This tutorial aims to provide participants with a comprehensive understanding of aircraft trajectory analysis using deterministic rule-based methods and machine learning techniques. Over the course of three hours, participants will learn how to access trajectory data, implement analysis techniques in Python, and design machine learning algorithms for more advanced studies. By the end of the tutorial, participants will acquire the necessary knowledge to analyse and interpret aircraft trajectories in diverse real-world scenarios.
Introduction. An overview of use cases for aircraft trajectory analysis, including commercial aviation, general aviation, and other low-altitude activities. The introduction will cover a wide array of possible applications, including situational awareness, airspace management, safety assessment, optimization, and collaborative decision-making.
Accessing trajectory data and meta-information. A practical guide on accessing open-access trajectory data. Participants will gain an understanding of various data formats and standards commonly used in the aviation industry, including ADS-B, Mode S, ADS-C for trajectories, and other open sources of data for complementing context information. We will demonstrate how to effectively access and parse trajectory data using Python.
Deterministic approaches to trajectory analysis. We will explain data cleaning, trajectory filtering and smoothing, essential steps for preparing trajectory data for analysis. Then, we will explore methods for integrating trajectory data with external sources such as weather data and flight plans, enhancing the contextual understanding of aircraft movements.
Machine Learning approaches. Unsupervised machine learning techniques for trajectory clustering, classification, and anomaly detection. We will learn how to uncover patterns and insights within trajectory data. Additionally, we will address the benefits of machine learning approaches over rule-based methods in different contexts. Graph Neural Networks and generative models will also be covered.
Designing a Toolchain for Trajectory Analysis. Best practices for conducting transparent and reproducible research, emphasizing the importance of open data, transparent methodologies, and reproducible analyses.
Conclusion. Suggestions for further exploration and research in aircraft trajectory analysis, emerging trends, potential areas for innovation, and future research.
