Document
COCKPIT TECHNOLOGY FOR THE PREVENTION OF GENERAL AVIATION RUNWAY INCURSIONS Lawrence J. Prinzel III and Denise R. Jones NASA Langley Research Center Hampton, VA General aviation accounted for 74 percent of runway incursions but only 57 percent of the operations during the four-year period from fiscal year (FY) 2001 through FY2004. Elements of the NASA Runway Incursion Prevention System were adapted and tested for general aviation aircraft. Sixteen General Aviation pilots, of varying levels of certification and amount of experience, participated in a piloted simulation study to evaluate the system for prevention of general aviation runway incursions compared to existing moving map displays. Pilots flew numerous complex, high workload approaches under varying weather and visibility conditions. A rare-event runway incursion scenario was presented, unbeknownst to the pilots, which represented a typical runway incursion situation. The results validated the efficacy and safety need for a runway incursion prevention system for general aviation aircraft.
Introduction data and has shown the following correlating factors: The Problem • Weather not a factor (89%) The FAA defines Runway Incursions as, "any • Pilots taxiing w/o clearance (62%) occurrence at an airport involving an aircraft, • Landing/ departing w/o clearance (23%) vehicle, person or object on the ground that • Landing on the wrong runways (10%) creates a collision hazard or results in loss of • Pilot distractions (17%) separation with an aircraft taking off, intending • Pilots disoriented or lost (12%) to take off, landing or intending to land."
• Unfamiliarity w/ ATC procedures (22%) Runway incursions are a serious aviation safety hazard, particularly for general aviation • Unfamiliarity with the airport (19%) operations. According to the Federal Aviation The Solutions Administration (FAA, 2005), during the four year period from fiscal year (FY) 2001 through These statistics provide a sobering view of the FY 2004, there were approximately 257 million need for solutions to the problem of runway aircraft operations and 1,395 runway incursions incursions, in general, and one with a focus reported at United States towered airports – involving the human on the flight deck. The approximately 5.4 runway incursions for every FAA has voiced its commitment to reducing the one million operations. General aviation severity, number, and rate of runway incursions accounted for 74 percent of these incursions but by implementing a combination of technology, only 57 percent of the operations. Seventy-six infrastructure, procedural, and training percent of the most severe incursions (114 of 150 interventions. These solutions include Airport incursions) involved at least one general aviation Movement Area Safety System (AMASS); aircraft. Airport Surface Detection Equipment Model 3 (ASDE-3), ASDE Model X (ASDE-X) radar; The Etiologies multi-lateration systems; in-pavement loops; Runway Status Lights; (RWSL); enhanced Statistics show the causes of these incursions are principally pilots (62%) followed by air traffic controller training; airport surface operations advisory circulars; improved airport surface controllers (35%), meaning that the leading causes of runway incursions both involve human markings; improved education, training and awareness; and revised pilot/controller factors. The main causal factor (56%) for pilot- related occurrences was the pilot’s failure to communications phraseology.
follow an ATC clearance (Khatwa, 2002).
These efforts target improved awareness and Further, the FAA has analyzed runway incursion enhanced surveillance, but none of these initiatives directly involve technology solutions experiment was to evaluate several candidate for the flight deck. Taken together, the proposed RIPS elements, adapted for GA operations, and FAA solutions may still not provide a compare them to current electronic flight bag comprehensive solution without addressing the (EFB) capability for prevention of GA runway flight deck. The NTSB currently lists the 6 incursions.
“most wanted” aviation safety improvements, Method including “stop runway incursions/ground collisions of aircraft” and has specifically General Aviation Pilots recommended that the FAA implement Sixteen GA pilots served as participants in the technology that, “give immediate warnings of experiment with an equal distribution of flying probable collisions/incursions directly to flight experience used to represent the Part 91 crews in the cockpit” (cf. NTSB, 2000; 2006).
population: low-time (< 400 hours) visual flight The NASA Runway Incursion Prevention rules (VFR), high-time (> 400 hours) VFR, low- System has been designed to provide a flight time (< 1000 hours) instrument-rated and high- deck solution to the problem of runway time (> 2000 hours) instrument-rated.
incursions.
Facilities/Equipment NASA Runway Incursion Prevention System The simulation experiment was conducted at Leveraging on NASA research (e.g., McCann et NASA Langley Research Center (LaRC) using al., 1998), RIPS integrates airborne and ground- the Integration Flight Deck (IFD) transport based technologies to provide: (1) enhanced category fixed-base high-fidelity flight simulator surface situation awareness to prevent blunders (Figure 1). The IFD, normally a Boeing 757 and errors and, (2) runway conflict alerts to cockpit, was adapted for the experiment to take prevent runway incidents if blunders or errors do advantage of its excellent visual, tactile, and occur.
audio capabilities. A six-degree of freedom non- As presently envisioned, RIPS provides linear simulation model of the Cessna 206 (C- enhanced situation awareness using the pilot’s 206) was used for this experiment. A collimated head-up display (HUD), Primary Flight Display out-the-window scene was produced by an Evans (PFD), and an Electronic Moving Map (EMM) and Sutherland ESIG 4530 graphics system by displaying airport map information, surface providing approximately 200 degrees horizontal traffic, and graphical guidance during rollout, by 40 degrees vertical field of view at 26 pixels turn-off, and taxi. The system also continuously per degree.
monitors for potential incursions and pilot An electronic flight bag (EFB) display was used blunders, and if detected, provides aural and to present the airport surface map display graphical alerts. These alerts are presented concepts described below (Figure 1). This visually on the displays and aurally throughout display was 10.4” (26.4 cm) diagonal with a the cockpit. Research during both simulation resolution of 1280 x 1024 pixels.
(e.g., Young & Jones, 2001) and flight tests (e.g., Jones, 2001) for commercial and business An electronic research display (RD), was aircraft operations have demonstrated that these installed on the instrument panel directly in front technologies can significantly increase situation of the left seat and control yoke. The RD was awareness and reduce the occurrence of runway composed of two 10.4” (26.4 cm) diagonal liquid incursions.
crystal displays and simulated the “Baseline Round Dials” – that is, the standard set of Research Objectives Cessna-206 aircraft instruments: airspeed, The greatest incidence of runway incursions is attitude, altitude, vertical speed, directional gyro, attributable to general aviation (GA) aircraft turn and bank indicator, tachometer, and operations; therefore, mitigating the occurrences Instrument Landing System indicators (Figure for the GA operator could significantly enhance 2).
safety. Because RIPS has demonstrated tremendous potential for eliminating the causes of runway incursions for commercial and business aircraft operations, the research turns naturally toward system efficacy for GA operations. The objective of the present alerting (e.g., “Warning, Traffic Departing Two Five”) (Figure 4).
When traffic data was provided, it was “broadcast” at a 1 Hz rate. Own-ship position data was updated at a 20 Hz rate. Positional errors, noise, or uncertainties were not introduced into these data.
Figure 1. Integration Flight Deck Experimental Design The experiment was designed as a 4 (display) by 4 (weather) by 6 (task) partially factorial, mixed- subjects design. The between-subject factor was display and each participant pilot flew 19 approaches with the one of the four display concepts. On the last approach, a runway incursion was staged to assess the utility of the display concept for runway incursion prevention.
The evaluation subject was not expecting a runway incursion.
Figure 3. Plan-View Surface Map with Traffic (BMOT) Figure 2. Baseline Instruments Display Concepts Four display concepts were evaluated: (a) Baseline with a Moving Map and Own- ship (BMO), (b) BMO + Traffic Display (BMOT), (c) BMO + audible runway incursion alerting (BAMO), (d) BMOT + audible and graphical runway incursion alerting (BAMOT).
Figure 3 presents an example of the plan-view surface map with traffic displayed (BMOT).
These display concepts were designed to represent a range of Electronic Flight Bags (EFB) and alerting applications, typical of GA aircraft. As shown in Figure 3, “Traffic Display” refers to the graphical representation of surface traffic on the moving map display. “Runway Figure 4. Plan-View Surface Map with incursion alerting” involves the addition of Traffic and Alerting (BAMOT) computer-generated audible and/or graphical Evaluation Tasks the most prevalent type of GA runway incursion (i.e., taxiing onto runways or taxiways without During the 19 experimental trials, pilots clearance) during weather conditions when they performed 6 approach tasks at the Reno/Tahoe most often occur (i.e., day VMC). This scenario International Airport (KRNO): (1) 3-nm straight- would engender at least a “category D” severity in approach to Runway 34R with a full-stop rating (see below) from the FAA dependent upon landing; (2) 3-nm straight-in to 34R with wave- pilot response to the event.
off initiated at 200 ft Above Field Level (AFL); (3) 3-nm straight-in approach to 34L, with Runway Incursion Alerting sidestep to 34R for a full-stop landing; (4) 3-nm The Runway Safety Monitor (RSM) incursion straight-in to Runway 25 with a full-stop detection algorithm (Green, 2006) was used to landing; (5) Circle-to-land on Runway 25; from generate the alerting function for the BAMO and a Runway 34L approach and, (6) 9.56-nm BAMOT display concepts. The RSM monitors straight-in approach with go-around initiated at traffic that enters a three-dimensional virtual 200 ft AFL.
protection zone around the runway that is being Weather used by the own-ship. Incursion detection is based on the operational state of the own-ship Four weather conditions were used to create the and traffic, as well as other criteria (separation experimental scenarios: (1) 3 miles visibility, and closure rate). Identification, position, and 1000 ft. ceiling; (2) 3 miles visibility, 2000 ft.
altitude data is used to track the traffic in the ceiling; (3) 1 mile visibility, 1000 ft ceiling; and protection zone. Traffic data projections are (4) 1 mile visibility, 400 ft. ceiling.
calculated within RSM since, from flight test Procedure experience, reliable position updates are not received at consistent intervals. RSM generates Each pilot participated in an extensive briefing a warning alert, which occurs when a runway and training session that was designed to mask incursion is detected and evasive action is the runway incursion focus of the experiment.
required to avoid a potential collision.
Pilots flew approach tasks designed specifically Information provided with each alert includes to set-up the necessary conditions for identification of the incurring traffic and presentation of the runway incursion scenario.
separation distance to potential conflict. RSM The runway incursion scenario was presented on was developed for NASA by Lockheed Martin.
the last experimental run. Pilots were not Results informed of the total number of runs. Post-run scales (i.e., Situation Awareness Rating FAA Runway Incursion Severity Ratings Technique, NASA-Task Load Index) and questionnaires were administered. The FAA performed an independent analysis of the experimental data, using an FAA runway Simulated ATC clearances were given to all incursion severity rating (FAA, 2005), and traffic to recreate the ATC communication categorized the runway incursion incident data “party line” environment for participant pilots.
from this study.
Surface and airborne traffic were simulated to represent typical operations at KRNO (Reno). • Category A – Separation decreases, extreme action taken to narrowly avoid Runway Incursion Scenario collision, or collision occurs; The scenario began with the C-206 on approach • Category B – Separation decreases, aligned with Runway 34R, 3 nm from the significant potential for collision; threshold at 1010 ft AFL and 90 knots. The weather condition was day with 1000 ft ceiling • Category C – Separation decreases, and 3 miles visibility. The incursion traffic ample time and distance to avoid started at the Runway 34R hold line near the collision; Runway 34R threshold. The incursion traffic • Category D – Little or no chance of then taxied into position on the active runway collision but meets definition of runway while the participant pilot was on final approach incursion.
(approximately 2 nm from the threshold). The incursion is categorized as a pilot deviation (i.e., Using these classifications, the 16 “rare event” incursion traffic not cleared for departure on runway incursions produced fourteen scenarios 34R). The runway incursion scenario represents in the less hazardous Category C and D map and/or incursion alerts, but the differences incursions, one resulted in a Category A were not statistically significant at the α = 0.05 incursion, and one resulted in a Category B level.
incursion (see Table 1). The 14 less hazardous No statistically significant differences were Category C and D incursions were mitigated by found between the display concepts for the the EPs by conducting a go-around and gaining distance to the incurring traffic when the pilots separation from the traffic. Traffic awareness initiated a go-around or for the EP’s reaction was provided by either the display concepts or time from the incursion event occurrence (based visual acquisition out-the-window.
on a Multivariate Analysis of Variance The Category A runway incursion occurred with (MANOVA) test p > 0.05).
the EP flying the BMOT display concept.
Despite the traffic indications on the surface map and out-the-window visuals, the EP 6000 35 demonstrated no awareness of the runway traffic, over-flew the traffic and landed.
The Category B incident occurred when the EP over-flew the runway traffic (at 146 ft AFL) before conducting a go-around. The EP was aware of the incursion after having received an Feet audible alert (BAMO display concept) but Seconds continued to descend to visually acquire the Distance to Traffic traffic to confirm the alert. This incident would Time to Traffic have been classified as a Category D incursion if the EP had initiated the go-around at first 0 0 awareness of the alert.
BMO BMOT BAMO BAMOT Table 1. Results Categorized by the FAA Runway Incursion Severity Ratings Figure 5. Initial Traffic Awareness Pilot Display Rating #1 BMO D Mental Workload and Situation Awareness #2 BMOT C- After the incursion scenario, no statistical #3 BAMO B+ differences were found for overall mental #4 BAMOT D workload or situation awareness, p >.05, between #5 BMO D the display configurations. Subjectively, the EPs gave significantly better ratings for audible #6 BMOT D alerting displays for runway incursion detection, #7 BAMO D F (3,15) =17.955, p<.05; likelihood of runway #8 BAMOT C incursion prevention in real-world, F (3,15) = #9 BMO C 10.948, p<.05; and level of perceived safety, F(3,15) = 8.814, p<.05.
#10 BMOT A #11 BAMO D Pilot Preference #12 BAMOT C For those displays that had alerting (BAMO, #13 BMO D+ BAMOT), there were no significant differences in timeliness of the alerting in terms of being #14 BMOT D+ able to take evasive action. However, when #15 BAMO D pilots were asked to rate all four display concepts #16 BAMOT D on the perceived efficacy of the alerts (F (3,15) = 10.948, p<.05) and the additional safety value Runway Incursion Detection Reaction Time added (F(3,15) = 8.814, p<.05) analyses revealed As shown in Figure 5, the incursion traffic was significant effects between the displays.
typically acquired sooner when the EP was Subsequent post-hoc Student Newman Keuls provided with a traffic display on the surface tests showed that pilots reported that the BMO display condition was significantly poorer than refine the system concepts with targeted the other three display conditions, which were enhancements toward support of GA operations.
not significantly different from each other.
References Conclusions Federal Aviation Administration Office of Runway Safety, August 2005, FAA Runway The experimental objective was to determine Safety Report.
how different EFB and alerting concepts supported pilot situation awareness and Green, D.F. (2006). Runway safety monitor resolution of runway incursions. The results algorithm for single and crossing runway show large individual differences in response to incursion detection and alerting. NASA CR- an incursion event regardless of display concept.
2006-214275 However, only one pilot’s performance was judged as a severe runway incursion risk for Jones, D. (2001). “Runway Incursion Prevention collision (with the BMOT display). Despite System – Demonstration and Testing at the having traffic on the display and other cues, the Dallas/Fort Worth International Airport.” In pilot was unaware of the traffic and landed just Proceedings of the 20th Digital Avionics Systems beyond the incursion aircraft, resulting in a near- Conference. Daytona Beach, Florida miss.
Khatwa, R. (2002). An analysis of runway The addition of audible alerting was found to incursions, 1990-2002. Meeting of the Flight enhance runway incursion detection.
th Safety Foundation (FSF) 55 annual Furthermore, had the experiment simulated the International Air Safety Seminar (November 4-7, alerting system also being installed on the 2002). Dublin, Ireland.
incursion aircraft, it is less likely that situation would have become a runway incursion event McCann, R.S., Foyle, D.C., Hooey, B.L., Andre, (i.e., category “D”); since the incursion aircraft A.D., Parke, B., & Kanki, B. (1998). An would have received an alert before taxing onto evaluation of the taxiway navigation and the active runway.
situation awareness (T-NASA) system in high- fidelity simulation . SAE Transactions: Journal of The results generally match past research on Aerospace , 107 , 1612-1625.
commercial and business aircraft operations - the incursion alerts provided sufficient time and National Transportation Safety Board, Safety awareness to avoid a potential incursion conflict.
Recommendation Letter to the FAA Adminis- Post-run briefings revealed that a surface map trator , A-00-66, July 6, 2000.
with own-ship and traffic along with audible alerts was considered an optimal incursion National Transportation Safety Board. (2006).
prevention display for GA aircraft, while an Most Wanted Transportation Safety Improve- audible alert alone was considered a minimally ments - Aviation Issue Areas: Stop Runway effective display. Over half of the pilots Incursions/Ground Collisions of Aircraft.
evaluated would have liked maneuver guidance www.ntsb.gov/Recs/mostwanted/aviation_issues for conflict resolution in conjunction with .htm incursion alerting. In general, the pilots reported feeling substantially safer during runway Young, S.D., & Jones, D.R. (2001). Runway incursion incidents with onboard alerting.
incursion prevention: A technology solution.
Th Proceedings of the Flight Safety Foundation 54 There is tremendous potential to significantly Annual International Air Safety Seminar, 54 , 1- enhance safety for all classes of aircraft by using 22. Athens, Greece: Flight Safety Foundation.
flight deck awareness and alerting for runway incursion prevention, such as that demonstrated herein using the NASA Runway Incursion Prevention System. Future research will further