jonaskaad/bachelorproject

GitHub repository for our bachelor project for 6th semester at SDU

Recognition of Patterns in Flight Tracking Data

Bachelorproject @ SDU, F23. Software Engineering

Students:

• Jonas Solhaug Kaad ([email protected])
• Victor Andreas Boye ([email protected])
• Alexander Vinding Nørup ([email protected])

Supervisor: Kamrul Islam Shahin - [email protected]

In collaboration with ForeFlight.

Development / Running locally

Prerequisite: dotnet 7.x: https://dotnet.microsoft.com/en-us/download

1. Clone the repository
2. Copy: FlightPatternDetection/appsettings.json to FlightPatternDetection/appsettings.Development.json and change: mysqlConnectionString to point to a MariaDB server with FlightPatternDetection/Database_Creation.sql loaded in.
   • Optional: Change Logging.LogLevel.Default to Debug in FlightPatternDetection/appsettings.Development.json
3. Find the latest NAVDB and place it here FlightPatternDetection/NAVDB.sqlite.
4. Optional: Load in some fallback flights into FlightPatternDetection/FallbackHistoryData/. Get some from a good friend or create your own by using avalaible python scripts in scripts/ (requires FF-vpn). (Remember that flights should be stored as .JSON)
5. Debug and run. Use the FlightPatternDetection project as your startup project. Use the http or https launch-profile.

bachelorproject's Issues

Sequence Diagram

Create a Sequence Diagram for the project to describe how it all works together

Fetch flight data from endpoint

The program should be able to fetch external flight data from some kind of endpoint.
Ideally it should be the one by ForeFlight. This will need the VPN, so for now 2-3 test files handled from somewhere else as proof of concept will be fine.

Should be able to:

• Fetch the data and download it
• Load the data and save it if changes are needed
• Save relevant parts to the database

Flight Visualizer

Make a view that can load a KML or JSON file and show it on a map-view.

Use leaflet.js?
You might be able to use this, but it doesn't seem that developed: https://github.com/Mehigh17/BlazorLeaflet

Research basis for using image recognition

Find out how hard image recognition is, and how viable it might be.

Research Holding Patterns

Find clear definitions for holding patterns, how to fly and thus how to recognize them.

Load NavDB

Load NavDB locally to the project

Color lines on the map based on different parameters

The lines could be colored based on:

• Flight is in the airport (On ground)
• Takeoff (Climbing)
• Cruise altitude (When flight is cruising at designated altitude)
• When a holding pattern occurs

Make it easier to mass validate and visualize flights.

Use Case Diagram

Create a use case diagram to show how the user interacts with the system

Implement the database logic

Use Entity Framework Core and MySQL

Research basis for using "AI" for pattern-matching in datasets

Find or create training data for Machine learning

In order to use machine learning, we need some data to train the machine.
Hopefully, we can find some data but if not, then we need to find flight data that ends up in holding patterns and some that don't.

Store flights in database

Store the flights in the database tables:

• Flight
• Airport
• Route Information
• Holding Pattern

Setup the project on GitHub

Setup the project environment

Implement the engine project

The engine is the main powerhouse of the project.

Should do the actual calculations

Zoom and re-draw holding patterns

When a holding pattern has been found it should be possible to click on a button to zoom in to the area with the holding pattern. It should then re-draw the animation for the pattern.

Fix map boundaries

Fix the drawing on the map when it flies over the antimeridian. (See problem below)

• Could do something like -180 on the value

Setup Overleaf for project report

There currently is a document created, however we need to make sure it's ready. Maybe copy stuff from the contract?

Set up the requirements

Research, find and define the requirements for the project

Research Holding Patterns

Find clear definitions for holding patterns, how to fly and thus how to recognize them.

Find waypoints for flight

Find the waypoints for the corresponding flight from the AirNavDB.

• Could be with triangulation for the whole entire flight
• Could be within a radius of a single LAT / LONG coordinate

Improve simple pattern detection engine

We could improve the simple pattern detection engine in a couple ways.

• Ignore points with altitude 0
• Check that the inversions are close enough together to be part of a holding pattern
• Check if the inversion points are close to the general flight level and speed of general holding patterns

2. Presentation Pillar

Be ready for the 2nd presentation pillar. We should bring the following:

3 more diagrams:

• Sequence Diagram
• Use Case Diagram
• Database Diagram

Study the UML language and lingo and be ready to answer questions about the diagram. Kamrul will do some interview questions during the presentation.

Convert requirements and classify them into the modules. (Requirements that are related. Should be grouped together) 4 - 5 modules. At least one should be fully fleshed-out and be ready for the presentation.

• API
• DB
• Engine

"Reformulate the diagram correction". Update the Activity Diagram the reflect changes Kamrul requested.

PowerPoint titled "___ 2.1" that showcases all this work.

1. Presentation Pillar

Updates to map

• Add some kind of direction, to show which way the flight is going
• Add option to clear map. (Either on next flight load or button)
• Possibility to click on next flight (list of them)
  -- click on 10-15 flights to quickly load in

Database Diagram

Create an ER diagram to showcase the database