Flight trajectory grafting: Leveraging historical trajectories for more efficient arrival air traffic management

Inside the terminal maneuvering area (TMA), flight trajectories need to be determined to maintain safe and efficient arrival operations. Air traffic control officers (ATCOs) devise trajectories and provide instructions to pilots. The subjectivity involved in the decision-making exposes operational efficiency to factors such as workload, experience, and TMA complexity. Suboptimal trajectory solutions can increase arrival transit times, i.e., the time spent from entering TMA to landing, leading to congestion and flight delays. These adverse effects are particularly critical during peak hours. While existing methods provide efficient trajectory solutions, they often overlook critical embedded features that constitute trajectory solution feasibility in real operations. To address these challenges, we propose a trajectory grafting method to generate high-fidelity, feature-embedded trajectories compatible with existing air traffic management systems. Trajectory grafting utilizes historical trajectory segments as components to construct situational flight trajectories that conform to given traffic dynamics and constraints. Collectively, these trajectory segments constitute a feasible design space, thereby eliminating the need to explicitly model operational constraints, flight physics, and ATCOs’ workload. Our results demonstrate the benefits of this method, which reduces the average arrival transit time by 3% during peak hours. The benefits are further amplified by its compound effect, with up to 24% reductions in accumulated arrival transit times. • Creates grafted trajectories with operational features, enhancing applicability. • Reduces peak arrival transit times by 3.05% and separation deviations by 90%. • Reduces sequential arrival transit times by up to 24.1%. • Highlights the compound effect and interdependent nature of arrival transit times. • Shows most inefficiencies can be rectified with minimal late-stage adjustments.