Separation Assurance (SA) Highlights - Fiscal Year 2013
In January 2013, Aviation Systems and Human Systems Integration Division researchers completed what is believed to be the first human-in-the-loop simulation to address the age-old question regarding Trajectory Prediction (TP) performance: “How good is good enough?” That is, how much trajectory-prediction error can be accommodated by air traffic controllers before a NextGen Trajectory-Based Operations (TBO) concept becomes untenable? Two controller teams used an advanced set of tools over 22 typical Atlanta arrival-metering trials in which uncertainties in the form of wind forecast errors and errors in aircraft performance (e.g., descent rate) assumptions were varied from nominal levels to four- to five-times these real-world uncertainties. Real-time controller performance metrics on the number of clearance instructions, workload, and quality of service in terms of arrival metering accuracy and traffic separation were recorded. Preliminary observations suggest that there was no clear controller "breakpoint" (whereupon the operational concept became untenable). Even with the largest errors, controllers were able to compensate and adapt the advisories suggested by the automated tools to maintain safe separation and deliver arrival aircraft on-time. Reported workload ratings did increase with increasing errors, but on average remained within acceptable levels. Increased TP errors did not appear to be correlated with an increase in the number of corrective clearances, nor a reduction in the efficiency benefits of the NextGen tools. These findings suggest future research should further investigate the exact nature and impact of how the controllers adapted to the TP errors, and, if the same TP errors that were so effectively managed by the controllers are found to overwhelm a fully automated system that relies solely on corrective advisories, there may be a need to incorporate corrective learning in TBO automation as the controllers did naturally in the simulation. There may also be a need for additional development of the human-system concept for transition to more automated paradigms.
Aviation Systems Division engineers met with flight operations personnel from Regional-Jet (RJ) operator SkyWest Airlines to discuss NASA findings that suggest that SkyWest could save upwards of 50 pounds of fuel per RJ flight by altering their arrival procedures to use more optimal flight path angles that would avoid the need to deploy speedbrakes. NASA developed an algorithm that can inform operators such as SkyWest what flight path angle would be optimal given the prevailing winds, route of flight and other factors such as the descent-speed profile being advised by the Efficient Descent Advisor (EDA), where in use. These savings are over and above the fuel and environmental benefits provided by EDA, and the new algorithm/ tool would also reap benefits for un-delayed flights (e.g., into low-density "out stations"), essentially providing EDA-like benefits without using EDA. The SkyWest officials quickly validated that even a 20-pound savings would make an operationally and economically significant difference, since their policy is to plan fuel loads down to tens of pounds. They also confirmed the viability of the proposed operational concept for the tool. An invention disclosure is being written.
SA research in evaluating and improving trajectory prediction accuracy, development of the Advanced Airspace Concept (AAC), and the impact of speed, wind and weather uncertainties, conflict resolution, and multiple separation assurance agents, were presented and published in multiple venues, including the AIAA Guidance, Navigation and Control (GNC), Aviation 2013 conferences, DASC 2012, the Electronic Navigation Research Institute (ENRI) Workshop on ATM/CNS, and ATM 2013. Technical lead and principal investigator Dr. Todd Lauderdale also gave multiple seminars on Robust Automation to Reduce Air Transportation Delays for a number of university audiences (Cornell University, University of Toronto, University of Texas) as well as at the NASA Ames Summer Lecture series.