HexAero is a EUROCONTROL Python and R package designed for aviation professionals and data analysts. It allows for the determination of used airports, runways, taxiways, and stands based on available (ADS-B) flight trajectory coordinates. This tool aims to enhance aviation data analysis, facilitating the extraction of geospatial milestones for operational performance analysis.

The package is built on top of the H3 system developed by Uber Technologies, Inc. and open geospatial data retrieved from OpenStreetMap (OSM) and OurAirports.com. H3 is a hexagonal hierarchical geospatial indexing system that enables coordinates to be associated with hexagons (identified by an H3 tag). The hexagons used in this indexing system cover the Earth's surface in various resolutions (i.e., hexagon sizes). By labelling and thus associating certain hexagons (i.e., H3 tags) to various spatial features (e.g., airports, stands, taxiways, runways, or other regions extracted from OSM and OA), one can detect the flight trajectory entry and exit of these regions. In practice, this is done by transferring the (ADS-B) trajectory coordinates (latitude, longitude) to the associated hexagons covering the trajectory (i.e., by determining the H3 tags of the flight coordinates) and associating these to the predefined hexagons covering the geospatial feature of interest.

As an example, by creating a hexagon database for each arrival area (i.e., an area consisting of the runway and cone-shaped areas at each runway's end extending several nautical miles) of each runway at all airport, one can detect the entry and exit of flight trajectories within these areas and runways. This is done by merging the predefined arrival area databases onto the flight trajectory coordinate-associated H3 tags. In the case of overlapping arrival areas for multiple runways, heuristics are used to determine the most likely used runway. These heuristics take into account the distance to the runway of crossed sub-areas within the arrival area, the number of crossed sub-areas, whether the runway is touched, etc. Based on the extracted arrival area entry and exit information and the applied heuristics, a final score is determined which indicates the most likely used arrival or departure runway for each trajectory.

• Airport Detection: Identifies airports involved in a flight's trajectory.
• Runway Utilization: Determines which runways are used during takeoff and landing.

To install HexAeroPy, ensure you have Python 3.9+ installed. Then run the following command:

pip install HexAeroPy

Get started with HexAeroPy by running the following Python code:

from HexAeroPy import *

This will prompt you to download the geospatial metadata from Zenodo necessary to run the package. It will only cache once. The parquet datasets are available here: https://zenodo.org/records/10651018.

# Load test data

df = load_dataset(name='trajectories.parquet', datatype='test_data')

# Create unique id for each trajectory

df['id'] = df['icao24'] + '-' + df['callsign'] + '-' + df['time'].dt.date.apply(str)
df = df[['id', 'time', 'icao24', 'callsign', 'lat', 'lon', 'baroaltitude']]

# Identify runways

scored_rwy_detections_df, rwy_detections_df = identify_runways(df)

# Load approach hex dataset
egll = load_dataset(name="EGLL.parquet", datatype="runway_hex")

# Visualize approach cones
map_viz = choropleth_map(
        egll,
        column_name='gate_id_nr',
        border_color='black',
        fill_opacity=0.7,
        color_map_name='Reds',
        initial_map=None,
        initial_location=[df.lat.values[0], df.lon.values[0]],
        initial_zoom = 13,
        tooltip_columns = ['id', 'airport_ref', 'airport_ident', 'le_ident', 'he_ident', 'length_ft', 'width_ft',
   'surface', 'lighted', 'closed', 'gate_id']
    )

# Add a single aircraft trajectory to the map
df = df[df['id'] == '0a0048-DAH2054-2023-08-02']
add_trajectory(map_viz, df)

# Show
map_viz.save('egll_arrival.html')

Download the HTML here: egll_departure.html

• [pending implementation] Taxiway Analysis: Analyzes taxi routes for efficiency and optimization.
• [pending implementation] Stand Identification: Identifies aircraft stands, enhancing ground operation analysis.

We welcome contributions to HexAeroPy! Feel free to submit pull requests, report issues, or request features.

HexAeroPy is licensed under the MIT License - see the LICENSE file for details.

Special thanks to EUROCONTROL and the Performance Review Commission (PRC).

For support or queries, please contact us at [email protected].