This project focused on integration of data from different sources (airport and airline data) and from different databases (safety-related data and operational data), and with the application of data mining techniques to identify risk patterns to intervene proactively on predicted risks.
Developed a state of the art pattern matching system that provides functionality for data…
integration and data mining. Data integration allows consideration of safety and operational data across different databases and different stakeholders. Data mining techniques can be used to detect patterns of variables (risk patterns) and combinations of values that affect the likelihood or severity of safety events.
The risk pattern matching function monitors continuously safety and operational performance indicators and attempts to match these to patterns identified from historical data. This solution can scale up to millions of records without impacting on performance. The risk intelligence distribution function informs relevant stakeholders of any patterns or alarms that were raised by the risk pattern matching function.
The Corballis Pattern Matching server received bird strike risk patterns, and was able to monitor simulated operational data in order to match and identify patterns in that data. Notifications about identified patterns were generated and forwarded through the PROSPERO system to an alerting tool for risk intelligence distribution.