report flight deck simulations of station keeping · document id: sas_nup_wp2_sk_report_1.0...

Programme/Area: TEN-T/DGVII Project Number: EU/S/98/239 Project Title: North Atlantic ADS-B Network Update Programme, Phase I (NUP I) Document Id: SAS_NUP_WP2_SK_report_1.0 Internal Reference: NA Version: 1.0 Work package: 2 Date: 2001-01-17 Status: Final Classification: Public Author(s): Capt. Michael Agelii, SAS / Christian Olausson, SAS

Report

Flight Deck Simulations of Station Keeping

Flight Deck Simulations of Station Keeping Page 1

SAS_NUP_WP2_SK_report_1.0 Capt. Michael Agelii, SAS / Christian Olausson, SAS 2001-01-17

Document Identification Programme: TEN-T/DGVII Project Number EU/S/98/239 Project Title: North European ADS-B Network Update Programme, Phase I Project Acronym NUP I

Chairman of Steering Committee Mr. Bo Redeborn, SCAA

+46 1119 2388 [email protected]

Project Technical Manager Mr. Niclas Gustavsson, SCAA +46 1119 2273 [email protected]

Partners Luftfartsverket, SCAA Statens Luftfartsvæsen, DCAA Deutsche Lufthansa, DLH Scandinavian Airlines System, SAS Deutsche Flugsicherung GmbH, DFS Airbus Industrie, AI Direction Générale de l'Aviation Civile, DGAC

Document title Flight Deck Simulations of Station Keeping Document Id SAS_NUP_WP2_SK_report_1.0 Organisation Internal Reference NA Work Package No 2 Version 1.0 Status Final Classification Public Date 2001-01-17 Author(s) Capt. Michael Agelii, SAS / Christian Olausson, SAS Organisation maintaining document NUP File Simreport1.0.doc Printed 2001-01-22

Copyright: Scandinavian Airlines System SAS, Flight Operations, Standards & Development

This document may be freely downloaded and distributed from the NUP website.

Nothing in this document may be altered without prior permission from the authors.

www.nup.nu



Table of Contents TABLE OF CONTENTS .............................................................................................................. 1

TABLE OF CONTENTS .............................................................................................................. 2

ACKNOWLEDGEMENTS ........................................................................................................... 3

EXECUTIVE SUMMARY............................................................................................................. 4

1. INTRODUCTION............................................................................................................... 5

2. OBJECTIVES.................................................................................................................... 5

3. CONCEPT OF DELEGATED AIRBORNE SEPARATION DAS....................................... 6 3.1. DAS APPLICATION ...................................................................................................... 6

3.1.1. Objectives and benefits...................................................................................... 6 3.1.2. Broadcast parameters........................................................................................ 6 3.1.3. Delegation of separation from ATC to crew....................................................... 6

3.2. ATC SIMULATIONS SATSA ......................................................................................... 7 3.3. STATION KEEPING ....................................................................................................... 7

3.3.1. Types of Station Keeping................................................................................... 8 3.3.2. Scope of Station Keeping .................................................................................. 8 3.3.3. Time or Distance................................................................................................ 8 3.3.4. Definition of Separation...................................................................................... 8 3.3.5. Separation Assurance........................................................................................ 9

4. SIMULATION PREREQUISITES.................................................................................... 10 4.1. SIMULATION PARTICIPANTS ........................................................................................ 10 4.2. SIMULATOR ............................................................................................................... 10 4.3. TECHNICAL SYSTEM................................................................................................... 10

4.3.1. Hardware interface........................................................................................... 10 4.3.2. Software interface ............................................................................................ 11 4.3.2.1. MMI5000 software........................................................................................11 4.3.2.2. CATS as a synthesizer of data.....................................................................11 4.3.2.3. Communication protocol...............................................................................12 4.3.3. Data Streams ................................................................................................... 12

4.4. SCENARIOS ............................................................................................................... 12 5. SIMULATION ACCOMPLISHMENTS............................................................................. 13

5.1. BRIEFING .................................................................................................................. 13 5.2. FLIGHTS IN SIMULATOR .............................................................................................. 13 5.3. DEBRIEFING .............................................................................................................. 13 5.4. MODIFICATION DURING SIMULATION ............................................................................ 14

5.4.1. Scenarios ......................................................................................................... 14 5.4.2. Ground Speed input......................................................................................... 14 5.4.3. Fallback and escape procedures ..................................................................... 14 5.4.4. CDTI modification ............................................................................................ 14

6. ANALYSIS AND EVALUATION...................................................................................... 15 6.1. TECHNICAL ANALYSIS ................................................................................................ 15

6.1.1. Technical Analysis Objectives ......................................................................... 15 6.1.2. Technical Analysis Methods ............................................................................ 15 6.1.2.1. Limitations of the study.................................................................................15 6.1.2.2. Methods used to analyze data .....................................................................15 6.1.3. Technical Analysis Findings............................................................................. 16 6.1.3.1. Technical analysis findings in general..........................................................16 6.1.3.2. Approaches and departures.........................................................................17 6.1.3.3. Lateral deviation from track ..........................................................................17



6.1.3.4. Deviation from assigned separation.............................................................17 6.1.3.5. Distance........................................................................................................18 6.1.3.6. Time..............................................................................................................19 6.1.3.7. Excluding turn segments ..............................................................................19 6.1.3.8. Multi-link effects............................................................................................20 6.1.4. Conclusions from the technical analysis.......................................................... 21

6.2. HUMAN FACTORS HF ANALYSIS ................................................................................. 22 6.2.1. HF Analysis Objectives .................................................................................... 22 6.2.2. HF Analysis Methods ....................................................................................... 22 6.2.3. HF Analysis Findings ....................................................................................... 22 6.2.3.1. Limitations of the study.................................................................................22 6.2.3.2. Workload ......................................................................................................23 6.2.3.3. Man – Machine Interface (MMI) ...................................................................23 6.2.3.4. Operational remarks.....................................................................................23

6.3. OPERATIONAL ANALYSIS............................................................................................ 23 7. OPERATIONAL CONCLUSIONS ................................................................................... 24

7.1. DEPARTURES ............................................................................................................ 24 7.2. APPROACHES ............................................................................................................ 24 7.3. TIME/DISTANCE ......................................................................................................... 25 7.4. WORKLOAD............................................................................................................... 25 7.5. PILOT OPINIONS OF CONCEPT..................................................................................... 26

REFERENCES.......................................................................................................................... 26

APPENDICES ........................................................................................................................... 27 A. ABBREVIATIONS AND DEFINITION OF TERMS .................................................................... 27 B. NUP TIGER TEAMS ....................................................................................................... 33 C. ADS-B ......................................................................................................................... 34 D. CDTI – MMI5000......................................................................................................... 35 E. SCENARIO MAPS ........................................................................................................... 36 F. HUMAN FACTORS EVALUATION ...................................................................................... 37 G. CALL FOR VOLUNTARY PILOTS........................................................................................ 47 H. PILOT QUESTIONNAIRE .................................................................................................. 48 I. INSTRUCTOR QUESTIONNAIRE........................................................................................ 52

Acknowledgements The following persons have been directly involved in the simulations, the preparatory work and analysis associated with the simulations: Capt. Michael Agelii, SAS / Project Leader Capt. Arne Liljeqvist, SAS / Flight Instructor Flight Operations Engineer Christian Olausson, SAS / Technical Analysis Ret. Capt. Gunnar Fahlgren, Loop / Human Factors Analysis Phd. Leif Rydstedt, University College of Trollhättan Uddevalla / Human Factors Analysis In the technical development of simulation system software the following persons contributed: M.Sc. Per Åkesson, Carmenta AB / MMI5000 Software M.Sc. Örjan Råberg, Carmenta AB / CATS Software Flight Simulator Engineer Svein Rugland, SAS / Flight Simulator Software Flight Simulator Engineer Jan Johansson, SAS / Flight Simulator Software Many thanks to Saab Celsius Transponder Tech who contributed by permitting use of the SAAB VDL4 Binary Link Protocol. A special thanks goes to the pilots who volunteered in the simulations and greatly contributed with their professional skills, attitude and advice.



Executive summary Since the middle of the 1990’s Scandinavian Airlines System SAS (Flight Operations, Standards & Development) has taken an active part in research and development within the field of GNSS. The main reason for this involvement is to promote the development of the ADS-B concept and ASAS applications. Ultimately this is the road to an ATM system that will be able to cope with future air traffic demands. In line with the above stated intentions SAS performed simulations in the summer of 2000 to gain empirical knowledge about an ASAS application called Station Keeping. These simulations were performed as part of the R&D activities within the EU sponsored NUP (NEAN Update Programme) where SAS is one of seven partners. The Objective with this study was to gain operational knowledge about the concept of Station Keeping ("In Trail Spacing" is the american term). The authors do not claim scientific validity of the study. Instead these simulations are to be considered an empirical experiment in a near realistic operational environment. It is a complement to the many theoretical studies within the field of ASAS applications. Station Keeping is defined as a method of manoeuvering an a/c (trailer) to maintain a distance or sector relative another aircraft (target). During this procedure separation responsibility for the trailer relative its target is delegated to the crew of the trailer. When applying Station Keeping a defined minimum or a given time/distance will be assigned by ATC to the trailer. A trailer may be linked to a target who in turn is linked to another target. Such a chain of links is defined as a multi-link. A total of six SAS line pilots took part in the simulations. The simulator used in the trials was a full flight motion and visual MD-80 simulator. The simulator was equipped with a CDTI designated for Station Keeping guidance. The evaluation focuses on two main areas: 1. The technical evaluation analyzes navigational performance. 2. The Human Factors evaluation focuses on the pilot's operational behaviour. The findings are summarized in Operational Conclusions. The technical analysis findings can be summarized as follows: • Time separation seems to give higher performance than distance separation. • The limits for which the aircraft could stay within 95 % of the time, generally varied between (+/-) 5-10

seconds from assigned separation time. • Turns turned out to be the segment causing the largest degradation in separation performance. The

reason is twofold; definition of the separation distance and difficulty in maintaining target track during turn.

• Multi-link effects were briefly investigated by flying five aircraft following each other and executing Station Keeping in a chain. This limited experiment does not indicate any major problems with multi-link Station Keeping.

The human factors analysis findings can be summarized as follows: • Time separation gives smoother operations and less workload. • Much experience was gained regarding the pilot's information need and man-machine interface. • Much experience was gained about what factors tend to increase pilot workload and means to

mitigate increased workload. • Flying Station Keeping is a skill that can be learned and implemented in pilot training. • Most pilots expressed the feeling that the concept "felt right". It was expressed as a "natural" method

of operation that pilots control separation to preceeding aircraft once they have the tools. The operational conclusion from these simulations can be expressed in the following statements: • Time separation is by far the preferred separation criteria from an operational viewpoint. • It is highly recommended to develop FMS STAR´s that go all the way from TMA border to an

intercept of the LOC for Station Keeping procedures. This would greatly alleviate pilot workload. • The CDTI for operational use must be integrated with existing ND/PFD displays. • Station Keeping is feasible without Autopilot/FMS or Autothrottle system integration provided the

above points are considered. • Pilot controlled separation between same type of a/c is feasible with a +- 5 second precision.

*****************************



1. Introduction Since the middle of the 1990’s SAS (Flight Operations, Standards & Development) has taken an active part in research and development within the field of GNSS (Global Navigation Satellite System). The objective is to support the process of utilising satellite navigation techniques to develop operational applications within the field of Air Traffic Management and Navigation. The projects of CARD (CNS Applications Research & and Development) and NEAN (North European ADS-B Network) have focused on technical development issues. During autumn 1999 project structures to turn technical development into operational applications use were created. Operational teams (Tiger Teams, ref. appendix A) were formed in the NEAN Update Program (NUP). ATC and Airlines co-operate with the objective to develop new operational procedures within the gate-to-gate concept. Strong efforts and preparatory work have been put into developing a foundation for the certification of the procedures and technique. The teams have during the spring of 2000 worked on formulating descriptions and definitions to be the basis for safety analysis. The technique for datalink communication used within NUP, ADS-B/VDL Mode 4, has now been adopted by ICAO as a future global standard. The Flight Deck Simulations of Station Keeping was carried out during july – august of 2000 by SAS within the framework of NUP. The objective has been to aquire knowledge about the concept of Station Keeping when applied in a flight operational environment. The experience from these simulations will together with the ATC simulations performed by the swedish ATC SATSA (sect.. section 3.2) form the foundation for design of an airborne architechture that supports Delegated Airborne Separation (DAS)* applications. The simulation at SATSA was the first ATC simulation to be carried out within the framework of NUP, and is part of the process to confirm that the procedures for DAS meet ATC operational requirements. * DAS is an example of ASAS (Airborne Separation Assurance System).

2. Objectives These simulations do not claim scientific validity. Instead they are to be considered an empirical experiment with the general objective to gain operational knowledge about the concept of Station Keeping. The objective with Flight Deck Simulations of Station Keeping was: • To define limits with regard to the maneuvering of aircraft while executing Station Keeping. • To measure which relevance different flight parameters, (data to be broadcast) have on the precision

with which Station Keeping is being executed. • To map and evaluate the workload imposed on flight crew executing Station Keeping. • To outline safe, efficient and practical procedures and methodology to be used when flying Station

Keeping. Note: It was not the objective to develop or evaluate a GUI or other man-machine interface in the simulations. Nor was the objective to develop system functionality or technical solutions. The findings from these simulations will serve as the foundation for the development of GUI and systems which is to be performed within the frame of WP 3A within the NUP project.



3. Concept of Delegated Airborne Separation DAS

3.1. DAS Application The procedures for Delegated Airborne Separation (ref 1.) have been prepared and developed by the Stockholm Tiger Team (app. B).

3.1.1. Objectives and benefits The assumption is that delegation of separation to the pilot can increase efficiency in the TMA. The present situation of congestion in Arlanda TMA requires an increased throughput of traffic during peak hours. The ROT (Runway Occupancy Time) can be said to constitute the ultimate limit of the flow. It is desirable to achieve a flow that comes as close as possible to the ROT taking factors such as vortex into account. Today, radar resolution and conventional ATM procedures set the limit to a significantly higher value than the ROT. The precision in maintaining minimum separation is not sufficiently high. This is due to the uncertainty of exact position (radar resolution) and the long time lag in corrective adjustments. At first radar must reveal a discrepancy, then the controller must discover it and then decide on and transmit instructions. Thereafter, the pilot can initiate corrective actions. Effectively, there is too much slack generated in assuring the distances between aircraft. With the use of ADS-B technology (app. C) and separation assurance delegated to the pilot-in-command it is likely that separation assurance distances can be decreased, which will allow for separations closer to the ROT. The formal separation distances will not be decreased, but the precision in maintaining those distances will be much enhanced, thus improving the throughput of traffic.

3.1.2. Broadcast parameters ADS-B over data link with augmented DGNSS position updates serves as the technical foundation of DAS. Concept prerequisites are: • both ATC and aircraft will use ADS-B information; • suitably developed and adapted CDTI’s (Cockpit Display of Traffic Information) and CWP’s

(Controller Work Positions) must be available; • the CDTI shall be integrated with the aircraft FMS (Flight Management System); • CWP ADS-B data should be an integral part of the controller working tools; • datalinks will be used for transmission of the ADS-B data; basic ADS-B data broadcast parameters include: • callsign, • position (lat / long), • altitude, • ground speed, • track and • vertical speed. When DAS is implemented radar will be a complement to ADS-B for the foreseeable future (2015). During a relatively long implementation phase of ADS-B, radar position symbols can be re-broadcast via ADS-B from ground stations. This method is known as TIS-B and gives the opportunity of mixing ADS-B equipped aircraft with non-equipped for surveillance purposes, thus being displayed for both ATC and pilot.

3.1.3. Delegation of separation from ATC to crew To achieve DAS a procedure called Station Keeping (SK) (ref sect. 3.3) is used. SK is a method of manoeuvering the aircraft to maintain a position or sector relative another aircraft. SK is assigned to the flight crew from ATC. DAS can be used between aircraft irrespective of if they are destined to or originating from the same airport or runway. It can be used for climb, descent or aircraft at same level.



Eurocontrol EEC has defined three levels of delegation of separation assurance, (ref. 2.). In an increasing order these can be defined as limited, extended and full delegation. Limited delegation: The controller is in charge of both identification of a situation and management of a solution to a situation. Only implementation of solution and monitoring is delegated to the crew. Extended delegation: The controller is in charge of identification of a situation and delegate to the crew identification and implementation of a solution and the monitoring. Full delegation: The crew is responsible for all tasks related to separation assurance, identification of situations and solutions, implementation and monitoring. In the Arlanda DAS application “limited and extended” delegation will be applied. In the DAS concept identification and implementation of a solution and monitoring is delegated: - Choosing suitable speed and track to execute assigned SK is delegated to the crew of trailer. Controller-Pilot communication will be based on VHF radio with standardised phraseology. Data link messages (CPDLC) may constitute a future option.

3.2. ATC Simulations SATSA The ATC simulations was carried out to investigate the concept of DAS. The simulation was initiated by the NEAN update project (NUP) and prepared by the Stockholm NUP Tiger Team and the Swedish Air Traffic Services Academy (SATSA). In contrast to the Flight Deck Simulations dealt with in this report, the ATC simulations focused on ATM control and flow issues. (ref. 3.) The operational environment was Stockholm-Arlanda Terminal Control Area TMA. The traffic simulated was 38 landings/hr for runway 01R and 40 departures/hr for runway 08. Separate simulations were done for arrival and departure phases. The ATC simulation clearly demonstrated that Station Keeping (SK) in the arrival phase contributes to an increased capacity and a reduced workload on ATC. The precision of keeping minimum separation on final is improved and thus resulting in shorter flying distances. SK also leads to less monitoring of separation on final, spacing/vectoring tasks are minimised, and R/T load decreases. The Arrival and Director operators can concentrate on the planning, sequencing and positioning of aircraft for Station Keeping. The departure constraints for DAS in the simulation (aircraft in clean configuration mode i.e. 250KT / 4-5000ft before applying Station Keeping) made SK less applicable between departing aircraft from the same runway. Using SK will obviously increase pilot awareness as the pilot will be more involved and can support a smooth traffic flow. This is part of the fundamentals for collaborative decision making.

3.3. Station Keeping The phrase “Station Keeping” also includes the enhanced concept of “extended Station Keeping” in this paragraph. Station Keeping is defined as a method of maneuvering to maintain a distance or sector relative another aircraft. Station Keeping defines a longitudinal and lateral relationship between two aircraft. This relationship is called a link. The aircraft which the relative position or sector in a link is based upon is called the target. The aircraft responsible for executing the Station Keeping is called the trailer. When applying Station Keeping a defined minimum or a given time/distance will be assigned by ATC to the trailer. A trailer may be linked to a target who in turn is linked to another target. Such a chain of links is defined as a multi-link.



3.3.1. Types of Station Keeping Extended Station Keeping can be executed as a procedure using one of four methods, NAV TRAIL, TARGET TRAIL, OFFSET or BEHIND. OFFSET can be used in combination with TRAIL. • NAV TRAIL. The trailer is flying lateral according to a published or defined navigational clearance.

Maintaining a longitudinal relationship to the target. • TARGET TRAIL. The trailer is flying along the targets flown trajectory. Method based on targets

positions in past time, (position interval 10 sec). Maintaining a longitudinal relationship to the target. • BEHIND. The trailer is flying in a sector behind the target. This sector is bounded by the target

bearings 150 – 210 degrees relative the target true track. Maintaining a longitudinal relationship to the target. (Note: BEHIND was tested during the first trials and found to be impractical, hence abandoned in later trials.)

• OFFSET. The trailer is flying offset lateral according to a published or defined navigational clearance or the targets flown trajectory. Maintaining a longitudinal relationship to the target. (Note: OFFSET was not tested during the trials).

3.3.2. Scope of Station Keeping

Due to standard procedures during takeoff and first part of climbout speed control is a very limited option to use. Hence Station Keeping is not be applied until after the trailer has reached a standard climbout speed of 250 kts IAS in clean configuration. During approach it is impractical to apply Station Keeping in landing configuration. Hence Station Keeping ends at the outer marker when established inbound on the approach, (apprx. 1500'). Station Keeping can very well be applied in stable cruise conditions, however the scope of this report is limited to climbout and approach conditions.

3.3.3. Time or Distance Station Keeping is normally used in combination with a separation defined as an assigned time or distance. This separation can be defined as a distance in NM between the aircraft in the link, or as a projected flight time in seconds between the aircraft absolute positions in real time based on the trailers GS. The assigned time/distance may never be less than X sec/X NM as regulated by authorities or company procedures (suggested minimum time/dist. 65 sec / 2.5 NM). The assigned time/distance may depend on a/c type and/or airborne and ground equipment or the characteristics of the airfield/airspace etc.

3.3.4. Definition of Separation This definition is developed by Tiger Team Arlanda (appendix B) and enhanced by EADS (ref. 4.) The description of the separation minima in effect is a key element of the safety assessment of the end-to-end system. Specification of separation minima includes specification of horizontal (lateral, longitudinal), and vertical separation minima. Vertical separation : obtained by assigning different cruising levels. Longitudinal separation : obtained by maintaining an interval between aircrafts operating along the same, converging or reciprocal tracks; expressed in time or distance. Separation distance is the shortest possible distance to fly from any instant frozen trailer position (A) to the same instant frozen target position (C) including a shortest way turn with 25 degree bankangle (see figure 1). Special case: When 25 degree bankangle does not allow trailer to track to target pos (target within turn radius) then Separation distance is defined as great circle track between target and trailer. This definition is used as the official realtime separation.



Figure 1: Shortest distance to fly between A and C, is from A to B via s, then from B to C. Different, is the Minimum distance : great circle distance between target and trailer positions. Used as minimum safe distance, never to be broken. Lateral separation : obtained by maintaining aircrafts on different routes or in different geographical areas.

3.3.5. Separation Assurance Separation assurance is achieved by: 1. CAA defining a minimum safe separation. 2. ATC identifying and assigning a minimum or a given separation. 3. The crew executing the assigned separation. All types of Station Keeping must be used in combination with a minimum or a given separation. The separation will be defined as a time or distance between the aircraft in the link, (ref. sect. 3.3.3 above). When executing Station Keeping “extended delegation” of separation responsibility is applied, (ref. sect. 3.1.3 above). The value of the separation should be chosen such that any unpredicted lateral or vertical manoeuvre by any of the aircraft in the link will not infringe on safe separation. No immediate action should be required by the other aircraft. Consideration must also be given to the risk of vortex from the target aircraft when deciding on the separation. Vertical separation may be provided by ATC in combination with Station Keeping.



4. Simulation Prerequisites

4.1. Simulation participants A total of six SAS line pilots took part in the simulations. Three Captains and three First Officers were selected arbitrarily from a pool of 10 volunteers. The basis for selection was availability at respective session date and first come first serve. The pilots were divided in pairs of one Captain and one First officer and flew one full session of 4 hours per pair. The age of the Captains are in the 45-55 year range with a flight experience of around 10.000 flight hours. The F/O´s are in the 35-40 year range with a flight experience of some 2000-4000 hours. No one of the pilots had any previous experience of the concept of Station Keeping before the study.

4.2. Simulator The simulator used in the trials was a full flight motion and visual MD-80 simulator situated at SAS Flight Academy Stockholm/Arlanda airport. Type: Full flight simulator with IMAGE IV visual system. Manufactured by: SINGER - LINK MILES, delivered 1990, upgraded by CAE 1998. Certification: JAR STD 1A level DG Host computer: IBM RISC 6000. Motion system: 6 DOF Hydrostatic system

4.3. Technical system The purpose of the soft and hardware designed and used in the study was to provide means to perform the simulations with satisfactory results. It was never intended as development of avionics.

4.3.1. Hardware interface The MMI 5000 diplays used in the NEAN trials, has been used as cockpit hardware interface. It should be noted that the MMI 5000 hardware can be used for real flight tests at a later stage. The hardware was used mainly for display purposes. Little effort has been put into the design of the display interface. A few inputs through hardware is used e.g. start/stop, mode changes etc. The design of the input interface to the hardware was of little concern for these simulations.

Figure 2: MMI 5000 Cockpit Display (CDTI)



4.3.2. Software interface The software used in the simulations was the MMI5000 internal software, the CATS data synthesizing software and the VDL4 Binary Link Protocol for data communication between the flight simulator and the CATS platform.

4.3.2.1. MMI5000 software The MMI5000 CDTI operates with a nav/com application software based on the QNX realtime OS. This software includes moving map, flightplan, nav data base, rwy incursion warning, CPDLC etc. A specifically designed Station Keeping module has been integrated. Version 1.58 and 1.60 has been used during the test. The software is designed by CARMENTA AB in the C programming language.

Figure 3: Station Keeping module of the MMI 5000 software interface (appendix D)

4.3.2.2. CATS as a synthesizer of data CATS is the acronym for Carmenta Air Traffic Surveyor. This WIN NT based application is a tool for displaying ADS-B vehicles on a map. It has the functionality to record data from ADS-B units over time and then replay recorded data. This feature has been used during the simulations to record data from the flight simulator. CATS has also been the tool for synthesizing recorded data with realtime data. The synthesized data is then transmitted onto a serial connected MMI5000. At the same time the new synthesized data can be recorded.

Figure 4: CATS software interface showing a multilink of 5 a/c



4.3.2.3. Communication protocol The protocol used for communication of data is VDL mode 4 (VDL4). The update rate used was one burst of data per second. VDL4 is defined in ICAO SARPS (ref. 5) and the implementation used here is the SAAB Binary Link Protocol.

4.3.3. Data Streams The simulations comprises four main streams with regard to data. • Generating target aircraft data • Generating simulator flight data • Display of synthesized target and simulator flight data • Recording of synthesized target and simulator flight data Target a/c data consists of recorded flight data from earlier simulator flights. Such flights comprises both scenario flights and flights being part of the study.

Figure 5: Schematic design of system data flow as implemented in these simulations. Simulator flight data was generated as realtime input from the simulator with the following parameters: • callsign • position (lat / long) • altitude, (barometric) • ground speed • true track • vertical trajectory • roll (bank) angle The data processing for evaluation purposes has been handled outside the actual simulation process and performed as far as possible with standard database and calculation software. The input to the evaluation software is the data recorded in files by the simulation software.

4.4. Scenarios The prerequisites for the scenarios were: ISO atmospheric conditions, no wind, 54 metric tons grossweight (medium weight MD-80). SAS standard operating procedures complemented with Station Keeping procedures were used. For details on the scenario SIDs and STARs used see appendix E.

MultiMode CATS(Synthesizing software run on a laptop)

MMI5000(CDTI)RS232

Recorded Target a/c data

Flight simulator(real time data) RS232

Laptop(running WIN NT)



5. Simulation Accomplishments

5.1. Briefing The week before each simulator session the respective pilot received a selection of documentation presenting the simulations, the concept of Station Keeping and some technical background material. A total of some 4-5 hours reading of which apprx. 2 hours was recommended to be read before the simulations. A two hour briefing was performed in connection with each simulator session. The briefing focused on the operational concept, its difficulties and possibilities. It was stressed that the input received from the participating pilots during and after the session were to be an extremely vital part of the evaluation.

5.2. Flights in simulator Three sessions lasting 4 hours each were performed. The first two hours of each session was flown as a Standard Instrument Departure SID followed by SK airwork at FL 180 – 240 and ended with a Standard Arrival STAR, approach and landing. The objective with these first two hours was to let the pilots get familiar with flying SK. These flights were recorded but has not been analyzed in the technical evaluation since they must be regarded as practice only. The last two hours were spent conducting one takeoff and climbout followed by 3 – 4 approaches. The objective here was to observe pilot behavior and measure SK performance in a simulated normal operational environment. During the entire session pilots alternated being Pilot Flying PF and Pilot Not Flying PNF for each complete departure or approach. The data recorded from these flights constitute the basis for the technical evaluation.

Figure 6: Pilots executing Station Keeping, CDTI installed on top of glareshield.

5.3. Debriefing During the debriefing which lasted for approximately one hour the pilots were asked to give there opinion on the concept, methods and procedures and also the CDTI interface in the form of an informal interview/discussion. During the debriefing both pilots, the project leader, the instructor and at least one human factors specialist were present. The pilots were also asked to answer questions in a questionnaire regarding the information in the CDTI. Finally each pilot was asked to fill out a HF questionnaire at home within the next few days.



5.4. Modification during simulation

5.4.1. Scenarios Initially quite a few different scenarios to different runways was planned. During the study it became apparent that it would be more useful to concentrate on fewer scenarios since the operational problems would be more or less the same for all scenarios. By limiting the number of scenarios to a total of 4 with an increased number of runs for each scenario the basis for evaluation was improved. Due to the fact that time was found to be superior compared to distance (ref. sect. 7.3) it was decided to focus on time based separation rather than distance based. Hence the low number of distance scenarios tried.

5.4.2. Ground Speed input Due to a misunderstanding during software development the speed parameter tapped from the simulator for the first of three study sessions was Indicated Airspeed IAS instead of Ground Speed GS. The impact on the study due to this mishap was minimal. The only deviation revealed was that the own GS displayed was too low (GS = IAS). For the following simulations this mishap was corrected by adjusting the software. It is considered that this deviation did not affect the outcome of the study in any way.

5.4.3. Fallback and escape procedures Initially fallback and escape procedures were planned to be tested. However, due to time and availability constraints concerning the flight simulator it was decided to postpone this part of the study to a later time and instead concentrate on the main objectives (see section 2).

5.4.4. CDTI modification After the first simulator session some modifications to the MMI software was implemented as a result of pilot response and input from that first session. These modifications included: • Upside-down turn of distance/time tape • Modified scale on distance/time tape • Enhanced algorithm for dist/time calculation • Boundary markers and arcs displayed • Warning when target IAS changes

Figure 7: Simulation data in Microsoft Excel format

File version: 2.0Datasource numberFormat Type Device name

8 VDLM4 Serial COM111 PLAYBACKFile C:\Program\Carmenta\MultiMode CATS\STKP 1.0\Log Files\trs2tb.CRF

Datasource numberIdentity Format Date Time MillisecondsCategory Latitude Longitude Speed Course Altitude Vert. speedBanking AngleReceived data8 CC0001 VDLM4 2000-08-06 12:12:37 464 131 58.827980 17.323332 287.49460042.173733 19002.296588N/A N/A 4b cc 00 01 01 c8 dc 09 15 96 ff 31 06 c7 05 dc 1e 15 0f 76 d6 08 00 05 03 dc 18 45 00 00 00 00 00 00 008 CC0001 VDLM4 2000-08-06 12:12:37 754 131 N/A N/A N/A N/A N/A 100.0000000.000000 50 cc 00 01 01 01 f6 81 008 CC0001 VDLM4 2000-08-06 12:12:38 485 131 58.827980 17.323332 287.49460042.173733 19002.296588N/A N/A 4b cc 00 01 01 c8 dc 09 15 96 ff 31 06 c7 05 dc 1e 15 0f 76 d6 08 00 05 03 dc fc 48 00 00 00 00 00 00 008 CC0001 VDLM4 2000-08-06 12:12:38 756 131 N/A N/A N/A N/A N/A 100.0000000.000000 50 cc 00 01 01 01 f6 81 008 CC0001 VDLM4 2000-08-06 12:12:39 487 131 58.827980 17.323332 287.49460042.173733 19002.296588N/A N/A 4b cc 00 01 01 c8 dc 09 15 96 ff 31 06 c7 05 dc 1e 15 0f 76 d6 08 00 05 03 dc f1 4c 00 00 00 00 00 00 008 CC0001 VDLM4 2000-08-06 12:12:39 767 131 N/A N/A N/A N/A N/A 100.0000000.000000 50 cc 00 01 01 01 f6 81 00

11 BB0004 VDLM4 2000-08-06 12:12:40 298 131 N/A N/A N/A N/A N/A -1600.0000000.000000 50 bb 00 04 01 01 f6 f0 008 CC0001 VDLM4 2000-08-06 12:12:40 508 131 58.828182 17.323683 287.49460042.174309 19002.821522N/A N/A 4b cc 00 01 01 85 e1 09 15 d0 07 32 06 c7 05 5c 2c 15 0f 86 d6 08 00 05 03 dc d5 50 00 00 00 00 00 00 008 CC0001 VDLM4 2000-08-06 12:12:40 769 131 N/A N/A N/A N/A N/A 200.0000000.000000 50 cc 00 01 01 01 f6 82 00

11 BB0004 VDLM4 2000-08-06 12:12:41 39 131 58.902172 17.433952 286.71706342.540413 17991.272966N/A N/A 4b bb 00 04 01 a8 a7 10 15 41 20 3c 06 c3 05 ed b0 36 0f 16 5e 08 00 05 00 3b 60 d7 00 00 00 00 00 00 0011 BB0004 VDLM4 2000-08-06 12:12:41 319 131 N/A N/A N/A N/A N/A -1600.0000000.000000 50 bb 00 04 01 01 f6 f0 008 CC0001 VDLM4 2000-08-06 12:12:41 510 131 58.829578 17.326092 286.91144742.166485 19006.889764N/A N/A 4b cc 00 01 01 3e 02 0a 15 46 40 32 06 c4 05 fc 74 14 0f 02 d7 08 00 05 03 dc da 54 00 00 00 00 00 00 008 CC0001 VDLM4 2000-08-06 12:12:41 790 131 N/A N/A N/A N/A N/A 200.0000000.000000 50 cc 00 01 01 01 f6 82 00

11 BB0004 VDLM4 2000-08-06 12:12:42 61 131 58.903435 17.436192 286.71706342.509647 17962.565617N/A N/A 4b bb 00 04 01 40 c5 10 15 bd 54 3c 06 c3 05 dc df 33 0f ab 5a 08 00 05 00 3b 55 db 00 00 00 00 00 00 0011 BB0004 VDLM4 2000-08-06 12:12:42 341 131 N/A N/A N/A N/A N/A -1600.0000000.000000 50 bb 00 04 01 01 f6 f0 008 CC0001 VDLM4 2000-08-06 12:12:42 531 131 58.830929 17.328419 286.52267842.141590 19011.384514N/A N/A 4b cc 00 01 01 e4 21 0a 15 d0 76 32 06 c2 05 84 2d 12 0f 8b d7 08 00 05 03 dc be 58 00 00 00 00 00 00 008 CC0001 VDLM4 2000-08-06 12:12:42 792 131 N/A N/A N/A N/A N/A 200.0000000.000000 50 cc 00 01 01 01 f6 82 00

11 BB0004 VDLM4 2000-08-06 12:12:43 72 131 58.904739 17.438503 286.71706342.484260 17933.858268N/A N/A 4b bb 00 04 01 d4 e3 10 15 eb 8a 3c 06 c3 05 db 8c 31 0f 40 57 08 00 05 00 3b 4a df 00 00 00 00 00 00 0011 BB0004 VDLM4 2000-08-06 12:12:43 362 131 N/A N/A N/A N/A N/A -1600.0000000.000000 50 bb 00 04 01 01 f6 f0 008 CC0001 VDLM4 2000-08-06 12:12:43 543 131 58.832325 17.330820 286.13390942.115379 19014.173228N/A N/A 4b cc 00 01 01 9d 42 0a 15 18 af 32 06 c0 05 34 c7 0f 0f e0 d7 08 00 05 03 dc b3 5c 00 00 00 00 00 00 008 CC0001 VDLM4 2000-08-06 12:12:43 813 131 N/A N/A N/A N/A N/A 0.000000 0.000000 50 cc 00 01 01 01 f6 80 00

11 BB0004 VDLM4 2000-08-06 12:12:44 144 131 58.906006 17.440739 286.71706342.466320 17905.150919N/A N/A 4b bb 00 04 01 83 01 11 15 50 bf 3c 06 c3 05 60 e8 2f 0f d5 53 08 00 05 00 3b 3e e3 00 00 00 00 00 00 0011 BB0004 VDLM4 2000-08-06 12:12:44 364 131 N/A N/A N/A N/A N/A -1600.0000000.000000 50 bb 00 04 01 01 f6 f0 00



6. Analysis and Evaluation

6.1. Technical Analysis

6.1.1. Technical Analysis Objectives The objectives with the trials from a technical point of view was to: • Produce data as an input to the evaluation of the viability of Station Keeping as a method to control

the traffic flow. • Compare time to distance as separation variables with respect to precision. • Investigate the limits to be required for the flight crew to perform within.

6.1.2. Technical Analysis Methods 6.1.2.1. Limitations of the study

Several interesting results were retrieved from the available data. However, there were some limitations that make it desirable to do further testing in some areas before conclusions can be regarded as reliable. • One of the major limitations of the trials was the relatively small amount of data available. There were

three simulator sessions lasting four hours each. Effective "airborne" time for analysis was about two hours each session. A total of twelve flights could be regarded as useful. Some of the other flights were regarded as training only and a few were discarded due to technical failure or project management mistakes.

• Minor software changes were made after the first session (ref sect. 5.4.2 and 5.4.4), which may have affected the performance of the flight crew.

• There were few flights made with distance as separation variable, which makes it difficult to make a fair comparison between distance and time.

6.1.2.2. Methods used to analyze data Twelve flights were analyzed. They were distributed on the different scenarios according to table A. The scenarios are described in appendix E.

Scen. # Dist. Time Tot. 4 - 2 2 5 1 7 8 6 - 1 1 7 1 - 1 S 12

Table A: Distribution of flights All desired data was collected from the simulator, converted to and analyzed in Microsoft Excel format. Excel was a natural choice as the CATS software has a built-in feature to convert data from crf-format (CATS record format) to Excel format. Example of an Excel data file is shown in figure 7. The Excel files contained data for each aircraft, updated once every second. Data for the trailing aircraft was displayed on one row and data for the target aircraft on the following row. Thus, separation distance (and time) was calculated from the position on row A to the position on row B, from C to D etc. However, occasionally a data row would be missing and therefore causing a very short distance to be calculated, from one aircraft's position to its own position one second later. In those cases the data have been manually edited, to make graphs easier to analyze. It was decided that it would be suitable to analyze the variation of a distance and its corresponding time between two aircraft, based on the ground speed of the trailing aircraft. The great circle distance between two aircraft's actual positions was considered to be irrelevant in all cases but that when one aircraft is flying along the same or opposite track. Therefore a variable called separation distance was defined.



Separation distance was defined as the distance to fly a 25-degree banking turn, from the current track to the track towards the target aircraft at the end of that turn, plus the distance from the end of the turn to the target aircraft's position (ref sect. 3.3.4). The time variable corresponding to separation distance was called separation time and was defined as the time it would take to fly the calculated separation distance with the aircraft's present ground speed. To achieve conformity between information displayed to pilots on CDTI and the analysis, the source code for the calculation of separation distance and separation time from the CDTI software was used. It was manually implemented into an Excel worksheet for the analysis. Initially limits were defined as an aim for the pilots to perform within. When flying with time as separation variable, the goal would be to stay within +/- 5 seconds from the assigned separation time. When flying with distance as separation variable, the corresponding limits would be +/- 0.2 NM. Several ways to evaluate the collected data were suggested by asking a number of questions: • What percentage of the flying time did the pilot flying keep the aircraft within the defined limits? • What limits would correspond to the pilot flying to keep the aircraft within those limits for at least 95%

of the time? • What is the maximum (positive and negative) deviation from the assigned distance or time during the

chosen scenario? • What specific events/maneuvers during the evaluated scenario causes the worst performance

according to the answers of the questions above? • If several aircraft fly in multi-link formation and one deviates from its assigned separation time or

distance, how will it affect the following aircraft? Will the deviation increase or decrease with each following aircraft?

Maximum deviation from assigned separation distance or time is defined as actual separation minus assigned separation, i.e. positive deviation means that separation is larger than intended. The separation variables were plotted as a function of time. The flight paths were also plotted to display graphically, during which part/parts of the flight separation increased or decreased. Findings from the technical analysis has been coordinated with information from the notes taken by the instructor during the simulations.

6.1.3. Technical Analysis Findings 6.1.3.1. Technical analysis findings in general

The distance generally varies a lot during the initial and final stages of a flight, while time varies much less. As indicated from the ATC simulations, time is also more interesting from a flow control point of view (ref sect. 3.2). After some simulation flights, it was therefore decided to focus on time as separation variable, which is the reason why there are relatively few flights to analyze with distance as separation variable. Many results show performance by the flight crews, better than expected. Most flights contained few mistakes by the pilot. Considering the overall results it is likely that these mistakes would have been avoided with more preparatory training. Table B shows the results from the analysis of twelve flights performed in the simulator. The flights have been numbered from 1 to 12. For each flight the table states scenario, crew (A - C), whether the flight was an approach or a departure and separation variable. The table also states percent of time that the aircraft was within defined limits, which limits would be required to stay within limits for 95 % of the time and finally maximum positive and negative deviation from assigned separation distance or time.



ID Scen.# Crew# Arr/Dep Dist/Time Within lim. [%] 95% lim. Max pos. Max neg.

1 5 A Appr. 70 sec. 69 % 8.7 sec. 8.0 sec. 10 sec. 2 5 A Appr. 70 sec. 79 % 8.5 sec. 14 sec. 5.8 sec. 3 6 A Appr. 70 sec. 83 % 9.4 sec. 5.6 sec. 17 sec. 4 7 A Appr. Dist. 30 % * 2.3 NM 1.7 NM 2.4 NM 5 5 B Appr. 60 sec. 95 % 5.1 sec. 2.5 sec. 8.8 sec. 6 4 B Dep. 60 sec. 74 % 11 sec. 3.0 sec. 12 sec. 7 5 B Appr. Dist. 59 % * 1.8 NM 0.048 NM 0.39 NM 8 4 C Dep. 60 sec. 69 % 9.8 sec. 1.2 sec. 10 sec. 9 5 B Appr. 70 sec. 95 % 5.3 sec. 4.5 sec. 12 sec.

10 5 B Appr. 50 sec. 91 % 9.4 sec. 10 sec. 19 sec. 11 5 C Appr. 50 sec. 83 % 19 sec. 7.2 sec. 26 sec. 12 5 C Appr. 50 sec. 94 % 5.6 sec. 2.3 sec. 10 sec.

* Commented under 6.1.3.5 Distance below. Table B: Compilation of results from the analysis. As can be seen from table B, most approaches were performed according to scenario 5. These approaches were made to Arlanda runway 26, via STAR TRS 2T. The approach route ends with an almost 180-degree turn to final (see picture 8). For all flights made with this scenario, the 180-degree turn turned out to be the segment that caused the worst performance.

6.1.3.2. Approaches and departures Only two departures were analyzed and both of them were performed with time as separation variable. The first flight was executed by crew B, and the second flight by crew C. Both flights show quite similar results. Ten approaches were analyzed, of which two were performed with distance as separation variable. It seems as if results are better for the approaches than for the departures, but only two departures cannot be considered enough to draw any conclusions about that. Basic maps of the approach and departure scenarios used can be found in appendix E.

6.1.3.3. Lateral deviation from track Deviation from track has been analyzed by manually observing graphs comparing target´s and trailer´s plotted tracks, (see figure 8). Since track deviation was found to be minimal in all cases no further analysis was deemed necessary. Segments performed in NAV mode (from TRS to TEB) is of course in line with ordinary RNAV performance. Segments in target trail (from TEB to Localizer) sometimes did show a tendency to deviate during turns. This deviation must be considered minimal from an RNP perspective. However it may affect performance in maintaining assigned separation, (ref sect. 6.1.3.4). Also in the case with multi-link (ref sect. 6.1.3.8) no significant deviations were observed.

6.1.3.4. Deviation from assigned separation When looking at the maximum deviation from the assigned separation distance or time, it can be seen that the magnitude of the deviation, is larger on the negative side than on the positive side. Positive deviation is defined as actual separation minus assigned separation. In all cases but one (flight 2, table B).It is also evident that negative deviation outweighs positive deviation with respect to the time deviated outside of limits. In part this may be due to the fact that it is generally faster to accelerate than to decelerate. In most cases the pilot flying didn't change the aircraft's configuration or use speed brakes to decelerate the aircraft.

Figure 8: Flight path for target and trailing aircraft approaching Arlanda runway 26.



However, the most evident reason for the negative deviations being predominant can probably be attributed to the fact that the definition of separation distance/time as implemented here, (ref sect. 3.3.4) does not consider the actual track that the trailer will follow behind the target. Accordingly the design of the algorithm used to calculate the separation distance in the CDTI is based on that same definition. This design feature was however pointed out by the instructor and considered by the pilots during the simulations. In sharp turns pilots let the indicated separation distance decrease on purpose since this is the natural result of the separation algorithm being calculated as "shortest turn" direct to trailer. By using such methodology the pilots would be close to the assigned separation when the turn along target track was completed. Pilots comprehended and learned this method quite quickly. Hence the negative deviation will be larger in both max deviation and amount of time deviated. This flaw in the design can of course be overcome with a definition of separation distance/time that follows target trail. Flight 2 was performed with positive deviation for most part of the flight and the largest positive deviation, 14 seconds, was due to a mistake during the left turn to final. According to the instructor present in cockpit it was a mistake, typically caused by lack of position awareness. The maximum positive deviation seems to increase with increasing separation time. Also this observation can with high certainty be attributed to the reason described above concerning the definition of the separation algorithm.

6.1.3.5. Distance Two of the twelve flights, that were subjects to the analysis, were performed with distance as separation variable. Comparing the percent of time that the aircraft was within the defined limits for each flight shows that using time as separation variable yields significantly better results. Flight 4 is not to be considered as a representative result for the simulations, but as a good example of how one mistake affects the overall result causing a complete failure. The pilot flying made a mistake that caused the separation to increase too much to be possible to correct by increasing speed. Flight 7 shows that the aircraft was within defined limits for 59 % of the time. However, this includes the time from the moment when the assigned separation distance is changed from 5 NM to 3 NM, until the actual separation distance again is within limits, i.e. in this case less than 3.2 NM. In this case it causes separation out of limit for 3 minutes and 3 seconds. The rest of time during this flight the aircraft was within the +/- 0.2 NM limit for 71 % of the time (see figures 9 and 10). Nevertheless, this is only in level with the worst results from the flights performed with time as separation variable. Flight 7 was within 1.8 NM from assigned separation distance for 95 % of the time.

0

1

2

3

4

5

6

07:45

:36

07:46

:37

07:47

:38

07:48

:39

07:49

:40

07:50

:40

07:51

:41

07:52

:43

07:53

:44

07:54

:45

07:55

:46

07:56

:47

07:57

:49

07:58

:50

07:59

:51

08:00

:52

08:01

:53

08:02

:55

Time UTC

Sep

arat

ion

dis

tan

ce

Figure 9: Separation distance, Flight 7.



-0.5

0

0.5

1

1.5

2

07:45

:36

07:46

:37

07:47

:38

07:48

:39

07:49

:40

07:50

:40

07:51

:41

07:52

:43

07:53

:44

07:54

:45

07:55

:46

07:56

:47

07:57

:49

07:58

:50

07:59

:51

08:00

:52

08:01

:53

08:02

:55

Time UTC

Sep

arat

ion

dis

tan

ce

Figure 10: Deviation from assigned separation distance, Flight 7.

6.1.3.6. Time Being able to fly with a constant separation variable throughout the scenario proved to be an advantage, just as predicted. Generally deviations from the assigned separation time can be derived to maneuvers (mainly turns, for example the earlier mentioned 180-degree turn on approach to Arlanda runway 26) and single mistakes from the pilot flying. Interesting observations from flights performed with time as separation variable: • The defined limit of keeping the aircraft within +/- 5 seconds from assigned separation time was

fulfilled 95 % of the time in two flights, no. 5 and 9 (see table B). • The maximum positive deviation was generally below 10 seconds. • The maximum negative deviation was generally between 10-20 seconds. • The limits for which the aircraft could stay within 95 % of the time, generally varied between (+/-) 5-10

seconds. • Six of eight "approaches" stayed within limits for more than 83 % of the time during the flight. • A large amount of the deviation can be attributed to phases of flight which include large track

deviations between target and trailer (turns). The main reason is likely to be flaws in the separation algorithm used, as mentioned earlier (ref. sect. 6.1.3.4).

6.1.3.7. Excluding turn segments One factor affecting separation performance during turns is lateral deviation from track. If deviating slightly on the inside or outside of a turn the separation distance as defined in these trials is notably affected. Considering this fact together with the problem of the separation algorithm mentioned earlier, it was considered necessary to analyze also the case when excluding turns. Hence, a turn was defined as any segment of the flight, where the difference between trailing aircraft's and target aircraft's course was more than 10 degrees. The data for these segments were then excluded and the remaining data, analyzed again with respect to: • Percentage of time within +/- 5 seconds from assigned separation time. • Max positive deviation from assigned separation time. • Max negative deviation from assigned separation time.

It can be concluded that, if excluding the turn segments of a flight from the analysis, the overall performance is improved. The only exception is flight 1, where a human error made performance deteriorate during a non-turn segment.



All flights except flight 1 were within +/-5 seconds from assigned separation time for more than 82 % of the time. Two flights were within limits 100 % of the time (see Table C). The average time within limits was 88.6 % as compared to 83.2 % when including turns. If excluding flight 1 (see comment above) the average time within limits on "straight track" segments is 92.2 %. As can be seen from Table C, the maximum deviations also changes in most cases when excluding turns. However, no significant patterns can be noted. Maximum negative deviation is equal to or less than 10 seconds for all cases except flight 11.

Incl. turns Excl. turns ID Arr/Dep Dist/Time Within lim.

[%] Max pos. Max neg. Within lim.

[%] Max pos. Max neg.

1 Appr. 70 sec. 69 % 8.0 sec. 10 sec. 57 % 8.0 sec. 10 sec. 2 Appr. 70 sec. 79 % 14 sec. 5.8 sec. 82 % 12 sec. 5.8 sec. 3 Appr. 70 sec. 83 % 5.6 sec. 17 sec. 87 % 5.4 sec. 2.9 sec. 5 Appr. 60 sec. 95 % 2.5 sec. 8.8 sec. 100 % 2.0 sec. 4.5 sec. 6 Dep. 60 sec. 74 % 3.0 sec. 12 sec. 90 % 3.0 sec. 10 sec. 8 Dep. 60 sec. 69 % 1.2 sec. 10 sec. 90 % 1.2 sec. 6.7 sec. 9 Appr. 70 sec. 95 % 4.5 sec. 12 sec. 100 % 4.5 sec. 3.0 sec.

10 Appr. 50 sec. 91 % 10 sec. 19 sec. 99 % 5.3 sec. 2.9 sec. 11 Appr. 50 sec. 83 % 7.2 sec. 26 sec. 85 % 6.7 sec. 23 sec. 12 Appr. 50 sec. 94 % 2.3 sec. 10 sec. 96 % 2.3 sec. 5.9 sec.

Table C: Performance including turns vs excluding turns The pattern seen when turns are included in the analysis, showing that maximum negative deviation generally is larger than the positive deviation, is not quite as obvious when excluding the turns. This fact verifies the conclusion that the definition of separation distance/time affects performance significantly in turn segments. Still, the largest negative deviations are larger than the largest positive deviation. This may depend on the fact that an aircraft decelerates slower than it accelerates as mentioned earlier. Since one of the two flights executed with distance as separation variable included a pilot error affecting the performance rather severely, those flights were not analyzed under these conditions.

6.1.3.8. Multi-link effects During the simulator sessions it was tried whether it was possible to "connect" several aircraft, all executing Station Keeping, together. Five aircraft formed a chain, arriving to Arlanda runway 26 with separation time 70 seconds between the first and second aircraft, and 50 seconds between the rest of the aircraft. The results from this flight were evaluated partly as separate flights (table B, flights 9-12), partly altogether in one graph (figure 11). In the graph (figure 11) it can be noted that separation distance suddenly drops to zero in one instant. The reason for this is a definition made to cover up for the case where the distance between the trailing aircraft and the target aircraft is too short to be able to calculate a 25-degree banking turn to target's position. The CDTI in these cases displays the great circle distance between the two positions, since the computer program cannot calculate a numerical value for the present conditions (ref.sect. 3.3.4).



-30

-25

-20

-15

-10

-5

0

5

10

15

15:59

:59

16:02

:52

16:05

:45

16:08

:38

16:11

:31

16:14

:23

16:17

:16

Time UTC

a/c 2 - a/c 1

a/c 3 - a/c 2

a/c 4 - a/c 3

a/c 5 - a/c 4

Figure 11: Deviation from separation times for a chain of aircraft (multi-link) performing Station Keeping. One of the interesting points by analyzing a scenario with several aircraft participating, is to see how an event executed by one aircraft (manoeuver, mistake etc) affects the following aircraft, i.e. if the effect is damped out or if it becomes uncontrollable. It is possible to see an event (see picture 11), time 16:10, first trailing aircraft) growing quite a bit through the first two links in the chain, but eventually decrease. This scenario would need to be improved, by adding at least three or four more aircraft and letting the first aircraft perform a schedule of events.

6.1.4. Conclusions from the technical analysis • Data from relatively few flights were available for the analysis. A larger number of flights would be

needed to verify observations made during the analysis. It would probably be possible to achieve results that could be used to establish realistic limits of variations of separation distances and times by performing a larger number of simulations. Also to answer questions that emerged during the evaluation of the collected data, more simulations would be desirable from some of the scenarios.

• The maximum deviation and amount of time deviated from assigned separation distance or time is generally larger on the negative side (closer to target than assigned). This is most likely due to two separate causes. 1. There is a limitation in the definition of the separation algorithm used in the trials. 2. Deceleration is slower than acceleration.

• When using distance as separation variable the overall performance depends notably on whether the time from change of assigned separation distance is included or not. One example showed that the aircraft stayed within +/- 0.2 NM for 59 % of the time, but excluding the period of time where the aircraft adjusted its actual separation distance (from 5 to 3NM), gave a corresponding result of 71 %.

• Time as the parameter for separation turned out to give the best overall results. • Turns i.e. differences in target and trailer tracks turned out to be the segment causing the largest

degradation in separation performance. The reason is twofold; definition of the separation distance and difficulty in maintaining target track during turn.

• Multi-link effects were briefly investigated by flying five aircraft following each other and executing Station Keeping in a chain. This limited experiment does not indicate any major problems with multi-link Station Keeping. This scenario would need to be improved, by adding more aircraft and letting the first aircraft perform a schedule of "events".

Dev

iati

on

fro

m a

ssig

ned

sep

arat

ion

in s

ec.



6.2. Human Factors HF Analysis Under this chapter only a brief summary of the objectives, methods and findings with regard to human factors is discussed. For the full report on the human factors analysis of the simulations refer to appendix F.

Figure 12: Flight instructor observing pilot performance and taking notes

6.2.1. HF Analysis Objectives The objective with the human factors analysis was to evaluate how the Station Keeping affects the pilots and there work environment concept when applied operationally. Some specific objectives can be derived: • How is pilot workload influenced? • What can be learned concerning the man-machine interface? • Observations about the practicality of different procedures. • How will pilots react to a new operational paradigm of ATM?

6.2.2. HF Analysis Methods Only six pilots participated in the trials. They were all male and had a long experience in the profession, with an average pilot duty close to 20 years. All six pilots were chosen from a small stock of pilots that had answered a call for volunteers. They were chosen only on basis of availability at the dates of the trials. These pilots received a written information about Station Keeping and before the simulator flight the objectives as well as the detailed instructions were given during a two hours briefing. After the simulations they were interviewed and asked to complete 2 different questionnaires. In addition the instructor´s observations has been taken into account. After each simulation the instructor answered a questionnaire regarding his observations. Instructor notes taken during the trials has also been analysed in connection with the operational and technical analysis of logged data.

6.2.3. HF Analysis Findings 6.2.3.1. Limitations of the study

When interpreting the results from this study, we must of course be aware of its limitations. The small number of participants in the test group and participation based on self-selection reduces the internal validity of the study. It must be clearly stated that the results from this limited study cannot be generalized, nor do the authors claim to have identified all possible problematic aspects of Station Keeping. The presented study should rather be considered as explorative and attempting to illuminate possible problems and advantages associated with Station Keeping.



It must be recognized that most of the problematic aspects of Station Keeping, found in the present study, can be improved, or even overcome, by re-design, proper training and education and by more developed routines and increased familiarity with the procedures and the equipment.

6.2.3.2. Workload One of the more critical findings, in this initial simulation flight with Station Keeping, was the increased workload reported by several of the pilots. Three of the six participants in the study reported that they had to make decisions under time pressure. Decision making under time pressure always occurs with some frequency in this type of work, and from this study the occurrence of time pressure compared with the ordinary work situation cannot be concluded. Some of the participants found that the increment in available information and improved overview more than compensated for the new work tasks. One pilot actually mentioned a percieved decreased workload while operating in NAV trail mode as there was less ATC instructions than normal. We must be aware of the potential risk in adding yet another task to the already complex and demanding work of the aircrew. It can although be expected that a more developed version of the equipment, combined with proper training and education in handling the system, substantially will reduce the perceived workload reported in this study. Furthermore, if we take into account that the majority of pilots expected that Station Keeping would improve flight safety, a possible conclusion is that the improved overview more than compensates for the increment in demands.

6.2.3.3. Man – Machine Interface (MMI) Another critical finding from this study, which definitely calls for further attention is the quality and intelligibility of the information presented on the display. Although some improvements were made already after the first simulator flight, several critical comments were made about vagueness and difficulty in interpreting some of the information. This evaluation gave a lot of information about the design and placement of the display as well as what kind of information the participants wanted. We are convinced that the quality of the presented information can and will be improved in the future.

6.2.3.4. Operational remarks Station Keeping with Time separation was so unanimously regarded as the best method of maintaining a separation when executing Station Keeping. Flying Station Keeping is like a prolonged semi automatic ILS or VOR approach when the Pilot Flying (PF) is occupied with the manoeuvring of the aircraft and Pilot Not Flying takes care of ATC communication, checklist reading, monitoring and serving the PF. From the remarks of the participants, there are sound reasons to believe that the high Passenger Comfort will remain unaffected by Station Keeping, if proper training in Station Keeping is provided and Time Keeping is used. The overall conclusion from this study is that Station Keeping still must be considered to be in a phase of development. This study indicates that there are definite advantages with Station Keeping. One of the most salient advantages is the increased overview of air traffic and the potential benefits for flight safety that not only ATC but also the pilots have access to this. The most appearing issues, that need to be better elaborated before taken to use in operational flight, is how to deal with the potentially increased workload for the pilots, and how to better adjust the display itself as well as the presented information. Nevertheless, it was our impression that all pilots, the instructor and the observers were surprised at how fast pilots learned the method of Station Keeping.

6.3. Operational Analysis The operational analysis is an integration of both the technical and human factors related observations, studied in a setting of standard flight operations. By relating events, deviations and discrepancies observed in the technically logged data, to the human factors observations, some operational factors can be concluded. This analysis is of course rather subjective in its nature. The conclusions drawn from the operational analysis is presented in the next chapter "Operational Conclusion".



7. Operational Conclusions

7.1. Departures According to the procedures defined in the DAS concept (ref. sect. 3.3.2) Station Keeping during departure is limited to after trailer has reached a climb speed of 250 kts+. The departure scenario when following a Standard Instrument Departure SID is a fairly stabile situation from an operational viewpoint. Standard procedures are followed with regard to both lateral navigation and speed control. This means that SID´s gives conditions suitable for Station Keeping using time separation once the standard T/O and cleanup procedure is completed. Using distance separation is however not possible until after both target and trailer have completed acceleration to their final climb speeds. Varying speed performance for different a/c types and grossweights will of course affect the ability to perform Station Keeping with high precision. During the simulator trials only a limited number of departures were performed (six). Of these only three were conducted after the pilots had aquired some SK experience. In spite of the limited number of trials our conclusion that the assumptions made could in principal be confirmed. If the trailer starts its takeoff at the exact assigned time after the target start of takeoff, then SK using time separation is very feasible provided aircraft performance does not differ largely. Only NAV trail mode was tried. Based on the overall experience from TARGET trail during these simulations it must be assumed that TARGET trail will work in a similar manner during departure as was the case for approach.

7.2. Approaches After a few approaches during the simulations it became rather obvious that the most operationally suitable way to excute Station Keeping approaches would be to use NAV trail up to initial approach fix IAF and thereafter TARGET trail until established on the localizer LOC. Also it seemed obvious that time separation would be preferable from the pilots point of view. Since indications from the ATC simulations (ref. sect. 3.2) indicated that time separation would be very useful in an ATM perspective it was decided to focus on approaches using NAV trail and time separation. This resulted of course in rather limited knowledge about other types of trail and distance separation. However with the limited number of trials that was available we found it important to concentrate on what we thought to be the most realistic way of developing procedures for Station Keeping. The conclusion drawn from the 10 approaches executed in NAV trail to IAF then TARGET trail to LOC and time mode using 50, 60, and 70 secods separation is the following: Lateral navigation is no different in NAV mode since it is executed flying autopilot (AP) coupled to the FMS flightplan. When passing IAF heading (HDG) select mode is used. This of course increases the pilot workload considerably especially during turns. Following the target trail by use of the trailer dots on the map display is not complicated but requires quite a lot of attention. Pilots indicated that they preferred to use the trailer dots display rather than the track deviation indicator TDI at the bottom of the CDTI (appendix. D). The difficulties executing these approaches were almost exclusively encountered during the near 180 degree turn from downwind to intercept LOC. The problems can be related to the effect of different turning radius when noticeable speed-differences between trailer and target existed. Also the CDTI information was insufficient in turn since the actual indicated separation before trailer turn is based on a rather approximative algorithm (ref. sect. 3.3.3 + 6.1.3.4). Note: This algorithm should be enhanced with more experience. Last but not least, this part of the approach gives a very high workload when piloting pilot must navigate both vertically laterally and longitudinally in a long turn at the same time as normal cockpit duties are rather frequent.



Except for the situation in turn described above the impression is that pilots could obtain a very high standard of precision in maintaining the assigned time separation, refer to section 6.1 for details. With the above in mind it is recommended to develop FMS STAR´s that go all the way from TMA border to an intercept of the LOC for Station Keeping procedures. This would greatly alleviate pilot workload and Station Keeping approaches would be operationally feasible without system integration with AP/FMS. If TARGET trail is to be used in Station Keeping approaches it is our belief that the lateral navigation would have to be integrated with the AP/FMS through a specific trail mode.

7.3. Time/Distance The operational methodology when trying to maintain an assigned distance to a preceeding aircraft (target) differs quite a bit from trying to maintain an assigned time when the trailer adjusts its speed. This is not conceptually evident, therefore it requires some thought to get used to how to adjust your own speed as a trailer using time instead of distance separation. DISTANCE is quite straightforward to understand. If target increases/decreases its speed, the trailer must immidiately decrease its speed with the same increment/decrement to maintain the assigned distance. TIME is a bit more complicated to comprehend but on the other hand gives a better possibility to maneauver into position. If target increases/decreases it speed. The trailer must also increase/decrease its speed but not with the same increment/decrment. When the trailer speed decreases the time to fly to the target position will actually increase, therefore the trailer must decrease its speed with a less decrement than the target to maintain the assigned time separation. The same phenomena is valid for increase in speed. The trailer must increase its speed with a less increment. The conclusion is that time separation gives a smoother (less) correction in speed to maintain separation and also to aquire an assigned separation when not locked in position. In short the difference could be described as the following: In distance separation it is the target that controls the longitudinal relationship and the trailer must follow slavishly by adjusting its speed to navigate longitudinally in 3D. (compare to a stiff steelrod) Using time separation on the other hand the trailer controls the relationship by adjusting its own speed to navigate longitudinally in 4D in relation to the targets real time position. (compare to a telescopic rod controlled by the trailer). The simulations revealed that this difference between the two modes is quite noticeable for the pilot and it is experienced as much easier to control time than distance with high precision.

7.4. Workload When analysing the findings from the human factors study (ref. sect. 6.2.3.2. + appendix F) regarding the workload situation in cockpit when executing Station Keeping it is obvious that in general there was a marked increase in workload. If Station Keeping is to be implemented in operational procedures it is important that the pilot workload is not noticeably increased. Hence it is important to break down and analyse the reasons for the increased workload. In these trials the increase in workload can be attributed to three main areas: • An added navigational dimension (longitudinal) • Insufficient Human Machine Interface HMI • Novelty of concept (limited training) As pointed out earlier in this report the most important means of mitigating the increased workload attributed to an added navigational dimension is to use time separation and/or to automate the keeping of



assigned longitudinal separation. If longitudinal separation is maintained through pilot control, then it is important that lateral navigation can be managed by automation to keep the workload at an acceptable level. The most important enhancement of the HMI used in this study is to integrate the display in the existing ND/PFD displays. In addition it is of course imperative that extensive HF research and existing knowledge will build a sound foundation for the design of a future Station Keeping interface. It is our strong conviction that a well designed CDTI will greatly alleviate the pilot workload experienced in these simulation trials. The last factor affecting workload, namely novelty of the concept may be the easiest to mitigate but will probably still have quite a large effect when it comes to decreasing the workload. Based on flight instructor experience an educated guess would be that approximately the following training would be sufficient to achieve a comfortable workload level: • 4 hours of theoretical system training. • 4 hours of dedicated Station Keeping training in simulator. • 10 flight applications of Station Keeping during route training. To conclude the operational viewpoint on the workload issue it can be said that: The workload experienced during the trials was too high to be acceptable in an operational environment. However the workload level can most certainly be decreased to a fully acceptable level if the mitigation means described above are implemented intelligently.

7.5. Pilot opinions of concept Most pilots indicated that it was not hard to learn how to fly Station Keeping. One pilot believed that Station Keeping could infringe on safety, the other 5 pilots were of the opinion that safety would be enhanced or not affected. Most pilots expressed the feeling that the concept "felt right". It was expressed as a "natural" method of operation that pilots control separation to preceeding aircraft once they have the tools. Although the number of pilots involved in the trials was rather low we are convinced that implementation of Station Keeping will be well received amongst pilots provided the tools, procedures and training are developed with the pilot in focus.

References 1) Operational Environment Definition (OED) of Delegated Airborne Separation in Approach and Climb-

Out Stockholm-Arlanda ver. 0.8 Michael Agelii, Åke Wall

2) Is limited delegation of separation assurance promising? Hoffman, Zeghal, Cloerec, Grimaud, Nicolaon

3) REPORT ATC Simulation Delegated Airborne Separation during Approach and Climb Out Simulations & Systems Department SATSA Swedish ATS Academy, Luftfartsverket

4) General OED (Operational Environment Definition) ver 1.0 Daniel Ferro, M. Le Berre

5) ICAO SARPS VDL mode 4

Flight Deck Simulations of Station Keeping Appendix A 27


Appendices A. Abbreviations and definition of terms Abbreviations ADS Automatic Dependent Surveillance ADS-B Automatic Dependent Surveillance-Broadcast AI Application Interface ANP Available/Estimated Navigation Performance AOC Air Operator Certificate ASN.1 Abstract Syntax Notation One ASTERIX All purpose Structured EUROCONTROL Radar Information exchange ATM Air Traffic management ATC Air Traffic Control ATIS-B Automatic Terminal Information Service - Broadcast ATS Air Traffic Service ATN Aeronautical Telecommunications Network CAA Civil Aviation Administration CC Cluster Control CDTI Cockpit Display of Traffic Information CNS Communication, Navigation, Surveillance CPDLC Controller Pilot Data Link Communication CRM Crew Resource Management CWP Controller Working Position DAP Download of Aircraft Parameter DAS Delegated Airborne Separation DGAC Direction Générale de l'Aviation Civile DGPS Differential Global Positioning System DME Distance Measuring Equipment EVA Extended Visual Acquisition FAR Federal Aviation Administration FIR Flight Information Region GNE Gross Navigation Error GNSS Global Navigation Satellite System GPS Global Positioning System ICAO International Civil Aviation Organisation IFR Instrument Flight Rules ISO International Standards Organisation JAA Joint Aviation Authority JAR Joint Aviation Requirements LAN Local Area Network LFV Luftfartsverket (SCAA) MET Meteorological NAV Navigation NUP NEAN Update Porgram OAT Operational Air Traffic OED Operational Environment Description ORT Optimising Runway Throughput OSA Operational Safety Assessment RCP Required Communication Performance



(B)RNAV ( ) Area Navigation RNP Required Navigation Performance RSP Required Surveillance Performance SARPS Standards and recommended Practices SAS Scandinavian Airline Systems SCAA Swedish Civil Aviation Administration (LFV) SK Station Keeping SMGCS Surface Movement and Guidance Control System SRG Safety Regulation Group SSR Secondary Surveillance Radar SUA Special User Airspace TMA Terminal Management Area TOD Top Of Descent VDL VHF Digital Link VFR Visual Flight Rules VHF Very High Frequency VOR VHF OMNIDIRECTIONAL RANGE

VORTAC VISUAL OMNI-RANGE TACTICAL AIR NAVIGATION

WGS 84 World Geodetic System 1984 XPDR Transponder



Definition of Terms Accuracy. A measure of the difference between the A/V position reported in the ADS-B message field as compared to the true position. Accuracy is usually defined in statistical terms of either: 1) a mean (bias) and a variation about the mean as defined by the standard deviation (sigma) or a root mean square (rms) value from the mean. The values given in this document are in terms of the two sigma variation from an assumed zero mean error. ADS-Message. An ADS-B message is a packet of formatted data that convey information used in the development of ADS-B reports or generate transmit ADS-B messages. ADS-B Report. An ADS-B report is information provided by ADS-B to an external application. ADS-B Subsystem. The set of avionics or equipment that performs ADS-B functionality in an aircraft or for ground-based, non-aircraft, participants. ADS-B System. A collection of ADS-B subsystems wherein ADS-B messages are broadcast and received by appropriately equipped participant subsystems. Capabilities of participant subsystems will vary based upon class of equipage. Airborne Collision. This occurs when two aircraft that are in flight come into contact. The word «collision» is not an antonym of the word «separation.» A/V Address. The term «address» is used to indicate the information field in an ADS-B message that identifies the A/V that issued the message. The address provides a convenient means by which ADS-B receiving units or end applications can sort messages received from multiple issuing units. Aircraft/Vehicle (A/V). Either 1) a machine or service capable of atmospheric flight, or 2) a vehicle on the airport surface movement area. Alert Zone. In the Free Flight environment, each aircraft will be surrounded by two zones, a protected zone and an alert zone. The alert zone is used to indicate a condition where intervention may be necessary. The size of the alert zone is determined by aircraft speed, performance, and by CNS/ATM capabilities. Call Sign. The term «aircraft call sign» means the radiotelephony call sign assigned to an aircraft for voice communications purposes. (This term is sometimes used interchangeably with «flight identification» or «flight ID»). For general aviation aircraft, the aircraft call sign is normally its national registration number; for airline and commuter aircraft, it is usually comprised of the company name and flight number (and therefore not linked to a particular airframe); and for the military, it usually consists of numbers and code words with special significance for the operation being conducted. Clearance. For convenience, the term "air traffic control clearance" is frequently abbreviated to "clearance" when used to appropriate contexts. Air traffic control clearance : Authorization for an aircraft to proceed under conditions specified by an air traffic control unit. Closest Point of Approach (CPA). The minimum horizontal distance between two aircraft during a close proximity encounter, also know as miss distance. Cockpit Display of Traffic Information (CDTI). A Cockpit Display of Traffic Information (CDTI) is a generic display that provides the flight crew with surveillance information about other aircraft, including their position. Traffic information for a CDTI may be obtained from one or multiple sources (including ADS-B, TCAS, and TIS) and it may be used for a variety of purposes. Any means of communicating the information is acceptable (aural, graphical, head-up, etc.) as long as the information is conveyed effectively. Requirements for CDTI information will vary based on intended use of the data (i.e., application). Collision Avoidance. An unplanned maneuver to avoid a collision.



Conflict. Any situation involving two or more aircraft, or an aircraft and an airspace, or an aircraft and ground terrain, in which the applicable separation minima may be violated. Conflict Avoidance. An unplanned maneuver to avoid a conflict. Conflict Detection. The process of projecting an aircraft’s trajectory to determine whether it is probable that the applicable separation minimum will not be maintained between the aircraft and another aircraft or vehicle. The level of uncertainty in the projection is reduced with increased knowledge about the situation, including aircraft capabilities, flight plan, short term intent information, etc. Conflict Management. Process of detecting and resolving conflicts. Conflict Probe. The flight paths are projected to determine if the minimum required separation will be violated. If the minima are not [projected to be] violated, a brief preventive instruction will be issued to maintain separation. If the projection shows the minimum required separation will be violated, the conflict resolution software suggests an appropriate maneuver. Conflict Resolution. The process of identifying a maneuver or set of maneuvers that, when followed, do not cause a conflict or reduce the likelihood of conflict between an aircraft and either 1, another aircraft or vehicle, 2, a given airspace, or 3, ground terrain. Maneuvers may be given to multiple aircraft to fully resolve a conflict. Conformance. The condition established when the surveillance report of an aircraft’s position at some time «t» (established by the Automated Tracking function) is within the conformance region constructed around that aircraft at its nominal position at time «t», according to the agreed upon trajectory. Cooperative Separation. This concept envisions a transfer of responsibility for aircraft separation from ground based systems to the air-crew of appropriately equipped aircraft, for a specific separation function such as In-trail merging or separation management of close proximity encounters. It is cooperative in the sense that ground-based ATC is involved in the handover process, and in the sense that all involved aircraft must be appropriately equipped, e.g., with RNAV and ADS-B capability, to perform such functions. Cooperative Surveillance Surveillance in which the target assists by cooperatively providing data using on-board equipment. Cross-link. A cross-link is a special purpose data transmission mechanism for exchanging data between two aircraft—a two-way addressed data link. For example, the TCAS II system uses a cross-link with another TCAS II to coordinate resolution advisories that are generated. A cross-link may also be used to exchange other information that is not of a general broadcast nature, such as intent information. Delegated Airborne Separation. Separation assurance between aircraft delegated from ATC to commander. Effective Update Interval. The time interval between successful message receipt with at least 98% probability of successful reception. For example, if ADS-B messages are sent at one second intervals in signal-to-noise conditions with 75% probability of success per transmission, then the probability of obtaining at least one message in three tries is = 1. - (0.25) 3 ~ 98.4%. Thus the effective update interval for this case = 1 sec x 3 = 3 sec. Effective Update Rate. The reciprocal of effective update interval, e.g. rate = 1/3 ~ 0.33 Hz for the example above. Extended Station Keeping. Method of maneuvering to maintain a distance/time or sector relative another aircraft. Free Flight. Free Flight is intended to be a safe and efficient flight operating capability under IFR in which the operators have the freedom to select their path and speed in real time.[TF3, Oct. 1995].



Garble. Garble is either non-synchronous, in which reply pulses are received from a transponder being interrogated by some other source (see FRUIT), or synchronous, in which an overlap of reply pulses occurs when two or more transponders reply to the same interrogation. Geometric height. The minimum altitude above or below a plane tangent to the earth’s ellipsoid as defined by WGS84. Geometric height error. Geometric height error is the error between the true geometric height and the transmitted geometric height. In-Trail Climb. In-trail climb (ITC) procedures enables trailing aircraft to climb to more fuel-efficient or less turbulent altitude. In-Trail Descent. In-trail descent (ITD) procedures enables trailing aircraft to climb to more fuel-efficient or less turbulent altitude. Interactive Participants. An ADS-B network member that is a supplier of information to the local ADS-B subsystem and a user of information output by the subsystem. Interactive participants receive messages and assemble reports specified for the respective equipage class. Latency. The latency of an ADS-B transmission is the time period from the time of applicability of the aircraft/vehicle position ADS-B report until the transmission of that ADS-B report is completed. Latency Compensation. High accuracy applications may correct for system latency introduced position errors using ADS-B time synchronized position and velocity information. Navigation Uncertainty Category (NUC). Uncertainty categories for the state vector navigation variables are characterized by a NUC data set provided in the ADS-B sending system. The NUC includes both position and velocity uncertainties. Near Term. Near-term applications are defined as those that can be supported by an initial ADS-B implementation and that may be operationally feasible within the context of a current ATC system or the ATC systems of the near future. The RTCA Task Force 3 report on Free Flight implementation suggests Initial Operational Capability (IOC) for ADS-B by June 1998. Passing Maneuvers. Procedures whereby pilots use: 1, onboard display of traffic to identify an aircraft they wish to pass; 2, traffic display and weather radar to establish a clear path for the maneuver; and 3, voice communication with controllers to positively identify traffic to be passed, state intentions and report initiation and completion maneuver. Position Uncertainty Category (PUC). The position uncertainty category (PUC) is needed for surveillance applications to determine whether the reported position has an acceptable level of position uncertainty. The category is based on the aircraft’s estimate of position uncertainty (EPU), as defined in 3.1.2 of the RNP MASPS[9] Protected Zone. In the Free Flight environment, each aircraft will be surrounded by two zones, a protected zone and an alert zone. The protected zone must remain clear of other A/Vs or obstructions to assure separation. It can be envisioned as a distance-based cylinder with radius equal to half the horizontal separation minimum and vertical extent equal to +- half the vertical separation minimum. The size of the protected zone is a direct reflection of the position determination accuracy. Received Update Rate. The sustained rate at which periodic ADS-B messages are successfully received, at a specified probability of reception. Resolution. The smallest increment reported in an ADS-B message field. The representation of the least significant bit in an ADS-B message field. Seamless. A «chock-to-chock» continuous and common view of the surveillance situation from the perspective of all users.



Separation. Separation exists between two or more aircraft when their positions and velocities are in accordance with standards and procedures that have been determined to be appropriate for the operations in which the aircraft are engaged. Station-keeping. Station-keeping provides the capability for a pilot to maintain an aircraft’s position relative to the designated aircraft. For example, an aircraft taxiing behind another aircraft can be cleared to follow and maintain separation on a lead aircraft. Station-keeping can be used to maintain a given (or variable) separation. An aircraft that is equipped with an ADS-B receiver could be cleared to follow an FMS or GNSS equipped aircraft on a GNSS/FMS/RNP approach to an airport. An aircraft doing station-keeping would be required to have, as a minimum, some type of CDTI. State Vector. A vector of information concerning the current kinematic state of an aircraft or vehicle; in ADS-B the state vector consists of 3 dimensional position and 3 dimensional velocity where the dimensions are in an orthogonal coordinate system. Tactical Parameters. Tactical information may be used to enhance the performance of designated applications. System designs should be flexible enough to support tactical parameters; however, it is not required to provide the parameters in all implementations. Time of applicability. Time of applicability of an ADS-B report indicates the time at which the reported values were valid. Transmission Rate. The sustained rate at which periodic ADS-B messages are transmitted. Traffic Situation Display (TSD). A TSD is a cockpit device that provides graphical information on proximate traffic as well as having a processing capability that identifies potential conflicts with other traffic or obstacles. The TSD may also have the capability to provide conflict resolutions. Trajectory Change Point (TCP). TCPs provide tactical information specifying space/time points at which the current trajectory of the vehicle will change. Velocity Uncertainty Category (VUC). The velocity uncertainty category (VUC) is needed for surveillance applications to determine whether the reported velocity has an acceptable level of velocity uncertainty.

Flight Deck Simulations of Station Keeping Appendix B 33


B. NUP Tiger Teams Description of NUP operational application development The core element of NUP is operational development teams. These expert teams, consisting of representatives for pilots, ATC, airports, general aviation (GA) and industry, are called TIGER TEAMS. In order to cover different phases of flight and ATM environments, seven TIGER TEAMS have been established throughout Europe, each team has its own specific task. The aggregted development work performed by the teams will form the basis for the analyses of operational applications. Tiger Team Stockholm … This team is led by an airline - SAS. Together with ATC and the Stockholm/Arlanda airport authority the possibility to use ADS-B for improved approach and departure procedures is being investigated. The basic concept is called "station-keeping" which is applied in different modes. The anticipated benefits include some 10-20 per cent increase in departure and arrival capacity. Tiger Team Copenhagen … Not only commercial airline traffic can benefit from ADS-B. The Danish CAA is leading a Tiger Team investigating the use of ADS-B for helicopter traffic in the North Sea. This environment is probably one of the most demanding in commercial aviation. The main drivers are to improve safety and regularity for the helicopter traffic between the mainland and offshore oilrigs in non-radar airspace. Tiger Team Frankfurt … This team is led by LUFTHANSA and is investigating the potential of using improved situation awareness based on ADS-B during the arrival and approach phase. Electronic means (Cockpit Display of Traffic Information, CDTI) are used to acquire preceeding traffic. It is anticipated that airborne visual separation could be maintained when weather conditions are close to minimum - and after further developments below - VMC values. The application is called Enhanced Visual Acquisition or EVA. Tiger Team Paris … The Paris team is looking to the use of ADS-B to improve safety and efficiency of operations on the airport surface, especially during bad weather conditions. The concept is normally referenced to as A-SMGCS (Advanced - Surface Movement Guidance and Control System). The work also includes the role and functionality of airport vehicles. The team is led by DGAC/Sofreavia. Tiger Team Reykjavik … This team is tasked with looking into ADS-B applications in non-radar environment in the North Atlantic airspace. The Icelandic CAA leads the team together with airlines from both Europe and Canada, bringing the experiences from the two continents together. The intention is to design a climbing and crossing application for delegated separation assurance. Tiger Team Maastricht … Taking into account the current situation in European airspace, ADS-B must be considered for implementation even in high-density airspace. In co-operation with the Eurocontrol ADS-Programme and Maastricht UAC, this team will explore the potential benefits of ADS-B in airspace with a well developed SSR/Mode-S infrastructure and relatively advanced level of ATC automation. The main driver is to investigate possibilities to share the workload between ATC and the cockpit. Tiger Team Nice … The Nice team is developing ASAS applications and procedures, based on the use of ADS-B, TIS-B and CDTI, to improve the mix of IFR and VFR traffic (helicopters and General Aviation aircraft). The main objective is improved safety but also a possible increase of capacity. The team is led by DGAC/Sofreavia.

Flight Deck Simulations of Station Keeping Appendix C 34


C. ADS-B Automatic Dependent Surveillance (ADS) enables aircraft to send data, including its position, to ground stations using various links, particularly satellites. A logical extension of this is Broadcast ADS (ADS-B). The principle behind ADS-B is that an aircraft should broadcast, regularly and frequently, an estimate of its position based on data from its navigation system. Since this information is broadcast, it could be received by other aircraft and ground stations and potentially used in a number of different ways. Schematic drawing of ADS-B information flow The NUP technical platform uses D-GNSS (Differential Global Navigation Satellite System) and involves a series of ground-based reference stations, which precisely know their positions. This is used as a reference to improve system accuracy. Air-to-Air applications: Airborne Separations Assurance Systems (ASAS) is a generic term for a number of applications centred around cockpit display of traffic information. These are intended to give the pilot a greater role in maintaining separation using new or existing procedures. Situation awareness, separation assurance, manoeuvring support and oceanic applications are examples of areas in which ASAS can be used. It is generally assumed that these applications will be implemented using some form of ADS-B. Air-to-Ground applications: ADS-B transmissions, which are received by a ground station, can be used to provide surveillance. Airport surface movement, area surveillance, separation monitoring, search and rescue and conformance monitoring are areas in which this could be used.

Flight Deck Simulations of Station Keeping Appendix D 35


D. CDTI – MMI5000

Flight Deck Simulations of Station Keeping Appendix E 36


E. Scenario maps

Departure SID (Dunker 2C) runway 01

Arrival STAR (Trosa 2T) runway 26 Arrival STAR (Trosa 2M) runway 01

Flight Deck Simulations of Station Keeping Appendix F 37


F. Human Factors Evaluation

Psychological Implications of introducing Station Keeping in flight navigation: An explorative study

Leif Rydstedt

University College of Trollhättan/Uddevalla (HTU) Gunnar Fahlgren

SAS Captain (Ret)

Technology behind Station Keeping and how Station Keeping operationally is performed can be found in main part of this report.

The purpose of this study is to evaluate psychological impacts of, according to Wickens (2000), aviation psychology rather than constituting a sub discipline. It applies several areas of psychology with overall aim to increase security and reduce human error in operational flight. Examples of sub disciplines with relevance for flight security are work and organizational psychology, motivational psychology, perceptual and cognitive psychology. When planning this study, the mentioned fields of psychology have been taken into consideration and in the discussion the results are approached from a viewpoint of these different perspectives. The specific objectives of the CRM evaluation during the simulator tests are: – How is pilot workload influenced? – How was the man-machine interface? – Observations about the practicality of different procedures. – How will Station Keeping influence passenger comfort?

METHOD Participants. As only six pilots would be used for the trial, a random selection of pilots from the MD-80 population was out of the question. We had to find pilots who were interested in the project and those who had time available for the simulator sessions. For that reason we distributed a leaflet ”Voluntary Test Pilot Wanted” (Appendix HF:A) to all SAS MD-80 pilots at the Stockholm base and of those who were available we selected six persons for the test. The participating pilots were all males, and had a long experience in the profession, with an average pilot duty close to 20 years (m=19,6; sd=11,5; n=5). One participant also reported seven years of experience as flight controller. The average experience with the MD-80 aircraft was 4,5 years (n=4). Procedure. To those six pilots that volunteered to participate we mailed a written information about Station Keeping and before the simulator flight the objectives as well as the detailed instructions were given during a two hours briefing. After the four hours flight in the MD-80 simulator we had a one and a half-hours debriefing. The questions regarding CRM (Appendix HF:B) were given to the pilots in a stamped envelope together with a book on Human Factors. The book was a gift as a ”thank you” for their homework. All answers arrived within four days after the debriefing.



We also asked the simulator instructor to answer some questions regarding his observations from a CRM point of view. (Appendix HF:C) The questionnaire. The questionnaire consists of 23 questions, mainly targeted at the issues specified by the objectives of the study. Questions with fixed response alternatives were mixed with open questions. All questions with fixed response alternatives were followed up with space for remarks and comments from the participant on the actual topic.

RESULTS “User friendliness” of the new navigation equipment. (Question 1) When requested to estimate the “user friendliness” of the new equipment on a scale with five grades, were 1 meant “easy” and 5 meant “difficult”, two participants each marked the alternatives 2, 3 and 4, respectively (m=3,0; sd=,89). From the open remarks we can conclude that inexperience and unfamiliarity with the new equipment could explain at least some of the reported difficulties. Most of the participants also held the opinion that the test version of the navigation equipment required to be further developed before taken to use for operational flying. Remarks: -Difficult to say. We only used the display information. This test equipment is not to be used for operational flying. -Needs further development. -Hopefully more usable when the system is integrated in pilot’s instrument -Information must be on Nav Display and be larger compared to this test display. -As being the first trial the equipment was good. Improvements and simplifications are possible. -The work became easy as we continuously got support from the instructors. Location of the display, during this test, made the use of it a bit difficult. Intelligibility and sufficiency of the exhibited information. For the question (Q2) “Was it easy or difficult to understand the information given on the display in cockpit?” the same reply scale as above was utilized (1=easy—5=difficult); four participants marked the grade 2, whereas one participant each marked the response alternatives 3 and 4, respectively (m=2,5; sd=,84). Also for the question (Q3) “Did the display give you insufficient or sufficient information”? A five graded response scale – where 1 meant “insufficient” and 5 “sufficient” - was used. Two participants each used the alternatives 2, 3, and 4, respectively, to characterize the sufficiency of the exhibited information (m=3; sd=,89). From the open remarks regarding the quality of the exhibited information, we can conclude that the participating pilots did not consider the used version of the equipment developed to a stage where it can be used in operational flying. Remarks on intelligibility of the information: (Q2) -The Separation Tape and the Relative Speed Vector arrow give contradictory information. -A lot of information is displayed but logic and adjusted to pilots. -At this stage there is too much information on screen. -A logic display, but some information was difficult to translate during large course changes.



Remarks on sufficiency of the information: (Q3) -The dotted Target Trail is hard to see and difficult to follow. Poor information about Target turning speed -Some information should be removed and other information to be presented. -I need Target Heading and not only Vector. -Relative speed vector is misleading when flying Time Trailing. -Sufficient but much spread on screen. - I would like to have the Relative Speed Vector (The arrow) to indicate a ”trend-info” telling me what distance/time-separation will I have in 30/60 seconds with this speed I have at the moment. The information during turn must be improved if merging target from opposite direction or from side. Attention priority. (Q4) In an open question the participants were asked how they diverted their attention between Station Keeping and ordinary pilot duties, during the simulator flight. As can be seen from the remarks, the pilots reported that during this simulator flight, the new navigation display took a great deal of attention. Occasionally it was to such a degree that some of the pilots felt an increased workload. It must again be kept in mind that this was the first test of the new navigation equipment. And the novelty of the test situation makes it plausible to assume that the findings in this study in fact overestimates the amount of attention pilots are forced to give to the display. One of the pilots noticed that due to learning, already during the second flight, less attention was diverted to the display. Remarks -The display took 90% of my attention. -Maximum capacity was diverted to screen during first flights, but as learning improved less attention was needed to the screen. -Station Keeping requires an increased workload. Flying Pilot needs to keep a higher focus on instrument. -This flight had the charm of novelty. With some training and the screen integrated on Nav Display it would be easy and not steel too much attention. -A tendency for target fascination. Will be better when presented on HSI. -Good. The importance of orders, asking for nav aids and FMS work will, as pilot flying (PF), require a greater workload. Communication with ATC. (Q5) No one of the participants perceived any incompatibilities between the display and ATC communication during the simulator flight. “It was an easy communication with ATC like a normal clearance for approach” was the only remark on this issue. The participants were also asked to report (Q6) how communication, on a five graded scale, functioned between pilots in cockpit. (1 meant “without problems” – 5 meant “with rather great problems”: Two of the participants responded with 1, three participants marked 2, whereas one marked 3 (m=1,83; sd=,75) Remarks (Q6) -A reasonable sound level in cockpit is necessary. How does it work with intercom via head set? -A very good, positive and well trained communication (CRM) is needed. This is more important during Station Keeping than for ordinary flying. -A bit unfamiliar as it was the very first time this technology was used. -We had no specific call outs and had to improvise.



When asked (Q7) only one of the six participants reported that communication during any part of the simulator test was difficult to understand or caused misunderstandings. As can be seen from remarks below on this issue, the occurred misunderstanding seems to have been of a minor degree. Three of the six participants reported, on a five grade scale, that communication between cockpit and ATC (Q8) functioned “very good” (5) during the simulator flight, while the remaining three participants all marked response alternative 4 (m=4,0) Remarks on misunderstandings: (Q7) -A standard communication is needed. But the misunderstandings were easily handled. -Yes as we had no specific standard call outs. Remarks communication between cockpit and ATC (Q8) -Very easy like Cleared for approach. -As usual -Clearances easy to understand -With a more developed system communication will probably be reduced a lot. -At some occasion doubtfulness occurred as we, so far, are missing an established phraseology. Time pressure. (Q10) When the participants were asked to report on a five grade scale (1=almost never—5 rather often) if they felt forced making decisions under time pressure, two of them marked 2; one marked 3; two marked 4, while one participant reported 5, that is, he had rather often made decisions under time pressure during the simulation flight (m=3,33; sd=1,21). The only remark, by the participants, points out that the magnitude of this problem will be reduced by increased experience with the technology. As we got used to the equipment, we relaxed and found more time for decision making. Insufficient/confusing information. (Q11) (Q12) Two of the six pilots reported that they during the simulation flight were forced to make decisions based on insufficient data/information, while four of the pilots reported that they were forced to make decisions with vague/confusing information. Again we must bear in mind that what we here report are the results from the first simulation where the new navigation equipment has been used. From the remarks we can conclude that many of the problems regarding the quality of the information can be solved by relatively minor adjustments of the display. Remarks -Sometimes the information was vague. See my answer on question 3. (The dotted Target Trail is hard to see and difficult to follow. Poor information about Target turning speed) -Difficult to determine when Target commenced a turn or changed speed. This should be marked on Target Trail. -The presentation on the screen is now a bit too ruff to give a good overview -Readability was not good as information sometimes obscured each other. -Wrong decision was made due to misleading information on screen. (I think he referred to the Separation Tape,

which was modified for session 2 and 3 as described in the operational report/GF) -Once again, information during great course changes must be improved. Three of the six participating pilots (Q13) experienced that Station Keeping caused disturbances which hampered decision making and/or increased stress. In the open remarks we find that a majority of the participating pilots reported that the navigation equipment increased the workload by adding one more task to pilot work.



Some, but not all, of the participants found although that the increment in available information and improved overview more than compensated for the new work tasks. The impact, of this new navigation equipment, upon the pilot’s mental work load and stress level can not be finally decided exclusively from this limited study. Also here we must take in regard the novelty of the situation and the possibility that further experience, in utilizing the equipment, will reduce the workload. Remarks -Like flying a localizer approach and still have all other pilot duties. -More workload on pilot to maintain separation within specific tolerances, which in congested airspace, would increase stress. -Workload is increased but I do not feel more stressed. -A new task has to be performed, but on the other hand we are spared questions about ATC intentions, radio communication is reduced and we get better situation awareness. -Increased workload due to rudimentary interface but better overview of the situation in airspace. - On the contrary, information is so much better that decision-making is facilitated. In regard to the impact from Station Keeping on the complexity of navigation (Q14) the result from this study also gave a mixed picture. Three of the six participants held the impression that Station Keeping will make navigation easier, whereas the other three participants did not believe this to be the case. In regard to navigation, the foremost mentioned advantage with Station Keeping was the improved overview. Remarks -Good picture of where everyone is. Today’s presentation on screen will possibly cause a lack of situational awareness. -Easy to follow a target, at least when it is flying a planned route. -It is not a question of navigation. After a few test runs, separation –especially Time Keeping– functioned better than it does at Arlanda today. -There was a better overview and easier to foresee what would happen. -Not the way we used it during this trial. Flight Safety. (Q15) When requested to grade the impact of Station Keeping on flight safety on five-graded scale (1=great improvement—5=great deterioration) five of the six participant expected that flight safety will be greatly (n=4) or somewhat (n=1) improved due to Station Keeping. But one of the participant believed that it may cause some deterioration (m=1,67; sd=1,21). The participants expect that the main contribution to flight safety, from Station Keeping, will be brought about by better supervision of the air space, since not only traffic controllers but also the crew members will have direct access to this type of information. Most of the participants claimed, although, that a condition for improved flight safety by Station Keeping, is that the design of the display will be improved and better incorporated with the other systems and that proper training and education in handling the new equipment is given to the pilots. The one pilot that feared deterioration in flight safety argued that Station Keeping might cause an increased workload for the crew. Remarks (Q15) -Situation Awareness and visual look out will be reduced due to increased workload. Maybe HUD will help. -Crew will need good training on system in order to assure incorporation with other systems.



-Yes provided the information is incorporated on Nav Display and the symbols are clarified in order to keep workload on a reasonable level. -Instead of one person supervising airspace, there will be many. -I regard flight safety being very high today, but this system gives possibilities for reduced separation with the same high safety. Most positive and negative features, respectively, with Station Keeping. (Q16) (Q17) When requested to list the most positive features of Station Keeping, most of the pilots mentioned the improved overview and control of the air traffic situation. In association with this, one pilot also mentioned that it might contribute to reduce delays during approach and departure. Also here, however, one of the pilots stressed the need to further develop the technique. As the most negative features of Station Keeping, increased workload and the risk for confusing information were foremost mentioned. As can be seen from the remarks, also in regard to the negative aspects, some of the pilots expressed the belief that the negative factors can be overcome, or at least improved, with technical adjustments of the equipment. The most positive with Station Keeping: -An interesting technique, which might be further developed. -To be able to see other aircraft around. -The fine overview which gave a good control of other traffic. -To have a full control of the situation. -Better overview of traffic situation and possibilities for reduced delays during approach and departure. -The precision and traffic awareness. A cockpit crew is supervising separation to one aircraft ahead instead of one flight controller supervising separation between many aircraft. The most negative with Station Keeping: -Takes too much energy from normal cockpit work. -Could be confusing when changing from one trailing scenario to another. -To keep ones hand up on, or close to, glare shield for e long time. There is a risk to cover or change the settings e.g. during gusty weather. I also got tiered in the arm during the test. -Increased workload but with more developed procedures and methods workload will be reduced to the level we have today. -Missing information during great course changes. Task complexity of the simulation flight. (Q18) The participants were requested to grade the complexity in utilizing Station Keeping during the different scenarios presented during the simulator test flight. As can be seen in Table 1 below, the majority of the pilots perceived most of the scenarios as easy. Only the scenario “Station Keeping long final approach” was perceived as somewhat difficult by three of the six pilots.



Table 1. Perceived task complexity in utilizing Station Keeping under different scenarios during the simulator test flight. (n=6) _______________________________________________________________________________ Very Rather With some Very I easy easy difficulty difficult failed mean (1) (2) (3) (4) (5) _______________________________________________________________________________ -Station Keeping on route 1 3 2 - - 2,2 or initial approach -Station Keeping long final 1 2 3 - - 2,3 approach -Station Keeping during 5 - 1 - - 1,3 climb out -Merging to trail another 2 1 - 1 - 2,0 trailing aircraft (n=4) Difficulty of utilizing Time Keeping and Distance Keeping. (Q19) Information about the distance between trailer and target, (based on actual speed of the involved aircraft), could be presented on the display either in terms of spatial distance or in terms of time separation. When asked to estimate the difficulty, on a five graded scale (1=easy—5=very difficult), in utilizing the two different measures of separation, it was clear that the majority of the participants found Time Keeping (m=2,17 sd=1,17) easier to use than Distant Keeping (m=3,17 sd=1,17). In the open remarks on this topic the participants also emphasized that they had found Time Keeping providing better and easier interpreted information. Remarks -Time Keeping gave me better information. Surprising as we are used to distance. -Time Keeping appeared to be easiest as we got a quicker update. -I prefer to fly ”Time” -Distant Keeping was not flown so much in this test. -If we could get some information about within what time we will reach the assigned distance or time separation, I regard Station Keeping being rather easy to perform. Passenger Comfort. (Q23) In an open question the participating pilots were asked to give their view on Passenger Comfort when flying Station Keeping. As can be concluded from the remarks below, the vast majority of the participants held the opinion that with improved routines, Station Keeping would not give any major disadvantages in passenger comfort. -That is a forgotten factor. Maybe motion sickness tablets will help? -With training in handling, passenger comfort should not be degraded. We can always learn to handle new items smoothly. -Initially passengers will feel engines accelerate and decelerate. As we learn to fly Station Keeping the problem will be reduced. -Time Keeping must be used. Then it is much more pleasant to work when changing speeds. -A lot of power changes, which probably can be eliminated in a more developed system. -It is proportionally to the required accuracy in Station Keeping. The greater the accepted margins the greater Passenger Comfort.



Miscellaneous. (Q9) (Q20) (Q21) (Q22) All six pilots agreed that Captain and First Officer should change and perform Station Keeping alternatively as during ordinary departures and landings -Both need training and must have good knowledge about the system. -On route flying yes. During tests/trials longer periods 1 - 2 hours should be given to each pilot. All the six participants judged the briefing, as well as the debriefing, good or very good. Remarks on this was” Yes, and it also was very good that we received information as home studies”, ”A very constructive debriefing”. ”Very good. Our opinions were received with great interest. Meaningful.” Furthermore, despite that no time compensation was given, five of the six pilots gave positive comments on being offered to participate in the study, e.g.: ”Good to find pilots who really are interested to spend a day without compensation”. ”Perfect, good to give all pilots the opportunity to bid”. ”I really wanted to take part in the test”.

DISCUSSION When interpreting the results from this study, we must of course be aware of its limitations. The small number of participants in the test group and participation based on self-selection reduces the internal validity of the study. It must be clearly stated that the results from this limited study can not be generalized, nor do the authors claim to have identified all possible problematic aspects of Station Keeping. The presented study should rather be considered as explorative and attempting to illuminate possible problems and advantages associated with Station Keeping. Furthermore, in response to initial remarks from the participants, the MMI 5000 display was, to some extent, modified during the course of the study. As this modification obviously improved equipment, the negative responses to some of the question should be interpreted with caution, since actual design was changed between the first and the two following simulator sessions. Finally, we must be aware that this was the first simulator test flight with Station Keeping. Therefore, the outcome of this initial test should not be considered as once and for all established facts about the impacts of Station Keeping. It must be recognized that most of the problematic aspects of Station Keeping, found in the present study, can be improved, or even overcome, by re-design, proper training and education and by more developed routines and increased familiarity with the procedures and the equipment. One of the more critical findings, in this initial simulation flight with Station Keeping, was the increased workload reported by several of the pilots. The well-established relation between arousal and performance, with its characteristic inverted U-shape (Hebb, 1955), predicts that “an effective perception and information processing demands a moderately intensive but also moderately varied extrinsic stimulation” (Sandén, 1990, p.6). Furthermore, it is well established that difficult tasks have a lower optimal level of arousal than easier tasks (e.g. Sandén, 1990). Closely associated with increased workload is time pressure (Karasek and Theorell, 1990). Three of the six participants in the study reported that they had to make decisions under time pressure. Decision making under time pressure always occurs with some frequency in this type of work, and from this study the occurrence of time pressure compared with the ordinary work situation can not be concluded. We must also be aware that time pressure may have a negative impact on quality of the process of decision-making (Prince, Bowers, and Salas, 1999; Svensson and Edland, 1987).



We must be aware of the potential risk in adding yet another task to the already complex and demanding work of the aircrew. It can although be expected that a more developed version of the equipment, combined with proper training and education in handling the system, substantially will reduce the perceived workload reported in this study. Furthermore, if we take into account that the vast majority of pilots expected that Station Keeping would improve flight safety, a possible conclusion is that the improved overview more than compensates for the increment in demands. From a work psychological perspective the increased demands, by Station Keeping, in combination with increased control in the work situation, in terms of the so-called Demand-Control Model (Karasek and Theorell, 1990), might further add an active component of the job. In this type of works characterized by high demands and high control, it is an open question whether further increments of active components in the work content is associated with the benefits normally gained from this development of the work (Karasek and Theorell, 1990; Kristensen 1995; Kasl, 1996). Another critical finding from this study, which definitely calls for further attention is the quality and intelligibility of the information presented on the display. Although some improvements were made already after the first simulator flight, several critical comments were made about vagueness and difficulty in interpreting some of the information. This evaluation, although, gave a lot of information about the design and placement of the display as well as what kind of information the participants wanted. It is our belief that the quality of the presented information can be improved in the future Station Keeping with Time separation was so unanimously regarded the best method that distant keeping can be abandoned in future tests. Standard procedures and standard communication must be arranged before further trials. Flying Station Keeping is like a prolonged semi automatic ILS or VOR approach when the Pilot Flying (PF) is very occupied with the maneuvering of the aircraft and Pilot Not Flying takes care of ATC communication, checklist reading, monitoring and serving the PF. From the remarks of the participants, there are sound reasons to believe that the high Passenger Comfort will remain unaffected by Station Keeping, if proper training in Station Keeping is provided and Time Keeping is used. The overall conclusion from this study is that Station Keeping still must be considered to be in a phase of development. This study indicates that there are definite advantages with Station Keeping. One of the most salient advantages is the increased overview of air traffic and the potential benefits for flight safety that not only ATC but also the pilots have access to this. The most appearing issues, that need to be better elaborated before taken to use in operational flight, is how to deal with the potentially increased workload for the pilots, and how to better adjust the display itself as well as the presented information. Nevertheless, it was our impression that all pilots, the instructor and the observers were surprised at how fast pilots learned the method of Station Keeping.



References Kasl, S. V. (1996). The influence of the work environment on cardiovascular health: A historical, conceptual and methodological perspective. Journal of Occupational Health Psychology, 1, 42-56. Karasek, R., and Theorell, T. (1990). Healthy Work: Stress, Productivity and the Reconstruction of Working Life. New York; NY: Basic Books. Kristensen, T. S. (1995). The demand-control-support model: Methodological demands for future research. Stress Medicine, 11, 17-26. Prince, C., Bowers, C. A., and Salas, E. (1999). Stress and crew performance: Challenges for aeronautical decision making training. In N. Johnston, N. McDonald, N., and R. Fuller, (Eds.): Aviation Psychology in Practice. Ashgate Publishing: Aldershot, UK. Sandén, P-O. (1990). Work in the Control Room: Studies of sociotechnical systems, job satisfaction, mental load and stress reactions. Doctoral dissertation, Department of Psychology, Stockholm University. Svensson, O. and Edland, A. (1987). Changes of preference under time pressure: Choices and judgements. Scandinavian Journal of Psychology, 28, 322-330. Wickens, C. D. (2000). Human Factors in Aviation. Conference presentation given at XXVII International Congress of Psychology. Stockholm, Sweden.

Flight Deck Simulations of Station Keeping Appendix G 47


G. Call for voluntary pilots

Voluntary "testpilots" wanted!

SAS is looking for voluntary MD-80 pilots to participate in trying out new flight concepts. This is "frontier" research conducted within the European NUP-project in SAS MD-80 flight simulators. SAS is involved in a european research project called NUP (NEAN UpdateProgramme). The project aims to develop new operational concepts and technology around the VDL4 datalink. This datalink is based on the principle of selforganised and timedivised broadcast of data. The technique uses GPS time as reference and is originally developed by the swedish inventor Håkan Lans. SAS has participated in the development of the technology for the last 4-5 years. The technology is now mature and has entered the next phase of development, namely application development. The NUP project is focused on developing procedures and avionics that is certifiable and will open up possibilities for more efficient methods of Air Traffic Management, ATM while at the same time enhancing safety. The NUP partners besides SAS are; Lufthansa, Airbus, the swedish, danish, french german CAA´s and Eurocontrol. The project is financed by the EU commission and the project partners together. SAS role in the project is to provide operational expertise with regard to flight procedures from an ATM perspective. As part of the research SAS will conduct simulator tests of completely new operational procedures based on new avionics information. The report from these tests will form part of the knowledge base for development of avionics. To perform the tests some 6-8 MD-80 pilots are needed as "testpilots" to fly the new concept in the MD-80 simulator at SFA. In short, the operational concept focuses on "Station Keeping". This concept is a bit similar to the military "radarkolonn". However it is quite more refined and developed. The aim is to perform self-separation to preceding aircraft during approach and departure with very high precision in distance between aircraft. The tests will be conducted during july-august 2000. Selected pilots will participate in one simulator session each. A session will consist of apprx 2 hours briefing, 4 hours simulator flight tests and 2 hours debriefing. Pilots will be asked to study some background operational concept information (a couple of hours) provided ahead of the test. The debriefing will include evaluation questionnaires and interviews. No reference to individual pilots will be available in the report from the study. Selection of pilots will be performed on an individual basis through agreement between project management and the pilot, considering dates, availability of CDR, FO etc. Since this is a non-commercial research project no compensation is available. However I believe that the participation in development of future flight operational procedures will be very rewarding in itself. Those of you interested to participate in these simulator tests please contact me through the channels below. Showing your interest does not mean you are committed until we have agreed upon a date. You are of course also welcome to contact me for more information. Arlanda, June 18, 2000 Capt. Michael Agelii STOGS

Flight Deck Simulations of Station Keeping Appendix H 48


H. Pilot Questionnaire Questionnaire to be answered by each pilot after the simulator test flight. 1. Now you have tested a new method of navigation when approaching or departing Arlanda airport. You have also trailed another aircraft, which is turning and changing altitude. Was the equipment in cockpit easy or difficult to use/handle? Easy 1 2 3 4 5 Difficult Remarks. 2. Was it easy or difficult to understand the information given on the display ? Easy 1 2 3 4 5 Difficult Remarks 3. Did the display give you insufficient or sufficient information? Insufficient 1 2 3 4 5 Sufficient Remarks 4. You had to divert your attention between Station Keeping and ordinary pilot duties. How did this function? Remarks 5. Did you at any occasion experience an incompatibility between the display and ATC communication? Remarks 6. How did communication function between the two pilots onboard the Trailer? Without problems 1 2 3 4 5 With rather great problems Remarks



7. Was communication, at any occasion, difficult to understand or caused misunderstandings? Yes No Remarks 8. How did communication function between cockpit and ATC? Very bad 1 2 3 4 5 Very good Remarks 9. Should Captain and First Officer change and perform Station Keeping alternatively like take offs and landings? Yes No Remarks 10. Were you forced to make decisions under time pressure? Almost never 1 2 3 4 5 Rather often Remarks 11. Were you forced to make decisions with insufficient data? Yes No Remarks 12 Were you forced to make decisions with vague/confusing information? Yes No Remarks 13. Have you experienced that Station Keeping caused disturbances which hampers decision making and/or increases stress? Yes No Remarks 14. Is your impression that Station Keeping will make navigation easier? Yes No Remarks



15 Mark you opinion how Flight Safety will be influenced by Station Keeping. 1. Great improvement 2. Some improvement 3. No change 4. Some deterioration 5. Great deterioration Remarks 16. What do you consider to be the most positive with Station Keeping? Remarks 17. What do you consider to be the most negative with Station Keeping? Remarks 18. Grade the different scenarios with 1. Very easy 2 Rather easy 3 With some difficulty 4. Very difficult 5 I failed ..... Station Keeping on route or initial approach. ..... Station Keeping long final approach. ..... Station Keeping during climb out. ..... Merging to trail another trailing aircraft 19. How do you grade the difficulty regarding Distant Keeping compared to Time Keeping? Distant Keeping Easy 1 2 3 4 5 Very Difficult. Time Keeping Easy 1 2 3 4 5 Very Difficult Remarks 20. Was briefing OK? Remarks 21. How did you experience the Debriefing? Remarks



22. How did you experience the invitation ”Voluntary Test Pilots Wanted”? Remarks 23. What is your opinion about Passenger Comfort when flying Station Keeping? Remarks Did you occupy Left seat or Right seat? How many years as pilot ..... How many years on aircraft type .....

Flight Deck Simulations of Station Keeping Appendix I 52


I. Instructor Questionnaire

Questions to the simulator flight instructor and his answers. 1. How did you experience cooperation between Captain and First Officer? Answer: Different crews acted differently. Some had very little cooperation and there was a lack in the CRM - LOOP. This caused a One Man Show and a lack in Situation Awareness. Another crew functioned better with very good inputs from Pilot Not Flying. 2. Did you notice any misunderstandings/conflicts between pilots? Answer: Yes now and then, especially in the beginning when all was new for them. But I do not believe that they noticed it themselves. 3. How do you think distribution of work between pilots can be optimized? Answer: Pilot Not Flying must have a holistic view of the situation and give info to Pilot Flying: Info about deviations from assigned Time/Distant, info about distance to runway, own altitude, target speed and so on. He must be aware about which mode Pilot Flying is using on auto pilot and there must be a standard procedure for call outs. 4. Which type of Station Keeping could they perform most easily? Answer: All three flights indicated clearly that Time Keeping caused less problems and was easiest. 5. Which type of Station Keeping appeared to be most difficult? Answer: All three flights indicated clearly that Distant Keeping in combination with an Assigned Sector caused most problems. 6. How do you think Passenger Comfort might be influenced by Station Keeping. Answer: Station Keeping requires power changes which influence both Passenger Comfort, Noise and Fuel economy. This will improve by training, but never be better than it is now. Time Keeping will give best comfort for passengers. 7. Other remark. Answer: Station Keeping should be able to use after some training. The most usable mode is probably Time Keeping to a target flying a standard instrument approach for landing. Then the trailer knows what is going on and surprises will be reduced.

report flight deck simulations of station keeping · document id: sas_nup_wp2_sk_report_1.0...

Documents