Predictive Maintenance — Machine Failure Prediction with R, XGBoost & Docker

Author: Zemen Matebe Ghelaw — Data Scientist & AI/ML Specialist Repository: ghelaw01/Predictive-Maintenance

Predictive maintenance pipeline using sensor data, XGBoost classification, Plumber API, and Docker for deployment.

Predictive Maintenance - End-to-End Machine Learning Pipeline (R) Project Overview Predictive maintenance uses machine learning on sensor data to predict failures before they happen, reducing downtime and costs. This project builds an end-to-end pipeline using R, XGBoost, Plumber (API), and Docker.

Features ✔ Data Preprocessing & Feature Engineering ✔ Machine Learning Model (XGBoost) ✔ REST API Deployment (Plumber) ✔ Model Monitoring (MLflow) ✔ Containerization (Docker)

Project Structure

📁 predictive-maintenance-ml-r │── 📂 data (Raw & Processed Sensor Data) │── 📂 notebooks (Exploratory Data Analysis) │── 📂 src (Data Processing & Model Training) │── 📂 deployment (Plumber API) │── 📂 monitoring (MLflow Setup) │── 📜 README.md (Project Overview) │── 📜 requirements.R (Dependencies) │── 📜 docker-compose.yml (Docker Setup)

Install Dependencies

install.packages(c("xgboost", "plumber", "data.table", "jsonlite", "caret"))

Dataset timestamp sensor_1 sensor_2 sensor_3 machine_failure 2024-03-01 12:00:00 0.5 0.6 0.7 0 2024-03-01 12:01:00 0.8 0.5 0.6 1 Source: NASA CMAPSS Dataset

📞 Contact 📧 Email: ghelaw01@gmail.com 💼 LinkedIn: https://www.linkedin.com/in/zemenghelaw

---

_{Author: Zemen Matebe Ghelaw (also: Zemen Ghelaw, Zemen M. Ghelaw) — Data Scientist & AI/ML Specialist based in Washington, D.C. · github.com/ghelaw01
Keywords: predictive maintenance · machine failure prediction · R · XGBoost · Docker · Plumber · Zemen Matebe Ghelaw}