Member-only story
Data to Diagnostics: Analysing Ship Engine Anomalies for Predictive Maintenance
Contents
- Introduction 🚢
- Methods 🧭
- Machine Learning for Anomaly Detection: Navigating with AI 🤖
(One Class SVM, Isolation Forests, PCA) - Results: Findings from the Expedition 🗺️
- Conclusions: Anchoring Our Discoveries ⚓
🚢Introduction
Certainty in the smooth running of a ships engine is paramount in avoiding costly breakdowns, maintaining fleet performance and ensuring the safety of the crew. Early detection of anomalies in functionality plays a crucial role in achieving this goal.
Given my deep-rooted association with maritime vessels, this analysis holds special significance as success will save relevant plays vasts amount of resources. It focuses on identifying and predicting anomalous behaviour in ship engines, using a comprehensive dataset of 19,535 observations. This data captures six critical Engine Features:
- Engine-RPM
- Lubrication Oil Pressure
- Fuel Pressure
- Coolant Pressure
- Lubrication Oil Temperature
