Machine learning based en-route flight safety methods

Total time needed: ~5 hours
Learning Objectives
Jump start application of machine learning methods for trajectory prediction and use cases involving flight safety
Potential Use Cases
Flight safety, flight prediction, delay prediction, flight operations, predictive maintenance, aircraft monitoring
Target Audience
INTERMEDIATEData Scientists interested in aerospace verticals, flight analysts, operations research analyst in the aerospace verticals
Go through the following annotated items in order:
ARTICLE 1. IATA article on air passenger demand forecasts
  • Why are developments and advanced analysis for flight safety necessary ?
10 minutes
ARTICLE 2. The future of the national airspace system
  • What is the current state of the art and the future of air transportation systems that are (or will be) implemented ?
10 minutes
ARTICLE 3. Base of aircraft data: conventional approach to flight prediction
  • What is the conventional physics based approach to trajectory prediction ?
10 minutes
PAPER 4. Flight trajectory prediction using deep generative convolutional recurrent neural networks approach
  • How can we leverage deep learning in trajectory prediction ?
60 minutes
REPO 5. Deep trajectory prediction
  • How to implement deep learning methodologies for trajectory prediction ?
120 minutes
PAPER 6. Introduction of uncertainty using Bayesian methods in flight trajectory prediction
  • How can we include uncertainty in a manageable manner in flight trajectory predictions ?
60 minutes
VIDEO 7. Bayesian uncertainty integration with trajectory prediction explained
  • How can we implement Bayesian uncertainty analysis and LSTMs for trajectory prediction ?
20 minutes
OTHER 8. Historic flight dataset
  • Where can I access historic data to develop and test my methodologies ?
10 minutes

Concepts Covered