): `data/features`
- [Metrics](https://github.com/unbora/igbt-phm/tree/main/results/metrics): `results/metrics`
- [Prediction Plot](https://github.com/unbora/igbt-phm/tree/main/results/figures): `results/figures`
- [Saved Model](https://github.com/unbora/igbt-phm/tree/main/models): `models`

### Other components

- **Learner**

## Connections

- Learner → PHM Notebooks: studies (evidence: `README.md`)
- Learner → Workflow Runner: runs
- PHM Notebooks → Aging Dataset: examines (evidence: `README.md`)
- Workflow Runner → Observation Loader: loads (evidence: `src/rul.py`)
- Observation Loader → Aging Dataset: reads MAT (evidence: `src/data_loader.py`)
- Workflow Runner → Window Features: builds (evidence: `src/rul.py`)
- Window Features → Observation Loader: uses signals (evidence: `src/features.py`)
- Workflow Runner → Device Evaluation: evaluates (evidence: `src/rul.py`)
- Device Evaluation → Window Features: uses columns (evidence: `src/evaluation.py`)
- Workflow Runner → Feature Tables: writes (evidence: `src/rul.py`)
- Workflow Runner → Metrics: writes (evidence: `src/rul.py`)
- Workflow Runner → Prediction Plot: saves plot (evidence: `src/rul.py`)
- Workflow Runner → Saved Model: saves model (evidence: `src/rul.py`)
- PHM Notebooks → Window Features: studies feature stages (evidence: `README.md`)
- PHM Notebooks → Device Evaluation: studies RUL readiness (evidence: `README.md`)
