Gate Allocation System

GateAllocation System

InstitutionAffiliated:

Dateof submission:

Contents

Executive Summary 3

Abstract 4

Introduction 5

Objectives 11

Features of a Reliable Gate Allocation Model 12

Examples of Existing Gate Allocation Models 15

Developing the Model 24

GAMS Implementation 30

Conclusion 33

References 34

Executive Summary

Theresearch aims at establishing a model that can be used for thereduction of runway occupancy time. It is an essential research inthat it enables airports to implement such a system in improving onhow they manage and utilize the available resources. The model usedconsiders Singapore’s Changi Airport in coming up with a solutionto managing airport resources. In so doing, a lot of money is saved,and airports can also improve their customer service. The customerservice is seen in the reduced lengths of time that customers will berequired to wait in case their flights delay. It also entails thereduced time that passengers will have to wait in the ramp becauseanother aircraft occupies their gate. Algebraic modeling is used inthe breakdown of the research, after which the collected equationsand outcomes are modeled using GAMS software to achieve comprehensiveresults. Mathematical analysis is a key element in formulatingalgorithms that are to be used in this model. Their calculation andformulation give a clear view of the complexity of an airport and itsmany activities.

GateAllocation System

Abstract

Theresearch aims at reducing the runway occupancy time in moderninternational airports. It considers the ever-increasing number offlights as the airports grow, and the need to manage resourcescarefully. It also takes into consideration the hundreds of flightsthat land or leave an airport through shared runways each day. Thereis a need to control the usage of runways and to keep up with a day’sschedule despite the flight delay and emergency landings that mighttake place. The research is essential in ensuring that the futuregrowth of airports is not limited by the resources they have. It alsoensures that the number of flights can be increased according to theplanned expansion of the airport and its management. It increasednumber of flights should be accommodated within the day’soperations and should function without any hitches. At the end of itall, a model should be formulated and analyzed in such a way that itwill optimize the operations of the airport’s gates. The number oftowed aircrafts should be reduced to the minimum possible number.Cases of passengers having to wait on the ramp because anotheraircraft has occupied their gate should also be reduced. The economicbenefit of this research should be the realization of increasedefficiency of gates and runways. The number of delayed flights due tooccupied runways and gates should also be reduced, boosting customerconfidence. In the end, there should be increased profits generatedfrom operations of the airport due to the easy flow of aircrafts toand from it. The designed model should also utilize resources whollyand efficiently, curbing any wastage.

Introduction

Recently,there has been a rapid growth in air transport traffic. It is becauseof the increased demand for air traffic in transportation. Airtransport is fast and efficient and, therefore, necessary for peoplewith tight time schedules. Airplanes are used by business people togo to business tours or meetings. People who wish to travel todistant places or overseas also prefer air travel to sea travel dueto the speed of travel and the comfort and convenience. Air travel isalso safe in terms of collisions since each aircraft is guidedthrough a clear route, unlike vehicles that can easily collide.Moreover, with the increase in demand for quick transportation ofperishable raw materials, there has been a great increase in thenumber of cargo planes flying the air each day. Air travel has alsobecome affordable to many travelers, unlike in the past when only therich could afford it. Having noted these major uses and advantages ofair travel, the number of aircrafts jetting in and out ofinternational airports is bound to increase.

Airportsare designed to operate like systems. They have inputs, which areincoming flights, outputs, which are the outgoing flights, andprocesses, which are all the activities that take place on theground. These characteristics of airports make them very busy areas,with a lot of activity in terms of people and aircrafts. There arepeople coming in to take their flights and others arriving from otherareas via aircrafts. All these activities have to be coordinated insuch a way that there is no confusion. Everything has to be plannedper day to ensure a smooth flow of an operational day. Any mishapwould lead to a snarl, which would then create a gridlock due to thedependency of some activities on others.

Thesedependencies can be seen in examples such as: the runway has to beclear for airplanes to take off or land passengers have to wait fortheir plane to be ready for them to board and leave a plane has towait for its scheduled time for it to take off and a gate has to befree for an aircraft to use. Any mismatch of one of the mutualactivities would lead to interlocking of other activities, and aneventual snarl-up that would slow down business. The confusion can beso big that the airport might have to go through a small operationalbreak to work on reversing the confusion.

Togain control of the ever-growing airport activity, gate allocationhas to be looked into. Gate allocation is the scheduling of resourcesin terms of airport gates, to ensure that flights are well assignedto gates. Gate allocation models are used to accomplish this task,and are designed to suit a specific airport. They look for the dailyschedule and activity of the airport and its aircraft. These modelsis designed in such a way that they work on the arrival and departuretimes of aircrafts to determine when and where to allocate anaircraft and the right gates to assign them. They also ensure thatthe runways are always free for aircraft to land and take off whentheir scheduled time arrives.

Afailed model could be disastrous in the working of an airport. Designparameters have to be considered and analyzed with utmost care andprecision. The end product should be a smooth system that facilitatessmooth movement and flow of activities and passengers through theairport. It is in consideration that airports are just terminal thatconnect people to the areas they desire to go to and should,therefore, not involve long stays of either passengers or aircraft.

Agood gate allocation model should consider the fact that everythingcannot go as scheduled. It should take into consideration the delaysand changed flights that are common in the day to day life of airtraffic. There are a number of reasons why delays can occur. Theseinclude bad weather which usually hinders visibility and is thereforea hazard to pilots. Another reason for delayed flights is volcanicactivity, although this is usually a rare occurrence. It has howeverbeen recently observed in some Middle East countries. The volcanicactivity causes volcanic ash to fill the air, again causingvisibility issues. Flights can be grounded for days until theatmosphere is clear. The grounding of aircrafts causes a lot ofconfusion, although most modern airports usually have a plan toreduce the impact of such an occurrence. The delayed flights can alsobe due to the congestion of the runway. It is the case where someplanes have landed and still stay on the ramp because others arestill occupying their gates. Moreover, some planes are scheduled toland, and they have to be allowed to land if the runway is shared,resulting in delays of those wishing to take off.

Changedflights can be due to mechanical problems facing some airplanes, suchthat passengers have to be given another plane. These changes caneven occur at the point where passengers had already boarded theplane assigned to them, only to be notified of last minute changes.They can get transferred to another aircraft or asked to wait forsome time as theirs gets fixed. The flights can also be changed whenthe current one is big and with many empty seats whereas there is asmaller aircraft in which these passengers can fit.

Moderninternational airports have hundreds of flights going through themper day. With such a huge traffic, there are bound to be thousands ofpassengers using these same airports per day. The main aim of anairport is to ensure that they serve their customers well andefficiently so that the passengers can use their services again.Failure to please passenger’s results to loss of revenue as thereis expected a reduction in the flow of people through the airport.

Customersatisfaction in an airport means that the passengers should spend theleast time in the terminals. They should check in and get to theirflights in the least possible time. It does not mean, however, thatthey should be rushed through the process of verification andpresentation of tickets. A reasonable amount of haste should beapplied such that by the end of the processing, they should not betired of queuing for too long, neither should they be tired of beingrushed around and probably leaving behind some luggage due to thehurry with which they board their flights.

Havingtaken all these factors into consideration, a passenger should notfinish checking in only to be told that their flight will be delayed,or that they will have to change flights. Some are even redirected toother airports due to unavoidable circumstances. Such occurrences, ifnot short-lived, leave the airport no choice but to book hotels forthe delayed passengers before the next available flights ready. Theaccommodation is however not the best solution, especially when itcomes to people who need to attend to important matters that cannotwait. Such include urgent business meetings and hospitalappointments.

Theincreasing demand for airport services can prompt airport managementto expand the physical facility. The expansion can involve theincreasing of runways, hangars, gates, terminals, and aircrafts. Itcan work in decongesting the airport and its terminals. It will alsogive more room to incoming and outgoing flights. The gates will beavailable for more flights to use.

However,this physical expansion of the facility will only have short-termresults. In the long-term, the number of flights will have increasedsuch that the same problem of congestion and confusion will replayitself. Very many passengers will have been attracted to this airportdue to its open facilities and big terminals. However, the increasingnumber of flights will result in an increased number of passengerstoo. The result of this is congestion in all areas of the airport.With a poor schedule, there would be dozens of delayed flights perday. As a result of this confusion, passengers would start gettingfrustrated by the services in this airport. Aircrafts will also startgetting locked in unending waits and delays, and their airlines canopt for other available and easily accessible airports. The endresult is a failed airport business, where clients leave to findbetter alternatives. Recovering from such a failure has resulted intosome of the biggest failed airports giving up to this industry andclosing shop.

Ratherthan working on the physical expansion of the airport, it would beavailable to adapt a system that monitors and controls activities ofthe airport from all angles. It can best be achieved by using a GateAllocation System. Such a system would be reliable in ensuring thatthe activities of the airport run smoothly and that passengers arenot dissatisfied with the services offered.

Goodairport management involves the proper use of available resources.The main resources include time and space. Time in this case refersto the availability of slots that can support rescheduling offlights, counter the delay caused by aircrafts undergoingmaintenance, and still allow for emergency landings withoutinterfering with the normal day’s schedule. Scheduling in airportsis a very key parameter of success. It is done such that it coverseach day, specifically outlining activities of that particular dayand allocating time to them. The scheduling also involves planningfor disruptions. Adjustments must be possible in the system such thatwhen something goes wrong or requires more time than was earlierplanned for, it can automatically be accommodated. The accommodationmight go unnoticed by many airport users but doing it wrongly cancost the airport in terms of revenue earned that day.

Themodern airports have the advantage of existing in the computer era.They are usually supported by special systems created for them. Suchsystems are the brain and nerves of the airport. They pick up anydiscrepancies, such as delays or missed flights as entered in a frontoffice computer, then decide on how to deal with this occurrence.Actions to be taken are usually pre-set such that the systemautomatically adjusts itself. Occurrence of major disruptions cantrigger the systems to adjust their schedule and ground each aircraftto avoid disaster or accidents on the runway. These adjustmentsshould, however, be monitored on a full-time basis. It is normal forthem to make erroneous claims of the aircraft collision situation andthe availability of gates. The gates can also be indicated as secureand yet they are occupied.

Objectives

Theresearch aims at developing and implementing a Gate Allocation Modelthat can be applied for an airport with two parallel runways, such asSingapore’s Changi Airport. The airport is international andsupports hundreds of flights per day. The runways have to be made asefficient as possible. It means that optimization of the airport’sfacilities and resources should be done excellently. There are threespecific objectives of this research. They are as listed below:

  1. To develop a gate allocation model that improves the runway throughput

  2. To implement this model in GAMS software

  3. To test this model for real instances.

Withthese objectives, it is easier to carry out the research since theywill guide the researcher on the kind of data to collect and how towork on it. Specifying an international airport creates a model thatcan be shared among various big airports and still be scaled down forsmaller airports. It is easier to convert a huge airport model to asmaller version that can fit into the systems of smaller but busierairports and airstrips. There are some airstrips that are usuallyquite busy, such as those used in agricultural plantations to carryperishable goods. These goods that call for the setting up of theairstrip are usually highly valuable and in very high demand. Suchinclude perishable flowers that are normally harvested andtransported immediately for processing. Any delays in the gateallocation model should have the option of prioritizing such flights.Other flights that can be prioritized include medical ones or thoseof flying doctors. Private business flights are also given prioritywhen there are other passenger flights colliding with them in termsof time allocation.

Features of aReliable Gate Allocation Model

Areliable gate allocation model incorporates three main features amongother additional ones. The first of the three main features is theability to accommodate the proposed flights. Each of the availablegate allocation models are designed to accommodate a specific numberof flights. These flights are such that their movement in and out ofthe airport is considered to its maximum. The maximum number that theairport can accommodate is the number of aircrafts that can take offand land, get serviced and use the gates at the airport in one day.The number looks into the maximum utilization of resources, such thatthey will be stretched to their optimum limits. The use of theseresources should, however, be within the available limits too. Itwould be wrong to design a model that does not consider the use ofsome resources, or one that over-uses the resources available. Theresources and their usage should be of the same magnitude.

Toaccommodate the proposed flights, the inputs are normally taken asthe incoming flights. They are supposed to use a parallel runway.These incoming flights are usually taken through gates on landing.The gates should always be available within the shortest timepossible. It would be a failed model if it kept passengers on theramp for long just because the gates were fully occupied. It shouldalso reduce the number of towing activities that take place. Theaircrafts using the gates should also be of a reasonable number suchthat they do not overwhelm the facilities at the airport and causetheir depletion. The airport should be constructed with enough gatesthat can be modeled by a system for the efficient working of anairport.

Thesecond characteristic of a good gate allocation model is thepreparation of daily plans before the exact day of operation. Itmeans that the working schedule of the airport should be prepared inadvance. Flights and their arrival or departure times should be knownwell in advance. The activities taking part throughout the airportthat affect its operations should also be outlined before the exactday in which they are going to be applied. It gives the gateallocation model a basis of understanding the expected day ofoperations and comparing it to the real one. The comparison enablesthe model to identify any deviation from the set standards and whatmeasures it should put in place to achieve optimization.

Flightoperators should also be aware of the daily airport operations on anearlier date than the exact date of execution. It is a keyrequirement for ensuring the smooth flow of activities in theairport. Those who work in teams should also ensure that theyconsolidate their day’s activities and go through the activities ofthe day of operation beforehand to avoid giving conflictinginstructions on the movements of an aircraft on the ground. Therewould be confusion if each operator failed to go through theactivities list and the aircraft schedule for a particular day. Anaircraft that they had not anticipated for could decide to land atthe airport and if they had not gone through the day’s schedule andtime plan, then they would not notice that the flight landed at thewrong airport.

Thedaily plans are usually used by the gate allocation model to identifywhich planes will land that specific day and which ones will depart.The model can then plan the allocation for the runway withoutcreating any confusion.

Finally,a good and reliable gate allocation model should have a schedule thatis updated throughout the day of operation. The updating is necessaryto accommodate the disruptions that are common in the aircrafttransport industry. Such disruptions include delayed flights,probably due to bad weather, volcanic activity, natural disasters andtechnical issue with planes. Other disruptions can be the need toaccommodate an emergency landing. Emergency landings are common, butnot frequent. They involve a distressed aircraft deciding to land onthe nearest runway along their path. These emergency landings arejust random happenings. They are usually unpredictable and can alsobe a risk to the airport itself. A good gate allocation model has theintelligence that not all flights fly within the required range oftime. Some flights have to be delayed due to unavoidable reasonswhile others have to be rescheduled. Others have to be accommodatedwithout having been planned for. However, this is the work of anormal airport, and the gate allocation model should work towardssupporting these events the best way it can without depleting theavailable resources or disrupting the general business day.

Theimportance of having a gate allocation model is that it saves time intrying to create space for new aircrafts and flights. It alsoprovides room for organizational neatness. A well-planned airport hasa neat of arrangement of aircrafts, and each arrival knows where topark. It also allows the airport management to track activities ofthe airport and manage their resources, keeping in mind that they arelimited. In the end, the airports save a lot of money and attractmore investors and airlines. It means that they get more customersand are always on the limelight in terms of performance. With a goodand efficient system, there is no worry about financial management.It comes as an additional package that creates revenue for theairport.

Examples of ExistingGate Allocation Models

Oneexample of an efficient gate allocation system is the AirportSolutions. It was created and supported by Preston Aviation SolutionsPty Ltd. The model was adopted by the Kuala Lumpur InternationalAirport to manage the forecasted growth of this airport. As earliernoted, it is wise to counter growth by the use of a good model ratherthan acquire land for expansion. The proper use of a modeling systemhelps airports utilize their resources appropriately. They end uprealizing potential in areas of development where they once assumedthey could not grow.

TheAirport Solutions model was aimed at supporting the future expansionof this international airport. There were plans to increase the dailynumber of flights and passengers going through the airport. Withterminals and runways remaining the same, there was urgent need tocome up with a way of containing and comfortably accommodating thenew arrivals. The model was the key to a reliable resource allocationand utilization. The system was a complex one due to the constraintsthat it had to take in and the algorithms it had to use to optimizethe airport’s physical facilities without much physical expansion.

Thesystem was supposed to meet the day-of-operations requirements, andthis was done through strategic planning of its design and launch. Itwas to look at the gate resources and their allocation, ensuring thatthere was no overcrowding or delays due to occupied gates. It wasalso supposed to maintain the parking resources, distributing themaccording to times scheduled for each aircraft, and re-assigning themwhen necessary. By managing the parking resources, it was alsosupposed to be up to date on the day’s schedule just in case anexpected flight was delayed, and its resources had to be re-assigned.As part of customer service, the system was to look into baggagecarousels and check-in desks. It was in a bid to ensure that allpassengers checked in within the required time frame and that none ofthem exchanged or misplaced their luggage. The same system wasadopted in several international airports, including the AucklandInternational Airport.

Anotherexample is the implementation of the Airport Capacity Enhancement byBrisbane Airport. It will be done with the help of Airservices, amongother partners. The enhancement program is meant to study and analyzethe Brisbane Airport runways, and come up with recommendations andsolutions for the reduction of runway occupancy time. Brisbane hasgrown into an international city, and this has increased its demandfor air traffic space and time. The growth of Brisbane has had adirect effect on the Brisbane Airport, such that there are plans tobuild a new parallel runway by the year 2020. The growth is in termsof number of daily flights, which can be summed up to about 650 and700 per day and growing. With such a number of daily flights, thereis a need to come up with a system that ensures reduced time wastageon the runway. Even the slightest delay at the airport, be it interms of seconds, can cause diverse effects on the operations of anairport. Brisbane Airport gets a growth of approximately 4.6 percentper financial year in terms of customer numbers, with 910,000additional passengers in the 2011-12 financial year. It means that bythe year 2020, they will be beyond the million passengers mark, andif there were no proper management system in place, the airport mightbe overwhelmed. It could lead to operational failures and theeventual lack of customer trust, which implies that the business willgo down.

BrisbaneAirport reported increased movements, especially in the mornings andevenings. These busy hours experienced delays of 20 to 40 minutes andwere due to the time lost while clearing the runway.

Withtheir current 01/19 runway, it is important for Brisbane Airport tomanage their airport appropriately and have the least possible runwayoccupancy time, especially during the peak hours. However, currently,the management of the runway and its availability highly depends onthe pilots, flight crew, and air traffic controllers. Their abilityto work as a team and coordinate their activities appropriatelyenables the airport to reduce the runway occupancy time and increasethe number of flights that can jet in and out with the day tooperations. It is a clear indicator that even the best airport gateand runway management systems cannot succeed if their implementersare failing in their areas of responsibility.

AirportCapacity Enhancement identifies the capacity to acquire moreresources in terms of time and its usage, and this is usuallyachieved through the reduction of runway occupancy time. Researchshows that even the slightest delays of=n the runway, as little asfive seconds, can cause two missed slots within an hour. It meansthat flights will be delayed such that the number scheduled to usethe runway per hour will be reduced by two flights. Two flights froman international airport can mean losses and accumulative delays thatcould destroy their customer relations. Messing up with the customersin such a business is one of the major causes of downfall. The runwaycapacity can be improved by up to 15% by the use of the ACE programand its associated initiatives.

Forthe ACE program to work through the reduced runway occupancy time, ithas to consider the departing and arrival times of flights. These twoactivities are the reason why the airport has a runway, and thefailure of any one of them would be disastrous. The departingaircraft are considered first. With a free runway, the departingaircraft should be on its way without any delays whatsoever. The timemanagement is partly due to the pilot’s experience and realizationof the urgency of the matter. Pilots who are aware of the need toclear the runway for other flights usually have the highest responsetimes. Airlines should train their pilots on how to analyze therequirements of the runway and how urgent its availability is to theairport and its operations. Runway occupancy time can be reducedduring takeoff by reacting to take-off clearance with immediateeffect, being fully prepared for take-off, going through checklistsand completing them within the set time to avoid delays on therunway, and avoiding backtracking unless it is necessary.

Landingaircrafts should also be controlled as to give the minimum runwayoccupancy time. The pilots in charge of the aircrafts should be smartplanners who analyze the wind and the runway conditions, togetherwith the weight of the aircraft to determine the point at which theywill touch down. They should also be aware of the taxiways that canlead them out of the runway in a bid to create an efficient exitplan. The overall performance and decision-making is important indefining the time occupied by an aircraft on the runway. Even withthe ACE system, pilots are the people with the biggest responsibilityof reducing runway occupancy time. The ACE system is an importanttool for the Brisbane Airport and their intended reduction in runwayoccupancy time. The end result of this system is the properutilization of the runway resource to maximize on the number offlights and the number of customers transported to and from theairport on a daily basis. The end product is increased revenue interms of earnings. Another important benefit of this system is thatit will reduce the delays in flights, meaning that customers will flyat the scheduled times without fail. It will contribute to goodcustomer service and the eventual popularity of the airport.Popularity among passengers means that many of them will prefer usingthe involved airline’s aircrafts, bringing in more aircrafts andpassengers to this airport. With the increased business influx, thesystem will have to be modified if it gets stretched. However, theACE system is strong enough to handle millions of annual customersand thousands pf daily flights. It is also automatically and easilyadjustable in case of delays and emergency landings.

Inthe designing of such a system, it is important to consider itsviability. It should be included in an airport that can handle it,and should be adjustable such that it can be used in other airports.It should also look into gate allocation, whereby gates are scheduledaccordingly to avoid any delays or flight changes. The ACE systemmeets all these requirements and qualifies to be a solution to thegrowing air traffic industry. It is because it also looks at theprospects for physical growth, such as the construction of theparallel runway, and puts it into consideration despite the fact thatits construction is yet to be completed. The system will easilyinclude the new runway in its algorithms upon completion, anddistribute some of the flights that would have used the older runway.It will also include the terminals that will come with the new runwayand factor in the increased flights and passengers. Such a system isefficient and can manage an airport’s resources responsibly. Itwill result in the increased revenues of the airport and might evenrequire additional personnel to handle the increased number ofcustomers while still maintaining the high customer service levelsthat the airport boasts of.

MelbourneAirport, the second busiest airport in Australia, is another exampleof the importance of managing the runway occupancy time. The airportis characterized by heavy traffic due to its strategic position. Italso expects a high growth in the number of aircrafts and passengersconnecting to other countries through it. The growth is summed upwith the increasing number of airplanes and passengers arriving anddeparting from this airport on a daily basis to give amazingly hugedaily figures of traffic. Such statistics re expected to grow by morethan double by the year 2030. The issue, however, needs to be takenas a major challenge due to the limitations in resource-utilizationwithin the airport.

Theairport is characterized by delays and change of flights. Thenegative aspect is described by frustrated pilots who have to stayairborne for close to half an hour of additional time as they waitfor the runway to be cleared. The bad thing about commercial airtraffic is that each plane has to land on a runway unless it is acase of emergency landing and the aircraft is not near an airport.The use of these runways has increased in Melbourne airport and asolution has to be designed to accommodate the big number ofaircrafts and passengers. The airport has to work in collaborationwith the airlines using it and the air traffic control to get asolution that utilizes the available resources effectively. The twomajor resources essential in such a set up are time and space. Timeis important because it takes into consideration the urgency withwhich some passengers have to travel, and the keeping of appointmentsthat were previously planned for by passengers. The wastage of timeleads to delays of flights and the lateness of passengers in takingtheir flights. A good air traffic control unit checks the mentionedrequirements of time and manages them with utmost importance becauseof their high value. Delays on the runway are not welcome becausethey disrupt the day of operation.

Anotherimportant resource that the airport values is space. Space isrequired I terms of having good runways that accommodate the flightof aircrafts. They also have terminals through which passengersarrive and depart. These terminals are sometimes crowded, especiallyduring peak seasons and hours which experience the highest numberincoming and outgoing aircrafts. It also entails the gates throughwhich aircrafts have to go through. The space is easily expandable ifthe airport has enough land resources. However, construction of newrunways, gates and terminals is costly and requires a lot of too. Anairport requires a properly modeled system to be able to functionappropriately. The modeled system assists the airport in the properallocation of available resources, such that they are optimally used.Wastage should be reduced to almost nil so that the airport canrealize increased revenue with the same resources that it previouslyhad, the difference being the level of utilization.

AirAustralia partnered with Melbourne Airport and leading airlines suchas Virgin Australia to implement the European ACE system, just likeBrisbane Airport did. Airport Capacity Enhancement was essential inidentifying the latent capacity of Melbourne Airport. It also came upwith appropriate methods that would tap this capacity through thereduction of runway occupancy time. The ACE system applied itsresearch in which five seconds of delays, caused two missed slotswithin an hour. It means that there will be less two flights perhour. If this loss in number of flights was to occur within everyhour of each day, the annual losses of the airport would betremendous. The realization of the possibility of such a loss callsfor immediate and swift action towards its mitigation.

Forthe ACE program to work and reduce the runway occupancy time inMelbourne Airport, incoming and outgoing flights have to beconsidered. The take-offs are considered first. In an idealsituation, the departing aircraft should take off without any delayswhatsoever. The accuracy and promptness are however not easilyachievable, taking into consideration the normal delays that takeplace at every single airport. Time management during take-off ismostly in the hands of the pilots. Their realization of the urgencyof the mater determines how efficient they can be in clearing therunway. Pilots who are aware of the need to clear the runway forother flights usually have the highest response times. Airlinesshould train their pilots on how to analyze the requirements of therunway and how urgent its availability is to the airport and itsoperations. Runway occupancy time can be reduced during takeoff byreacting to take-off clearance with immediate effect, being fullyprepared for take-off, going through checklists and completing themwithin the set time to avoid delays on the runway, and avoidingbacktracking unless it is necessary.

Landingaircrafts should also be controlled as to give the minimum runwayoccupancy time. The pilots in charge of the aircrafts should be smartplanners who analyze the wind and the runway conditions, togetherwith the weight of the aircraft to determine the point at which theywill touch down. They should also be aware of the taxiways that canlead them out of the runway in a bid to create an efficient exitplan. The overall performance and decision-making is important indefining the time occupied by an aircraft on the runway. Even withthe ACE system, pilots are the people with the biggest responsibilityof reducing runway occupancy time. The ACE system is an importanttool for the Brisbane Airport and their intended reduction in runwayoccupancy time. The end result of this system is the properutilization of the runway resource to maximize on the number offlights and the number of customers transported to and from theairport on a daily basis. The end product is increased revenue interms of earnings. Another important benefit of this system is thatit will reduce the delays in flights, meaning that customers will flyat the scheduled times without fail. It will contribute to goodcustomer service and the eventual popularity of the airport.Popularity among passengers means that many of them will prefer usingthe involved airline’s aircrafts, bringing in more aircrafts andpassengers to this airport. With the increased business influx, thesystem will have to be modified if it gets stretched. However, theACE system is strong enough to handle millions of annual customersand thousands pf daily flights. It is also automatically and easilyadjustable in case of delays and emergency landings.

Inthe designing of such a system, it is important to consider itsviability. It should be included in an airport that can handle it,and should be adjustable such that it can be used in other airports.It should also look into gate allocation, whereby gates are scheduledaccordingly to avoid any delays or flight changes. The ACE systemmeets all these requirements and qualifies to be a solution to thegrowing air traffic industry. It is because it also looks at theprospects for physical growth, such as the construction of theparallel runway, and puts it into consideration despite the fact thatits construction is yet to be completed. The system will easilyinclude the new runway in its algorithms upon completion, anddistribute some of the flights that would have used the older runway.It will also include the terminals that will come with the new runwayand factor in the increased flights and passengers. Such a system isefficient and can manage an airport’s resources responsibly. Itwill result in the increased revenues of the airport and might evenrequire additional personnel to handle the increased number ofcustomers while still maintaining the high customer service levelsthat the airport boasts of.

Developing the Model

Themodel was developed with reference to Singapore’s ChangiInternational Airport. While at it, one should note some statisticson this airport. It is the world`s 17th busiest airport, handlingover 53 million passengers every year. It serves 100 internationalairlines to more than 70 countries. It is slightly over 30 years oldand has four terminals. The airport has 2 runways, each with a lengthof 4 km.

Inthe formulation of the model, it was considered that his was aNP-hard problem. It means that there is no specific algorithm thatprovides the required optimization. Therefore, a mixed integerprogramming model was applied to give a modeling solution for ChangiAirport.

Constraints,sets and parameters are defined as below:

F-the set of all flights

n-the total number of flights

fϵ F : = {1,:::n} – a flight

ef- the on-gate time of flight f, a constant for each flight f

lf- the off-gate time of flight f, a constant for each flight f

G-the set of all gates

m-the total number of gates available

gϵ G := {1 :::m} – a gate

F(g)- the subset of flights that can use gate g

Sh- the set of pairs of gates that cannot be used simultaneously due toshadow restriction

GR2 {GR1… GR5} – a subset of gates (group of gates)

M-the number of conflicting flights allowed to be allocated to one GR

Xg,f- a decision variable that takes the value 1 if flight f is allocatedto gate g or 0 if not

Ug,f1,f2- an indicator variable that is 1 if the two flights f1 and f2 areboth allocated to gate g

pf1,f2- the constant penalty function for each flight pair which indicatesthe penalty to apply if flights f1 and f2 are both allocated to thesame gate.

Rf,g- the constant penalty for each gate-flight pair which indicates thepenalty to apply if flight f is smaller than the maximal size of theflight that can be allocated to gate g

Af,g- the constant penalty for each gate flight pair which indicates howstrongly gate g is preferred by an airline that is operating flight f

Freqf,g- the constant which indicates how often an airline which isoperating flight f has used gate g, based on statistic data analysis

d- the constant penalty for usage of `dummy gate`

Cd(f),Ca(f)- a set of flights which can be in conflict with the departure time(Cd(f)) or arrival time (Ca(f)) of flight f if the flights areallocated to the same group of gates.

Eachflight has to be allocated to a gate. A `dummy gate` (gate m + 1) isalso included in this constraint to which ungated flights will beallocated:

Notwo flights can be allocated to the same gate at the same time. Iftwo flights overlap or are within a minimum time gap of each otherthen they must be assigned to different gates, Inequality 4. The gapis set to 10 minutes, since some time is needed to clear the gate,however, the precise value is not important for these experiments,since there is an objective to increase these gaps.

Themathematical formulation of the time gap constraint is given below,it introduces a new variable which is set if two flights areallocated to the same gate. The constraint is introduced only forgaps smaller than 1:5 hour, which are more than sufficient. The gapswhich are smaller than 1.5 hour are maximized in the objectivefunction using the new variable so that potential, minor delays onthe day of operation can be absorbed.

Somegates cannot be modeled simultaneously, because usage of one gateblocks the usage of another one. Moreover, sometimes one large gatecan be treated as either two smaller gates (for smaller aircraft) orone large gate (for a single large aircraft). These large gates aremodeled as three separate gates with a shadowing constraint betweenthem.

LetSh denote the set of pairs of gates which have to be usedexclusively. The shadow restriction can then be described as follows:

Whenan aircraft is at the airport for a long time, it is often towed froma gate to a remote stand so that it is not occupying a gate thatcould be used by other aircraft at that time. The aircraft is thenusually towed back to the same or a different gate some time beforethe scheduled departure time. Although having many towing operationsis not desired by airports there is often no other option if aterminal does not have enough gates to serve all aircraft. It isassumed in this model that an aircraft is towed away from a gate ifit is going to stay at the airport for more than 4 hours. The timecan be changed according to airport preferences and has been chosenbased on the data set that has been used in the experiments. Theselong flights are modeled as two one-hour long flights. The firstflight represents the arrival and lasts one hour starting from thescheduled arrival time. The second flight starts one hour before thescheduled departure time.

Aspreviously discussed, the middle part is not modeled and sufficientremote stands are assumed, which it is usually the case for mostairports. An aircraft needs no more than an hour to do all servicingafter it arrives and no more than one hour before it departures, thetowing procedures are considered to be included in this hour. Anexact time of towing is not specified in the model, so thecontrollers may decide when to tow the flight and do so when theprocedure causes fewest conflicts. The conflict constraint is definedas follows:

whereGR is one of the gate groups and Cd(f1) is a set of flights which canbe in conflict with the departure time of flight f1 if the flightsare allocated to the same group of gates. The flights which are inCd(f1) are identified from the scheduled arrival and departure times.It is assumed that arrivals up to 10 minutes before/after thedeparture of f1 can conflict (i.e. a 10 minute long time gap isrequired). Similarly if an aircraft departs close to the departuretime of f1 it would be in Cd (f1), but the time margin can be shorterbecause it is a less conflicting situation so a 3 minute long timegap has been used. The 10 and 3 minute gaps can be set to any valuesdesired but have here been set to be large values to investigate andillustrate the effects of this constraint. An analogous set existsfor arrival time conflicts so there is a separate set Ca(f1) whichcontains all flights which are in conflict with the arrival of f1 anda constraint equivalent to Constraint 7 is applied. The idea offinding conflicting flights can also be viewed as a type of timewindow method. The conflicts within time windows would be identifiedby two windows for the arrival conflicts (short and long gaps) andtwo windows for the departure conflicts. But in a typical time windowapproach an artificial step size would usually be used to move thewindows along the problem whereas here the positions of the windowsare implicitly defined by the flight times, with consequent benefitsof avoiding time discretization problems.

Ourconstraint allows up to M flight movements for each group of gates.The number of allowed flights should be chosen depending on theparticular airport preference.

Theobjective function is a weighted sum of four elements. It aims toensure that time gaps between allocated flights are large enough toabsorb minor delays, that the gate spaces are used effectively, thatthe airline preferences are maximized and that the number of remote allocations is minimized.

Thefollowing function is minimized:

Thefirst component of the function refers to the time gaps, the variableUg,f1,f2 is forced to be one if two flights are at the same gate. Thegap cost function pf1,f2 that is used in the objective function topenalize such allocations is given by:

wheregapf1,f2 = ef2 – lf1 if ef2 &gt ef1 or gapf1,f2 = ef1 – lf2 if ef1 &gtef2 . The shape of pf1,f2 is appropriate for the problem since thesmaller gaps are penalized much more heavily than the larger gaps.Moreover pf1,f2 saturates, which means that the penalty for reallylarge gaps is fixed. It is reasonable to argue that only the gap fromthe immediately preceding flight on the gate should matter and thatgaps from earlier flights should be irrelevant.

Thesecond component refers to the effective usage of the gate space. Thesize of a flight that is allocated to a gate should correspond to thesize of a gate, possibly being equal to the maximal size that gate gcan absorb. It cannot be larger, which is guaranteed by the firstconstraint. When it is smaller, then such allocation is penalized, tokeep the larger gates clear in case they are needed later if flightshave to be reallocated. The gate-flight size function rf,g that isused to penalize the allocation is given by the equation below, wherebiggestGateSize refers to the size of the biggest gate at a terminal.

Thethird component of the objective function refers to the airlinepreferences. The preferences are established based upon statisticaldata analysis, by checking how often particular gates have been usedby particular airlines. The airline preference factor af,g which isused in the objective function to weight the allocation of flight fto gate g is calculated according to the equation below, wheremaxFreq is the maximal value of all frequencies and refers to themost frequently used gate by the most popular airline at theterminal.

Thelast component refers to the `dummy gate` allocations. `Dummy gate`symbolizes remote stands or holding places at an airport representingaircraft which cannot be allocated to gates so usage of it isstrongly penalized.

GAMS Implementation

Themodel was implemented in General Algebraic Modeling Software tosimulate and test for optimization. The optimization was in terms ofreduced runway occupancy time (ROT), which is directly related togate allocation of an airport. Poor gate allocation leads to poorusage of the runway. The negative aspect is due to the delay offlights on the ramp as they wait for others to finish using the gatesthat they should utilize. Modeling this system reduces the need tophysically expand the airport. It creates an avenue that allow forthe optimum usage of resources. The increased efficiency leads to thereduction of costs and the increase in revenue earned from activitiesof the airport.

GAMSsoftware was easy to use in defining the variables, constraints, andparameters. In the end, the modeling was successfully done using thecode below:

Partof the output, as taken from a screenshot, is as below:

Conclusion

Theoptimization is shown with the resulting data presentation. Themodeling of the airport operations was successful in reducing therunway occupancy time and increasing the airport’s efficiency interms of handling flights and customers. The result of this researchis a recommendation of the implementation of this model in achievingoptimization of airport activities without necessarily having to addthe number of runways they have. It will also assist airports toachieve their growth prospects in terms of increased flights andpassengers in the near future. The future of airports lies with theirmanagement and optimization.

References

Goldthorpe,S. (1997). Sensitivityof runway occupancy time (ROT) to various rollout and turnoff (ROTO)factors.Hampton, Va.: National Aeronautics and Space Administration, LangleyResearch Center

Hueschen,R., &amp Hankins, W. (2001). Rolloutand turnoff (ROTO) guidance and information displays effect on runwayoccupancy time in simulated low-visibility landings.Hampton, Va.: National Aeronautics and Space Administration, LangleyResearch Center

Yoshikawa,T. (1959). Theinfluence of exit taxiway locations on runway occupancy time.

Jackson,T., &amp Moy, E. (1979). Ananalysis of runway occupancy time for large aircraft at San FranciscoInternational Airport.Berkeley, Calif.: Institute of Transportation Studies, University ofCalifornia.

Kissane,B., &amp Harradine, A. (2000). Mathematicalinteractions: Algebraic modeling.Chatswood, N.S.W.: Shriro Australia. &nbsp

Koenig,S. (1978). Analysisof Runway Occupancy Times at Major Airports.Ft. Belvoir: Defense Technical Information Center. &nbsp

Runwayoccupancy time of landing aircraft at Indianapolis Municipal Airport.(1958). Washington, D.C.: U.S. Dept. of Commerce, Civil AeronauticsAdministration.

TheProvision of exit taxiways at Washington National Airport and theireffect on runway occupancy time of landing aircraft.(1958). Washington, D.C.: U.S. Dept. of Commerce, Civil AeronauticsAdministration.

Weiss,W., &amp Barrer, J. (1984). Analysisof runway occupancy time and separation data collected at La Guardia,Boston, and Newark airports.Washington, D.C.: U.S. Dept. of Transportation, Federal AviationAdministration, Office of System Studies and Cooperative Programs.