Air Route Traffic Prediction for Flight Plan Adjustment
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
The National Airspace System (NAS) faces inefficiencies in managing increasing aircraft traffic, leading to frequent flight delays due to the manual tracking of a large number of aircraft and adverse weather conditions, resulting in increased costs for airlines and passengers.
Innovation Solution
An air route traffic prediction system that uses machine learning models to accurately predict aircraft traffic counts in airspace sectors by parsing and correlating incoming data, identifying salient features, and selecting the most accurate models to adjust flight plans and reduce congestion.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Productivity
If manual tracking methods are used to manage aircraft traffic, then operational simplicity is maintained, but productivity decreases due to inability to efficiently manage increased aircraft traffic
Solution Approach 1:
The patent replaces manual mechanical tracking methods with an automated machine learning-based prediction system. The system uses ML models to predict aircraft traffic counts in airspace sectors, automatically processing data from multiple sources including radar, satellite, and weather systems to generate traffic predictions without manual intervention.
Solution Approach 2:
The system enables self-service by allowing the machine learning models to automatically learn from historical traffic data and improve predictions over time without human reconfiguration. The models autonomously adapt to changing traffic patterns, seasonal variations, and emerging routes, continuously optimizing air traffic management.
2Quantity of substance
If more aircraft are tracked manually, then coverage increases, but loss of time increases due to delays in processing and responding to traffic data
Solution Approach 1:
The system performs preliminary actions by predicting future aircraft traffic counts in various airspace sectors before congestion occurs. The machine learning models analyze current and historical data to forecast traffic patterns, enabling air traffic controllers to proactively reroute aircraft or adjust schedules before delays happen, rather than reacting after problems arise.
Solution Approach 2:
The system implements continuous feedback loops where prediction results are fed back to controllers who can adjust traffic flow, and these adjustments are in turn fed back into the system to improve future predictions. The ML models continuously learn from the effectiveness of traffic management decisions, refining their accuracy over time.
3Reliability
If traditional air traffic control methods are used, then system simplicity is maintained, but reliability decreases due to inability to handle adverse weather conditions and high traffic volumes
Solution Approach 1:
The patent implements a universal prediction system that handles multiple functions: predicting traffic counts under various weather conditions, managing different types of airspace sectors, processing diverse data sources (radar, satellite, weather systems), and supporting various traffic management strategies. The machine learning models are designed to be adaptable across different operational scenarios and geographic regions.
4Loss of time
If manual air traffic control is used, then operational cost is reduced, but loss of time increases due to frequent delays
Solution Approach 1:
The system changes key operational parameters by transitioning from reactive manual control to proactive automated prediction. The machine learning models process and analyze multiple data parameters simultaneously (aircraft position, speed, weather conditions, historical traffic patterns) to generate comprehensive traffic predictions, enabling more informed and efficient traffic management decisions.
Data Source
AI summary
Methods, apparatus, systems and articles of manufacture are disclosed to improve aircraft traffic control. An example apparatus includes a network interface to obtain air route traffic (ART) data associated with aircraft flying in airspace sectors during a first time period, a database controller to generate a first database entry by mapping extracted portions of the ART data to a first database entry field of the first database entry, and ART sector services to execute machine learning (ML) models using database entries to generate first aircraft traffic counts of the airspace sectors during the first time period, in response to selecting a first ML model of the ML models based on the first aircraft traffic counts, execute the first ML model to generate second aircraft traffic counts during a second time period, and transmit the second aircraft traffic counts to a computing device to cause an aircraft flight plan adjustment.