Customer Stories
Airline A
Data Analytics on AWS Cloud
Improve data utilization by enhancing Hadoop data schemas for a data lake.
- ClientAirline A
- ProjectAWS-Based Engine Failure Prediction
- IndustryServices
Early Risk Detection and Prediction Using Anomaly Detection Models
- Purpose & Background
-
- Detect and predict aircraft aging to improve safety and equipment efficiency
- Analyze data collected from numerous sensors
- Develop and validate applicable analytical models
- Refine and apply models to specific needs
- Key Features
-
- Develop and validate models to detect anomalies from aircraft sensor data
- Validate models with operational data and assess real-world application
- Technology
-
- Data storage, processing, and analysis on AWS Cloud
- R-based analytical model development
Process
-
01
Current Status Analysis
Sensor data exploration
-
02
Pattern Extraction & Feature Engineering
- Feature 1 : Anomaly Pattern Mining - Cumulative Changes
- Feature 2 : Standardization Calculation - Cumulative Changes
-
03
Combined Prediction
Cluster analysis : Normal cluster extraction and combined anomaly prediction
