Aircraft-Specific Vehicle Routing Profile for Taxi Route Finding

This paper introduces an open-source method for creating aircraft-specific vehicle routing profiles to obtain distance-optimized taxi routes on airport surfaces. The airport is modelled as a directed graph with nodes for segment intersections and edges for taxiways, runways, and parking positions, using open-source data from OpenStreetMap (OSM). Missing information is appended using the Aeronautical Information Publication (AIP). Key attributes such as coordinates and geometries are queried through osmnx, with novel attributes derived from segment headings. Aircraft-specific parameters, including steering angle, takeoff/landing distances, and wingspan, are sourced from EUROCONTROL's openly available Aircraft Performance Database (APD). Combined with the aircraft-specific routing profile, Dijkstra's algorithm computes the shortest path with feasible runway entries/exists. Taxi times are derived using acceleration/deceleration rates and aircraft-specific performance data. We validate the model by comparing distance-optimized routes for Zurich Airport (LSZH) in 2023 against actual routes derived from Automatic Dependent Surveillance-Broadcast (ADS-B) data, using 743 routes validated with data from EUROCONTROL's Airport Operator Data Flow (APDF). Results show actual routes are 7.5% and 19.8% longer for arrivals and departures, respectively.