Airspace Operations Lab @ NASA Ames

In 2013, Chevalley, et al., presented a concept of shared airspace where departures fly across arrival flows, provided gaps are available in these flows. They explored solutions for separating departures temporally from arrival traffic. Arrival controllers were responsible for deciding whether to climb departures through gaps, based on the departure aircrafts' trajectory and on the estimated flying time across the arrival flow. It was found that aircraft climb efficiency increased with more accurate departure time from the runway. Although in this earlier simulation, workload, coordination, and safety were judged by controllers as acceptable, it appeared that controllers would need improved tools to support this procedure.

In the current follow-up study, decision support tools were developed to help controllers decide whether it was safe to climb aircraft through gaps in the arrival flow. In all three tool conditions, controllers could refer to a timeline to show how close in time the departures were predicted to be to the arrivals. In two of the conditions, controller could either see tie-points on videomaps, or could use a conflict probe to assess the separation of arrivals and departures dynamically.

The tools were tested in a Human-In-The-Loop simulation. The efficiency and safety of 144 departures from the San Jose airport (SJC) climbing across the arrival airspace of the Oakland and San Francisco arrival flows were assessed. The simulation replicated the airspace and the manipulation of the accuracy of departure release times as reported in Chevalley et al. (2013) but used different aircraft climb profiles.

Results show that again aircraft climb efficiency improved with departure time accuracy. Additional tools, such as the tie-points and the conflict probe, helped controllers make decisions to climb aircraft. In most cases, the tools helped controllers to keep aircraft vertically separated. For example, the tools helped controllers keep aircraft at safe altitudes longer when aircraft departed outside of their scheduled time. However, the tools did not prevent losses of separation. Seven losses of separation took place. Four of those were just below the required separation standards.

This paper presents problems involved in predicting separation, and with controllers using anticipated separation to make decisions. New procedures and more precise tools are needed to limit the use of anticipated separation and to give options to controllers to climb aircraft safely.