INTEGRATING BIRD STRIKE RISK INFORMATION INTOTHE AIRPORT MANAGEMENT SYSTEM

Isabel Metz, Maria Freese, Tobias Pett, Sebastian Schier

German Aerospace Center DLR

Abstract

Birds pose a remarkable threat to aviation safety. Since 1960, bird strikes – collisions betweenbirds and aircraft – have caused at least 30 hull losses of commercial aircraft, 19 of them resulting in atotal of 194 fatalities. Due to increasing population sizes, especially of large birds and a rise in globalair traffic, the danger of bird strike is expected to rise in the future. The risk is largest below 3000 ftand as such in airport environments.Within the airport system, air traffic control has a major role in guiding the air side movements. There-fore, air traffic control is an important partner in the so called airport collaborative decision makingprocess. In negotiations with airlines, airport operations and ground handlers, air traffic control has asignificant influence on decisions about handling of congested traffic situations or disruptive events.Supporting air traffic control with a bird strike advisory system might not only improve the traffic guid-ance, but also planning activities of airport management if air traffic control provides the additionalinformation to the other stakeholders.A current DLR project considers the integration of information about short- and long-term predictionsfor bird activity at and around an airfield to the airport management system. The project’s scopelies on evaluating how the additional information is taken into account for the decision-making of theairport stakeholders. For this purpose, workshops in a serious gaming environment and a Human-in-the-Loop simulation environment were performed. This paper describes the chosen approach as wellas the obtained results.

1 IntroductionCollisions between aircraft and birds are a seri-ous concern for aviation safety. Since 1960, birdstrikes have caused at least 30 hull loss acci-dents, where 19 of them resulted in a total of 194fatalities [1]. The majority of bird strikes occursin low altitudes: Studies from the United Statesof America (USA) and Europe reveal, that be-tween 88% and 95% of all bird strikes happenbelow 3000ft [2], [3], [4]. Hence, aircraft aremost endangered during departure and land-ing. The danger for serious aircraft damage in-creases with the biomass and the number of in-volved birds [2]. Therefore, major concerns arebird strikes involving flocks of large birds. Theresearch for avoiding collisions between birdsand aircraft has mainly been driven by the mili-tary forces. As such, air forces of different coun-tries such as the USA and the Netherlands havedeveloped Bird Avoidance Model (BAM) focus-ing on detecting and modelling bird migrationmovements [4]. The resulting information canbe used for flight planning in order to avoid ar-eas with high densities of (predicted) bird move-ments. In civil aviation, bird movement infor-mation is usually not taken into account in flightplanning [4].

In the beginning of the 21th century, the Eu-ropean Organization for the Safety of Air Nav-igation (EUROCONTROL) introduced a distinctway to perform airport management. The socalled Airport Collaborative Decision Making(A-CDM) allows airport stakeholders to share in-formation and negotiate on available resources.Using A-CDM procedures, events delaying airtraffic can be handled more efficiently [5].

This study evaluated the possibilities of integrat-ing bird strike risk information into A-CDM pro-cedures. The main focus of the study lied onthe question, how well A-CDM stakeholders candevelop strategies for minimizing bird strike riskand up to what extent prediction tools can sup-port this process. For this purpose, two work-shops with experts in the field of flight guidancewere conducted in simulated A-CDM environ-ments with capabilities to integrate bird strikerisk information.

This paper is structured as follows: First, we givean overview on the simulation environments wefocused. This is followed by a description of theapplied method and the gained results. Finallywe discuss the results and give suggestions forfuture work.

Deutscher Luft- und Raumfahrtkongress 2016DocumentID: 420220

1©2016

2 MethodThe DLR Institute of Flight Guidance providesdifferent experimental set-ups for testing A-CDMprocesses. This study focuses on the SeriousGame Decisions based on Collaborative Inter-actions in Teams (D-CITE) and the Human-in-the-Loop simulation environment Airport Con-trol Center Simulator (ACCES). To test, howairport stakeholders take into account informa-tion about bird strike risk, workshops with ex-perts were conducted. In the first workshop, theparticipants played an adapted version D-CITE.In the second workshop, other participants per-formed the tasks of the A-CDM stakeholders inthe ACCES. In both workshops, the participantswere asked to evaluate their own performancewith questionnaires. The results were used toanswer the three main research questions whichwe had formulated for this study:

1. How well can bird strike risk events begenerated in A-CDM simulation environ-ments?

2. How well can the A-CDM stakeholders de-velop strategies to minimize the bird strikerisk?

3. How much do prediction tools improve theperformance of the A-CDM stakeholders inminimizing bird strike risk?

2.1 Participants

In this study, eight experts in the field of flightguidance participated. Four of them played theserious game D-CITE, the other four took partin the simulation within ACCES. We had onefemale participant and seven male participants.The average age was 37.63 years (standard de-viation = 12.05, Range = 20 - 56 years). Inthe D-CITE workshop, they were assigned tothe roles of two airline agents, one airport agentand one ground handler agent. In the ACCES-workshop, the participants were assigned to theroles of two airline agents, one airport agent andone Air Traffic Control (ATC) agent.

2.2 Experimental Setup

The experimental set-up consisted of two work-shops. In the first workshop, an adapted ver-sion of D-CITE, which took into account un-certain events in the form of bird strikes, wasused. In the second workshop, A-CDM scenar-ios including the risk of bird strikes were per-formed in ACCES. Both workshops took approx-imately three hours. In the following chapters, abrief overview of the two experimental set-ups isgiven.

2.2.1 Serious Game – D-CITE

In general, serious games are used for train-ing or education purposes [6]. In contrast tokinds of simulations, the use of serious gamesoffers a number of advantages. It is possibleto represent complex human processes (e.g.,decision-making) in an abstracted way [7]. Seri-ous games are easier to understand as well asmore intuitive than complex simulations [8].

The DLR initially developed the serious gameD-CITE as a research instrument to analysecomplex decision-making processes in airportmanagement [9]. It also has successfully beenused as a training instrument [10]. D-CITE is around-based multi-player game and playable forfour to five players. The provided roles consistof two to three airline agents, one airport agentand one ground-handler agent [11]. The func-tion of an air traffic controller is considered as anautomated game player. A game master moder-ates and coordinates the game. All players havethe goal of optimizing the business of a genericairport, which has one runway and three stands.Therefore, they have to maximize their individualscore as well as a team score. For this purpose,all players have to pay attention to their financialin- and outcomes during playing each round ofD-CITE. The players get a general financial in-come for take-off of each aircraft and a bonusfor a punctual dispatch/ handling time. Further-more, there are some role-specific in- and out-comes. The team score serves as a perfor-mance and collaboration indicator of the teamand the individual money score as a base forindividual economic success. D-CITE’s game-mechanism includes monetary penalties for de-layed or cancelled operations as well as for de-cisions which lead to passenger dissatisfaction.These penalties decrease the financial income.Furthermore, they represent a reduction of theairport’s reputation which can lead to financiallosses in the future, as passengers might decideto fly from another airport.

D-CITE consists of two phases for each round:planning phase and simulation phase. Duringthe planning phase, the players have the task ofcollaboratively tuning an Airport Operation Plan(AOP), consisting of seven time steps and a lim-ited number of flights (cf. Figure 1).

Each column within Figure 1 represents onetime step. The rows indicate possible posi-tions respective phases of the aircraft: approachroutes from A to D, landing phase, stand positionfor aircraft on ground or departure slot. Mark-ers for each aircraft indicate the current posi-tion or state – these can be placed and shiftedduring the planning phase. The more the play-ers share useful information during the planning

Deutscher Luft- und Raumfahrtkongress 2016

2©2016

phase, the better the game works and the teamcan achieve a higher team score. After finish-ing the planning phase of one round, the sim-ulation phase starts with the simulation of eachtime step of this round (cf. Figure 2).

Figure 1: D-CITE overview for the planningphase [11]

Figure 2: D-CITE overview for the simulationphase [11]

In a prior version of D-CITE, critical events –problems, which also happen in reality – weredefined. Examples for critical events are a run-way closure for one time step, required mainte-nance of equipment or a thunderstorm. All thesecritical events result in limitations (e.g., withinone time step it is not possible to land or to take-off), which have to be considered by the play-ers. The critical events are used to incorporatean unpredictable moment in the game setting.This is not only an important game element, butalso leads to new challenges in rescheduling theinitial AOP of all players.

For this study, we performed some changesconcerning the critical events happening duringD-CITE. These are described in the followingchapter.

2.2.2 Serious Game – Workshop

All participants of the serious game-workshophave already played prior versions of D-CITE.Hence, no special training was necessary. After

a short briefing, D-CITE was started. The gameexists in a digital as well as in a paper-basedversion. Both versions have identical game-mechanisms. For this workshop, we decided touse the paper-based version of D-CITE, as it ismore flexible in adding new critical events andrescheduling the AOP. The set-up of the paper-based version can be found in Figure 3.

Figure 3: Set-up of the paper-based version ofD-CITE

During the workshop’s initial briefing, the par-ticipants were informed about the risk of birdstrikes and the consequences of such an oc-currence. Furthermore, they were provided witha pre-planned AOP. The participants had thetask to reschedule the contained flight plan byconsidering critical events. Moreover, they hadthe opportunity to plan a limited number of addi-tional flights per round. Three rounds of D-CITEwere played. The participants were not informedabout the total number of rounds. In each round,a critical event in form of high bird activity with acertain level of uncertainty was presented. Theplayers received the information about this criti-cal event on a so called game card (cf. Figure 4)at the beginning of the planning phase of eachround.

Figure 4: Critical event card

Deutscher Luft- und Raumfahrtkongress 2016

3©2016

As the movements of birds are only predictableup to a certain extent, the critical event wouldalways affect three time-steps – one step hada high probability for a bird strike occurrence,while the adjacent time-steps had a low probabil-ity for a bird strike occurrence. All other steps didnot have any probability for a bird strike. It waspredefined, whether a bird strike actually wouldoccur and if yes, in which of the time-steps. Ta-ble 1 visualizes the critical events presented tothe players during the planning phase of eachround as well as the realization of the criticalevents during the simulation phase.

Table 1: Critical events in form of bird strikesand their realization during the D-CITE

workshop

round critical event(planning phase)

realization ofcritical event(simulationphase)

round 1 step 4: high prob-ability, step 3 and5: low probability

step 3: birdstrike

round 2 step 6: high prob-ability, step 5 and7: low probability

step 6: birdstrike

round 3 step 3: high prob-ability, step 2 and4. low probability

no bird strike

The affected time-steps were chosen in a waythat the participants had to adjust their initialplanning if they wanted to avoid the bird strike.If a bird strike occurs, the consequences de-pend on the aircraft’s flight phase. Table 2 sum-marizes the consequences for bird strike occur-rences in the different flight phases.

Depending on the individual assessment for therisk of bird strike, the players had to developstrategies for risk mitigation. Because of the in-creasing number of aircraft per round, the com-plexity of the game rose.

After finishing the planning phase of one round,the simulation phase for this round starts. In thisphase, the planning for every step is simulatedin a particular order. Therefore, in a first step theplans of the airlines are put into action; secondly,the plans of the airport; thirdly, the plans of theground handler. This is followed by simulatingthe boarding/ passenger movements, the take-offs, payments and aircraft movements on finalapproach. Furthermore, the participants receivethe information, whether any bird strike has hap-pened. The occurrence of a bird strike can haveeffects on the AOP, for example when an air-craft needs to return as an emergency to the air-port and therefore requires the runway for the

next time-step. In this case, a rescheduling forplanned flights might be required.

Table 2: Consequences of bird strikeoccurrences

flightphase

process-oriented conse-quences

financial conse-quences

landing the hit aircraftis removed outof the gameas it has tobe repaired.The subse-quent take-offof this aircraft iscancelled

due to the can-celled take-off,the players re-ceive no moneyfor this aircraft

take-offprior torotation(timestepswithlow riskprobabil-ity)

delayed take-off for the hitaircraft, as ithas to roll backto the stand tobe checked fordamage

monetarypenalty for de-layed take-offfor the airlineof the affectedaircraft

take-offafterrotation(timestepswithhigh riskprobabil-ity)

aircraft has toreturn to theairport due toserious enginedamage inthe followingtime-step. Incase a flightwas scheduledto take off orland in thistime-step, it hasto be delayed togive priority tothe emergencylanding

monetarypenalty for de-layed take-offfor the airlineof the affectedaircraft as wellas any airlinewhich has todelay a take-offin order to givepriority to theemergencylanding

The bird strike events as well as emerging con-sequences result in most of the cases in finan-cial losses for all players. Hence, it is importantthat the participants develop a common strategyfor the risk mitigation.

After each round of D-CITE, the participantswere asked to fill out a questionnaire. After thegame, a debriefing took place. The impact ofthe critical events on the planning activities of allparticipants will be presented in the result sec-tion and discussed later.

2.3 Human-in-the-Loop Simulation

The DLR operates the ACCES, which is viszal-ized in Figure 5. This Human-In-The-Loop sim-

Deutscher Luft- und Raumfahrtkongress 2016

4©2016

ulator is used to design and validate new proce-dures and systems for the airport management[12]. On one hand, the ACCES offers most flex-ible working positions to generate different air-port control centre layouts. On the other hand, itprovides an elaborate software system to sim-ulate the A-CDM process [13]. The softwareconsists of different user interfaces. These en-able the airport stakeholders to interact with thesimulation and plan their flights [14]. The simu-lation core itself includes flight dynamics repre-senting each flight as a sequence of processes(e.g. cruise, initial approach, final approach,turnaround). Due to parameters such as aircrafttype, available stands or runway time, the pro-cess duration for landing, taxi or take-off canchange (cf. [15]). If, for instance, an aircraft’sdesignated gate is blocked by another aircraft,the taxi process of the first aircraft takes un-til the gate is available. The simulation coreis also able to simulate certain events that caninfluence the process duration. Among others,weather events such as a thunderstorm can beintroduced. This is modelled as a delay in theturnaround process within the simulation (theground handling staff is not allowed to work dur-ing this period) or a delay in the final approach(increased separation minima).

The airport management simulation software iscapable of simulating any realistic or virtual air-port as well as real or fictional traffic schedules.As the bird strike workshop was not dedicatedto the effects on a special airport, a virtual air-port with two runways, 15 gates and 10 remotepositions was selected. Figure 6 visualizes thelayout of the chosen airport design which wasalready successfully used in former DLR sim-ulations [16]. The traffic schedule was takenfrom those simulations and relies on realisticdata with changes to the virtual airport. Thereby,traffic volume and flight data meet current trafficconditions without including special effects anddependencies of any real airports.

Figure 5: ACCES

Figure 6: The layout of the virtual airport ZZZZ

The bird strike itself was designed as an un-certain event. Therefore, the simulation soft-ware was configured to randomly choose a fiveminute interval with high bird activity within a to-tal time span of 30 minutes. The total simulationtime lasted 45 minutes per scenario, of whichthe bird strike could occur in the last 30 min-utes. The participants were not informed aboutthe total simulation time. In the case that flightswould arrive or depart within this five-minute in-terval with high bird activity, the first aircraft touse the runway would experience a bird strike.The impact of this bird strike is a five hour delayfor this flight. This is justified by the time which isrequired to check the aircraft for damage and toperform small repairs. If a bird strike occurs, therunway is closed for 20 minutes in order to in-spect the runway. This affects all aircraft whichintend to depart or to land after the bird strike. Inthe beginning of each round, the stakeholderswere provided with a pre-planned flight plan. Incontrast to the D-CITE workshop, adjustmentsof the initial planning took place during the sim-ulation, i.e. while the scenario was already run-ning. As soon as the five-minute interval withhigh bird activity was reached, the participantswere informed and it was checked, whether anyaircraft had been hit by a bird. In case that abird strike had taken place, the participants wereprovided with the opportunity to adjust their plan-ning for subsequent flights before the simulationwas ended. In case that no bird strike had takenplace, the simulation was ended at that point.

To examine how advanced bird strike infor-mation is taken into account by the airportstakeholders, three scenarios were simulated inACCES. The first scenario served for trainingand familiarization with the simulation. As allparticipants had worked with this simulation en-vironment before, this brief introduction was suf-ficient. This first scenario included the closureof four gates and a runway and the stakeholderscould practice how to reschedule flights and findnegotiation mechanisms for the following sce-narios. In the second scenario, the stakehold-

Deutscher Luft- und Raumfahrtkongress 2016

5©2016

ers got informed about increased bird activityfrom minute 15 to minute 45 of the scenario.Thereby, the participants could negotiate on de-laying flights to avoid the phase of bird activityor take the risk of a bird strike. In the third sce-nario, the stakeholders got information about themigrating pattern of the flock of birds which wasinitially predicted to cross the runway betweenminute 15 and minute 45. This information wasprovided in form of a map giving time-stampedinformation about the bird’s current and past po-sitions (cf. Figure 7).

Figure 7: Information about the migrationpattern of birds

2.4 Research Instrument

The main research instrument was a self-developed questionnaire, which the participantsof both workshops had to fill in during and afterthe respective workshops. All questions weredesigned as closed questions. The answeringscale included values from one to four. Thereby,the participants could agree up to a chosen ex-tent within the four levels. This design was usedto get a distinct rather than an open feedbackfrom the participants. Moreover by not provid-ing neutral element (neither satisfied nor dissat-isfied) was used to force the participants to makea distinct decision on the aspect in focus. Theparticipants were asked to fill out the question-naire after each round in D-CITE respective aftereach scenario in the ACCES simulation.

The questionnaire covered three areas. Firstly,the participants were asked about the compre-hensibility of the bird strike risk within the cur-rent scenario. This question was rated on theabove mentioned scale form one to four. Thecomprehensibility was defined as acceptable incase that the average rating to that question isequals or above 3.0 (slight to full affirmation) andadditionally no individual rating is below 2.0 (to-tal rejection).

The second part was specifically designed forthe D-CITE workshop. The target of this work-shop was to observe, whether the participantsconsidered the bird strike risk as relevant and

if yes, how they develop different strategies onmitigating bird strike risk. Therefore, the partic-ipants were asked whether they took any birdstrike risk into account in their decisions. Weassume that at least one stakeholder takes birdstrike risk into account in case that the bird strikerisk is assessed as relevant. Moreover, the par-ticipants were asked, whether they would usethe strategy applied in the current round / strat-egy again. If airport management is able to min-imize the bird strike risk by their decisions it isanticipated that the participants adapt to the riskthroughout the rounds. Thereby we assume ris-ing acceptance of the strategies with increasinground number.

The third part of the questionnaire focused onthe ACCES workshop. In this part, it was anal-ysed, how well tools for predicting bird strike riskover time supported the decision-making pro-cess. In the third scenario of the second work-shop, the participants were provided with addi-tional information about bird strike risk in formof a map with historic time-stamped bird trajec-tories. In the questionnaire, they were askedwhether they felt that this additional informationimproved the predictability of the bird strike risk.

For this purpose, the participants were askedwhether additional information on bird strike risk,which the participants were provided with in thethird round, was perceived as helpful. We as-sume that the predictability in the scenario withthe map is rated higher than in the situationswithout the map.

2.5 Base for Statistical Analysis

Because of the fact, that only eight participants(four per workshop) were tested, the data will bepresented only descriptive (mean score as wellas frequency analysis). To do more complex sta-tistical analysis, it is necessary to test more par-ticipants.

3 ResultsIn both workshops, the participants were askedafter each round, how comprehensible the birdstrike risk was. The possible answers were

1= not comprehensible2= partially comprehensible3= principally comprehensible4=fully comprehensible

Over the three rounds of D-CITE, the partici-pants had a mean comprehension value of 3.25,what lies between principally and fully com-prehensible. The average comprehensibility in

Deutscher Luft- und Raumfahrtkongress 2016

6©2016

the Human-in-the-Loop simulation amounted to3.58. Both values are above the defined thresh-old of 3.0 and an individual rating below 2.0could not be found. The individual values perround and agent are visualised Figure 8.

Within the D-CITE workshop, the participantswere asked if they did take the bird strike risk intoaccount. The replies varied strongly betweenthe stakeholders. At least one stakeholder perround did consider the bird strike risk in his plan(cf. Table 3)

Figure 8: Comprehensibility of the bird strikerisk information for D-CITE (top) and the

ACCES simulation (bottom)

Table 3: Strategy Adaption within D-CITE (yes= I will plan flights in a way that they do not takeplace in any slot with bird strike risk / no = I will

still plan flights in slots with bird strike risk)agent round 1 round 2 round 3airline 1 No No Noairline 2 No Yes Yesairport Yes Yes Yesgroundhandler

No Yes No

The agents were furthermore questioned,whether they would pursue the same strategyin future rounds. Their answers per round canbe found in Figure 9. In average the strategyof round one was rated with 2.5 (fundamentalto slight adaptations) while round two and threewere rated with 3.5 (slight adaptations to noadaptations).

Figure 9: Willingness of the stakeholders torepeat their decisions how to take bird strike

risk into account

Figure 10: Predictability of the bird strike risk(round 1 and 2 without, round 3 with tool)

In the Human-in-the-Loop simulation, tworounds were being performed without any pre-diction tool. In the third round, the stakehold-ers were supported by information about thebirds. This should help them to estimate moreprecisely, when the danger of bird strike waslargest. Figure 10 shows the subjective impres-sion of the predictability per simulation round. Inboth rounds without support, they assessed thepredictability with 1.75 on average, in the lastround, where additional support was provided,the predictability increased to 2.25 on average.

4 DiscussionThis paper posed the following research ques-tions on considering bird strike risk withinA-CDM processes:

1. How well can bird strike risk events begenerated in A-CDM simulation environ-ments?

2. How well can the A-CDM stakeholders de-velop strategies to minimize the bird strikerisk?

3. How much do prediction tools improve theperformance of the A-CDM stakeholders in

Deutscher Luft- und Raumfahrtkongress 2016

7©2016

minimizing bird strike risk?

To answer the first research question, we eval-uated how well the information about the birdstrike risk and occurrence was understood bythe participants in both workshops. The re-sults section showed that the criteria (slight tofull comprehension in average and no individualtotal rejection) were fulfilled. Hence, it can bestated that the distinct way of modelling the birdstrike event is a successful approach for the ex-amined participants. Due to the small numberof different events (two workshops) and partici-pants (eight) the results must be limited to thiscase study. A general answer for all airport ex-perts and all ways of modelling the bird strikecannot be given based on those results.

To answer the second research question, it wasanalysed whether and if yes, in which way theparticipants took bird strike risk into accountwhen making strategic decisions. The resultsshowed that in each round of the D-CITE work-shop at least one participant considered the risk.Moreover, the strategy of round one was ratedwith more need for adaptation than the strate-gies of the later rounds. Confirming these prede-fined criteria, it can be stated that the examinedstakeholders considered the bird strike risk andagreed on a strategy. Regarding this finding, itmust be stated again that the number of partici-pants is too small to generally conclude that birdstrike risk can be minimized by airport manage-ment. Moreover, a so called expectation effectmight have influenced the results. Simulationsas well as gaming are often expected to chal-lenge the participants as these approaches arevery broadly known for their advantage in train-ing risky or uncommon situations [17]. The par-ticipants were briefed about the general prob-lem of bird strike previous to both workshops.Hence, they might have the expectation that theywould be challenged by a bird strike occurrence.For this reason, they might have reacted morecautiously than operational staff.

The second workshop – the Human-in-the-Loopsimulation in the ACCES – focussed on thepotential of supporting the airport stakeholderswith additional information and tools in order toanswer the third research question. Therefore,the participants conducted two rounds withoutany supporting information about bird strike risk.In the third round, they were provided with a birdactivity map as an additional information source.We assumed that the participants would ratetheir prediction abilities higher when being sup-ported by the map. This assumption was con-firmed by the obtained results. Nevertheless,the differences between the condition with andwithout the map need to be analysed carefully.

On one hand only one run with the map wasconducted. On the other hand, the measureddifferences are rather small (0.5 on the 4 pointscale).

We conclude that it is possible to generate birdstrike risk events in airport management valida-tion environments in an understandable way forthe participants. The obtained results from theD-CITE workshop indicate that there is a learn-ing effect in developing strategies to minimizethe risk of bird strike. Moreover we analysed thatcertain additional information does support thestakeholders in choosing an appropriated strat-egy. Nevertheless, the small size of the studyonly allows observing local effects that need tobe proved with a greater approach. In the endwe can summarize that there is potential to con-duct future research activities in airport manage-ment to reduce the bird strike risk.

5 Outlook and Future WorkFor future studies on the subject of integratingbird strike risk information into the A-CDM pro-cess, we suggest to include a larger number ofparticipants to gain a broader statistical base.For the D-CITE workshop, we propose to gener-ate bird strike risks randomly in order to generateeven more uncertainty for the participants. An-other study could focus on differences in strate-gies if an aircraft is hit or no aircraft is hit in thefirst round. Considering the ACCES workshopprediction tools with different qualities could beimplemented in order to test the trust of the air-port stakeholder in this kind of technical support.In this study, we focused on the strategies of theindividual stakeholders. A follow-up study couldinclude metrics for gaining information about theteam strategy of the A-CDM stakeholders andinvolve operational staff.

References[1] J. Thorpe, “Update to’100 Years of Fatal-

ities and Destroyed Civil Aircraft due toBird Strikes’,” in 31th Meeting of the WorldBird Strike Association, (Atlanta, GA, USA),2014.

[2] Atkins Ltd. and Food and Environ-ment Research Agency (FERA), “BirdStrike Damage & Windshield BirdStrike,” 2009. http://www.easa.

europa.eu/rulemaking/docs/research/

FinalreportBirdStrikeStudy.pdf. On-line: accessed 14 April 2016.

[3] I. Maragakis, “Bird population trends andtheir impact on Aviation safety 1999-2008,”

Deutscher Luft- und Raumfahrtkongress 2016

8©2016

2009. Safety Report, Safety Analysis andResearch Department.

[4] J. McKee, P. Shaw, A. Dekker, andK. Patrick, “Approaches to Wildlife Manage-ment in Aviation,” in Problematic Wildlife.A Cross-Disciplinary Approach, ch. 22,pp. 465–488, 2016.

[5] Huet, Dennis and Booth, David and Pickup,Simon, “A-CDM Impact Assessment – FinalReport.” EUROCONTROL, March 2016.

[6] C. Abt, Serious Games: The Art and Sci-ence of Games that Simulate Life. NewYork, NY, USA: Viking Press, 1970.

[7] W. Kriz, “Creating Effective Learning En-vironments and Learning Organizationsthrough Gaming Simulation Design,” Sim-ulation & Gaming, vol. 4, no. 34, pp. 495–511, 2003.

[8] Q. Zhou, The Princess in the Castle: Chal-lenging Serious Game Play for IntegratedPolicy Analysis and Planning. PhD thesis,Delft University of Technology, Enschede,the Netherlands, 2003.

[9] M. Freese, S. Drees, and M. Meinecke,“Between Game and Reality: Using Seri-ous Games to Analyze Complex InteractionProcesses in Air Traffic Management,” inSimulation and Gaming in the Network So-ciety, Springer, 2015.

[10] S. Schier, M. Freese, and T. Mühlhausen,“Serious Gaming in Airport Management:Transformation from a Validation Tool toa Learning Environment,” in Gaming andLearning Alliance Conference, (Utrecht, theNetherlands), 2016. Submitted.

[11] M. Freese, “Game-Based Learning - an Ap-proach for Improving Collaborative Airport

Management,” in 10th European Confer-ence on Games Based Learning, (Paisley,Scotland), 2016. Accepted.

[12] R. Suikat, S. Kaltenhäuser, B. Weber,J. Hampe, and F. Timmermann, “ACCES- A Gaming and Simulation Platform forAdvanced Airport Operations Concepts,” inAIAA Modelling and Simulation Technolo-gies Conference, (Toronto, ON, Canada),2010.

[13] F. Piekert, S. Schier, A. Marsden,N. Carstengerdes, and R. Suikat, “AHigh-Fidelity Artificial Airport Environmentfor SESAR APOC Validation Experiments,”in Air Transport Research Society WorldConference, (Singapore), 2015.

[14] S. Schier, T. Pett, O. Mohr, and S. J. Yeo,“Design and Evaluation of User Interfacesfor an Airport Management Simulation,” inAIAA Modelling and Simulation Technolo-gies Conference, (Washington D.C. USA),2016.

[15] S. Schier, F. Timmermann, and T. Pett,“Airport Management in the Box – AHuman-in-the-Loop Simulation for A-CDMand Airport Management,” in GermanAerospace Congress, (Braunschweig, Ger-many), 2016. Submitted.

[16] A. Papenfuss, N. Carstengerdes, andY. Günther, “Konzept zur Kooperationin Flughafen-Leitständen,” in GermanAerospace Congress, (Rostock, Germany),2016.

[17] T. Anand, P. Derby, and H. Thiruvengada,“Training for Metacognition in SimulatedEnvironments,” in Human-in-the-Loop Sim-ulations, pp. 117–130, 2011.

Deutscher Luft- und Raumfahrtkongress 2016

9©2016