Flight Data Analysis System for Automatic Event Detection
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current flight data analysis systems require extensive manual effort and resource allocation to analyze all recorded flight events, leading to a limited detection of potential issues due to the systematic analysis of only pre-recorded events, and the lack of automatic generation of new flight events or data sets.
Innovation Solution
A method that automatically detects and updates flight data sets by correlating values from previous flights, generating new subsets based on pairing probabilities, and iteratively refining these sets to enhance the detection of new flight events, reducing the workload for experts and enriching the database with relevant flight data.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Ease of operation
If systematic analysis of only pre-recorded flight events is performed, then expert workload is reduced through predefined event detection, but detection capability is limited and cannot identify new flight events
Solution Approach 1:
The system automatically generates new flight events by computing correlations between flight data sets without requiring manual expert configuration. The processing unit autonomously identifies correlated parameters, defines new flight events, and updates the database, enabling the system to self-improve its detection capability while maintaining low expert workload.
Solution Approach 2:
The system pre-computes and stores correlation information between flight data sets in advance. When analyzing flight data, it efficiently retrieves and applies these pre-computed correlations to automatically generate flight events, rather than requiring experts to manually define all possible events beforehand.
2Measurement precision
If all recorded flight data is analyzed to detect potential issues, then detection precision is improved, but resource allocation and time consumption increase significantly
Solution Approach 1:
The system segments the analysis process into two phases: (1) pre-computation of correlation statistics between flight data sets, and (2) efficient querying and event generation using these pre-computed correlations. This segmentation allows comprehensive analysis of all flight data while reducing real-time processing time through cached correlation information.
Solution Approach 2:
The system performs preliminary computation of correlation metrics between flight data sets and stores them in the database. During actual flight analysis, it leverages these pre-computed correlations to quickly identify potential events, avoiding the need to re-analyze all data from scratch and significantly reducing analysis time.
3Measurement precision
If manual expert analysis is used to define flight events, then detection accuracy is maintained through expert knowledge, but productivity is reduced due to extensive manual effort required
Solution Approach 1:
The processing unit automatically generates new flight events by computing correlations between flight data sets without requiring manual expert intervention for each event definition. The system self-updates the database with newly discovered flight events, dramatically increasing productivity while maintaining detection accuracy through systematic correlation analysis.
Solution Approach 2:
The system implements a feedback loop where newly generated flight events are added to the database and can be further correlated with existing data in subsequent iterations. This feedback mechanism allows the system to continuously improve its detection capability automatically, reducing the need for manual expert analysis while maintaining or improving detection accuracy.
Data Source
Figure 1
Figure 2
AI summary
The present invention relates to a method for analyzing a so-called first flight data set, the values of which were recorded during a flight of an airplane and in which at least one subset of flight data, comprising at least one datum from the flight data of the first set and/or at least one flight datum from a second set, under the condition that at least one datum from among the data thereof exceed the nominal value thereof, is defined by correlating the value of at least one flight datum of the first set with the value of at least one datum of the flight data of the second set. An in-flight event can then be detected from the flight data values of a subset and the flight data values of one of the second sets of flight data, and if an in-flight event is not detected from a subset, a pairing probability between at least one second set and said subset is associated with said subset, the first set is then updated by appending at least one new flight datum of at least one second set and/or by deleting at least one datum of the flight data thereof, depending on the pairing probability values, the thus-updated first flight data set is then rerecorded during a new flight and the method is iterated as long as a person skilled in the art is not able to make a decision on said subset.