File size: 2,364 Bytes
7055320 7064099 7055320 7064099 7055320 7064099 7055320 7064099 |
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 |
---
title: Engine Predictive Maintenance App
emoji: "🛠️"
colorFrom: purple
colorTo: pink
sdk: docker
pinned: false
---
# 🛠️ Smart Engine Predictive Maintenance App
This interactive Streamlit application predicts whether an engine is likely to be **Faulty (1)** or **Normal (0)** using real-time sensor readings.
It is designed to support **preventive maintenance decision-making** by identifying engines at higher risk of failure before breakdown occurs.
---
## ✅ Key Features
- **Single Engine Prediction** using manual sensor inputs
- **Probability-based output** for Faulty / Normal (where supported by the model)
- **Feature engineering built-in** (the app automatically computes engineered features to match the training schema)
- **Download engineered input row** as CSV for traceability
- **Bulk CSV Prediction** (upload a CSV and generate batch predictions)
- **Download bulk predictions** directly from the UI
---
## 🧠 Model Details
- **Algorithm:** Gradient Boosting Classifier
- **Training Data:** Engine sensor telemetry dataset
- **Target Variable:** `Engine Condition`
- `0 = Normal`
- `1 = Faulty`
**Reference Metrics (from model evaluation):**
- Recall (Faulty): ~0.84
- ROC-AUC: ~0.70
- PR-AUC: ~0.80
---
## 🧾 Required Input Features (Single & Bulk)
Your CSV or manual inputs must include **only the raw sensor columns** below:
1. `Engine rpm`
2. `Lub oil pressure`
3. `Fuel pressure`
4. `Coolant pressure`
5. `lub oil temp`
6. `Coolant temp`
The app computes additional engineered features internally (ratios, indices, and warning flags) to align with the model training pipeline.
---
## 📦 Bulk Prediction Instructions
1. Upload a CSV file with the 6 required raw sensor columns listed above.
2. The app will generate:
- `Predicted_Class` (0/1)
- `Faulty_Probability` (if available)
3. Download the results using the provided **Download Bulk Predictions CSV** button.
---
## 🚀 Deployment
This Space uses a Docker-based deployment with Streamlit running on port **8501**. Hugging Face automatically maps ports during deployment.
---
## 🔗 Project Links
- **Model Hub:** `simnid/predictive-maintenance-model`
- **Dataset Hub:** `simnid/predictive-engine-maintenance-dataset`
- **GitHub Repository:** *(add your repo link here once finalized)*
---
|