Data-Driven Trajectory Optimization on the Cgk–upg Route Using K-means and Bluesky Stochastic Simulation

Abstract

This study examines the operational efficiency of the Jakarta–Makassar (CGK UPG) route, one of Indonesia’s busiest domestic corridors, by integrating historical ADS B data analysis with stochastic simulation. Flight trajectories obtained from FlightAware were processed through data cleaning, standardization, Principal Component Analysis (PCA), and K-Means Clustering to extract a representative data-driven trajectory. This trajectory was evaluated alongside the RNAV T-5 and conventional VOR/NDB routes documented in AIRAC AIP AMDT 162. All routes were assessed using 300 Monte-Carlo simulations in the BlueSky ATM Simulator, incorporating randomized variations in cruise speed and altitude to reflect operational uncertainty. Performance indicators included flight distance, travel time, fuel consumption, and CO₂ emissions. Results show that the data-driven trajectory offers shorter distance and flight duration, requiring less fuel and producing lower emissions compared to the two procedural routes. These findings demonstrate that trajectory extraction based on historical data can produce more efficient flight paths and support the implementation of Trajectory-Based Operations (TBO) within Indonesia’s domestic airspace. The approach also contributes to national sustainability efforts by enabling potential reductions in carbon emissions from the aviation sector.