Aircraft Prognostics Using Mutual Information Graphs
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current Airplane Health Management (AHM) solutions are limited in their ability to gather and exploit data for real-time, in-flight prognosis, as they primarily rely on reactive diagnosis and only utilize a small fraction of the available data, resulting in inefficient maintenance processes.
Innovation Solution
A method and system for inferred information propagation that preprocesses time-series data by dividing it into subsets, computing Mutual Information (MI) values, constructing relationship graphs, clustering, and analyzing these graphs to identify features in aircraft components, enhancing the prediction of impending failures.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Device complexity
If only a small fraction of available data is used for prognosis, then data processing complexity is reduced, but prediction accuracy deteriorates
Solution Approach 1:
The patent segments the available data into different types (OOOI events, maintenance events, performance parameters) and processes only the most relevant segments for prognosis. This selective segmentation reduces data processing complexity while maintaining prediction accuracy by focusing computational resources on high-value data subsets.
Solution Approach 2:
The patent extracts and utilizes additional data types beyond traditional OOOI events, including maintenance events and performance parameters. This extraction of useful information from previously underutilized data sources improves prediction accuracy without proportionally increasing processing complexity, as the system selectively processes only the most prognostically valuable extracted data.
2Device complexity
If reactive diagnosis is used, then system simplicity is maintained, but maintenance efficiency deteriorates
Solution Approach 1:
The patent implements preliminary action by performing prognostic analysis during flight using available data to predict potential failures before they occur. This allows maintenance to be scheduled proactively rather than reactively, improving maintenance efficiency by planning interventions during convenient maintenance windows rather than during unscheduled groundings.
Solution Approach 2:
The patent introduces dynamic prognostic capabilities that adapt to flight-specific data patterns and conditions. The system dynamically adjusts its analysis based on the specific flight context, component behavior, and emerging trends, maintaining system simplicity while achieving superior maintenance efficiency through context-aware predictive analytics.
3Loss of information
If comprehensive data collection is implemented, then information completeness is improved, but data exploitation capability deteriorates
Solution Approach 1:
The patent extracts and prioritizes the most prognostically valuable features from the comprehensive flight data set. By identifying and focusing on key performance parameters and event patterns that most strongly indicate potential failures, the system maintains information completeness while improving data exploitation capability through selective feature extraction and analysis.
Solution Approach 2:
The patent transforms raw flight data into prognostically relevant parameters and metrics through processing and analysis. By changing the parameter representation from raw sensor readings to derived performance indicators and trend metrics, the system improves both information completeness and exploitability, making the data more actionable for failure prediction.
Data Source
AI summary
Methods and systems are provided for inferred information propagation for aircraft prognostics. The method includes receiving, by a processor, an original time-series of data points for a component as an input; preprocessing the input to divide the original time-series of data into subsets of data by applying a time-window over the original time-series of data points; and computing, by the processor, a Mutual Information (MI) value for each pair of variables within each subset of data. The method also includes constructing, by the processor, a sequence of relationship graphs using the computed MI values; clustering, by the processor, each relationship graph; and analyzing, by the processor, the time-ordered sequence of clustered relationship graphs to identify features in the component.