UBC Theses and Dissertations

UBC Theses Logo

UBC Theses and Dissertations

A simulation analysis of the passenger check-in system Arnett, John Douglas 1971

Your browser doesn't seem to have a PDF viewer, please download the PDF to view this item.

Item Metadata

Download

Media
831-UBC_1971_A4_5 A75.pdf [ 5.78MB ]
Metadata
JSON: 831-1.0101735.json
JSON-LD: 831-1.0101735-ld.json
RDF/XML (Pretty): 831-1.0101735-rdf.xml
RDF/JSON: 831-1.0101735-rdf.json
Turtle: 831-1.0101735-turtle.txt
N-Triples: 831-1.0101735-rdf-ntriples.txt
Original Record: 831-1.0101735-source.json
Full Text
831-1.0101735-fulltext.txt
Citation
831-1.0101735.ris

Full Text

A SIMULATION ANALYSIS OF THE PASSENGER CHECK-IN SYSTEM  by  JOHN DOUGLAS ARNETT B.A., Simon Fraseir U n i v e r s i t y ,  1970  A THESIS SUBMITTED IN PARTIAL FULFILMENT OF THE REQUIREMENTS  FOR THE DEGREE OF  MASTER OF BUSINESS ADMINISTRATION i n the Faculty °f COMMERCE AND BUSINESS ADMINISTRATION  We a c c e p t t h i s required  t h e s i s as conforming  standard  THE UNIVERSITY OF BRITISH September, 1971  COLUMBIA  to the  In presenting t h i s thesis in p a r t i a l f u l f i l m e n t of the requirements for an advanced degree at the University of B r i t i s h Columbia, I agree that the Library s h a l l make i t f r e e l y a v a i l a b l e for reference and study. I further agree that permission for extensive copying of t h i s thesis for scholarly purposes may be granted by the Head of my Department or by his representatives.  It  i s understood that copying or p u b l i c a t i o n  of t h i s thesis f o r f i n a n c i a l gain s h a l l not be allowed without my w r i t t e n permission.  Department The University of B r i t i s h Columbia Vancouver 8, Canada  Date  £>£*>7i*:*nA£*?.  /f7/  The objective of this thesis i s to provide a comprehensive management tool that w i l l aid in policy formulation and evaluation of Air Canada's Passenger Check-in System. The tool i s a computer simulation model that was verified to ensure that representative information could be derived from the model concerning the state of the real system.  The model can be used to deter-  mine the effects of a given policy or passenger arrival composition on the three performance objectives of the system; namely, the utilization of the baggage and ticket f a c i l i t i e s , a minimum passenger waiting time in the system, and the checking-in of passengers i n accordance with procedural policy. The simulation model describes the state of the system and subsequent assessment of the effect of a policy on the objectives with their statistics:  1. the average utilization of the f a c i l i t i e s , 2. the percentage of passengers exceeding two and a half minutes i n any one queue, 3. the average transit time per passenger (summation of delay times i n the system).  An analysis u t i l i z i n g the simulation model was undertaken i n three areas of policy management. They are: 1. To determine the implications of the behaviour of the Passenger Check-in System on policy formulation.  2.  To determine the f a c i l i t y policy that should be imple-  mented in order to achieve the present service policy as well as the maximum capacity of the system. 3.  To formulate alternative operating policies and test  for v i a b i l i t y prior to implementation of the policy. The results of the analysis were as follows: 1.  The service policy that a minimum of 15% of the Revenue  passengers be allowed to exceed 2.5 minutes of waiting time has been formulated correctly. 2.  The nature of the system is such that greater utilization  of f a c i l i t i e s w i l l not be achieved by a nominal increase in the allowable percentage of passengers exceeding 2.5 minutes. 3.  The f a c i l i t y policy and associated methodology has  been formulated so that the objectives of the system w i l l be attained. A.  The maximum capacity of the system occurs when the  arrival rate is in excess of 200 passengers per 15 minute period. 5.  The use of a single queue at the Revenue ticket counter  w i l l ensure greater attainment of the system objectives than the use of multiple queues. 6.  The combining of the baggage and ticket operations at  one counter is a viable alternative in the present system.  Page LIST OF TABLES  vi  LIST OF FIGURES  vii  CHAPTER I  II  III  INTRODUCTION  .  . .  1  The Sytems Concept  1  A Specific Application: The Airport System  5  The Thesis Defined  11  SYSTEMS IDENTIFICATION  13  A Review  13  The Passenger Check-in System  16  A.  Definition and Isolation of the System  16  B.  Boundaries of the System  28  C.  Formulation of the Objectives of the System  28  SYSTEMS ANALYSIS  31  A.  Statement of the Problem  32  Objectives of the Analysis  33  Design of the Analysis  35  Part 1 - Why Simulation  35  Part 2 - Construction of the Model  36  Part 3 - Verification  51  B.  Chapter  Page  C. Analysis of the System  IV  68  Step 1  70  Step 2  78  Step 3  82  CONCLUSIONS  90  The systems Approach Concept  90  Conclusions of the Systems Analysis . . . . . Areas of Future Research BIBLIOGRAPHY APPENDIX I - Air Canada F a c i l i t i e s APPENDIX II - Flowcharts of the Passenger Check-in System APPENDIX III - Program Listing and Sample Output APPENDIX IV - USER'S MANUAL APPENDIX V - Report to Air Canada  . . 91 93 94  Table  Page  3.1  Group Type Distribution  42  3.2  Group Size Distribution  3.3  Mean Service Times for Baggage Counters  48  3.4  Parameter Allocation  54  3.5  Simulated Arrival Stream as Compared to Observed  3.6  Comparison of System Statistics  62  3.7  Averages of System Statistics  65  3.8  Compatibility Tests for Optimal Policies  67  3.9  System Statistics from Analysis: Step 1 - Revenue  .  . . . .  Passenger System  43  61  72  3.10  Facility Policy for Various Arrival Rates  79  3.11  F a c i l i t y Policies and System Statistics  80  3.12  Revenue/Standby Ticket Operations in F u l l Separation  3.13  Comparative Statistics for Revenue Ticket Counter under Various Operating Policies Standby Operations Compared  3.14  . .  83 86 89  Figure  Page  1.1  The System as a Black Box  2  1.2  The Airport as a Black Box  6  1.3  Systems in Operation at the Airport  7  1.4  Strata of Airport Systems  9  2.1  Translation of General to Specific System  17  2.2  Arrival Stream  21  2.3  Revenue Passenger Baggage System  22  2.4  Revenue Passenger Ticket System  23  2.5  Standby Process Stream  26  3.1  Distribution of Load Factors  41  3.2  Arrival Stream Specification  45  3.3  Check-in Process Specifications  46  3.4  Probability Distribution for Ticket Counter Service Times  49  3.5  Search for Steady State - X = 100  57  3.6  Search for Steady State - X =  58  3.7  Effect of Arrival Rate on Transit Time  3.8  Arrival Rate and Corresponding Utilization  3.9  Arrival Rate and % Passenger Units exceeding 2.5 Mins.  3.10 /  50  Average Utilization and % Passenger Units exceeding 2.5 Min. Waiting  .  73 74  .  76 77  Figure  Page  3.11  Multiple or Single Queues  85  3.12  1 Standby Agent Performs Both Ticket and Baggage Operation  88  I wish to express my sincere thanks to Assistant Professors D. H. Uyeno and J. B. Sidney of the Management Science Department, Faculty of Commerce and Business Administration, the University of British Columbia. Dr. Uyeno, the committee chairman, was of great assistance in the preparation of my thesis. the undertaking of this thesis.  Dr. Sidney was a motivating force in Once again, thank you both.  I would also like to express my gratitude to Air Canada, in particular, Mr. C. Morgan, Passenger Service Co-ordinator, Vancouver. His assistance and advice was of great value.  INTRODUCTION THE SYSTEMS CONCEPT The "systems approach" concept has been utilized since the middle of the 1930's as a method to handle complex and dynamic problems.  The  methodology employed is commonly pragmatic in nature; and its output is generally designed for the ultimate user.  Due to this pragmatic  approach, there are two logical divisions in the procedure.  First,  there is the identification or definition of the problematic system. Second, there is the system analysis or the determination of a solution for the system. The f i r s t stage has three parts: A.  Definition and isolation of the system;  B.  Determination of the boundaries of the system;  C.  Formulation of the system's objectives.  The f i r s t part, one of major importance, is the explicit definition of the system.  For the purpose of explaining this primary step of  the system approach, the common analogy of the 'Black Box' w i l l be used. A Black Box is a system which receives certain inputs and produces outputs.  The conversion or transformation of the inputs into outputs is  accomplished through the interaction of the components located within  the Black Box.  (See Figure  Inputs  Figure 1.1  -  1.1).  BLACK BOX  Outputs  The System as a Black Box  This definition should identify the elements of flow (inputs and outputs) and the conversion process (relationship of inputs to outputs).  The behaviour of the system should be defined as to its degree  of predictability. The second part is the determination or setting of the boundaries of the system that w i l l be analyzed.  Consider once again the Black  Box concept.  Any Black Box can be sub-divided into smaller components or  sub-systems.  By the same token, each system can be sub-divided into  smaller sub-systems or i t may be a sub-division of a larger system. Thus, there is a hierarchy of systems and the specific system under consideration must be identified as to its limits. The third and final step in the identification of a system is the establishment of the operational objectives of the system.  Hare, Van Court, Systems Analysis and World, 1967), Chapter I.  This  (New York: Harcourt, Brace  formulation of standards w i l l be employed to measure the system's performance.  It is also a logical key to the development of a useful and  efficient analysis.  The system must be defined as to what i t is supposed  to do and in what manner. The second stage of the systems approach is the analysis of the system defined.  It also has three parts:  A.  The statement of the problem and the formulation of the objectives;  B.  The design of the analysis;  C.  The determination of viable alternatives.  A statement of the problem w i l l generally be directed towards the objectives of the system.  Are these objectives being met?  Is there  a problem in the control of the system? The formulation of the objectives of the analysis w i l l give formal direction to the analysis, define the ultimate user of the analysis and highlight the areas of research.  This is an explicit statement  of how the analysis w i l l be carried out. The second area is the design of the analysis.  A model of the  system may be formulated to test the system under various operating conditions in order to determine those conditions which lead to optimal performance.  Often these models are computer-based.  This model must  be consistent with the f i r s t . The result of the third part is the determination of what might be accomplished in the real system.  The presentation of the viable  alternatives w i l l enable the decision maker to appraise the problematic system and implement a decision.  In summation, then, the Systems Approach i s : 1.  Systems I d e n t i f i c a t i o n A. B. C.  2.  D e f i n i t i o n and i s o l a t i o n of the system. Determination of the boundaries of the system. Formulation of the system's objectives.  Systems Analysis A. B. C.  Statement of the problem and the objectives analys i s . Design of the a n a l y s i s . Determination of v i a b l e a l t e r n a t i v e s .  of  Turning the discussion from the theoretical to a specific application should clarify the systems approach. The application of the system identification to the Airport is simplistically modelled in Figure 1.2.  The input/output flow elements  of the system are defined by their state - either arriving or departing. In describing a real system, there is an associated time element - t^. The arrivals (t^) must occur before the departures  (t^) in the modelled  system as well as in real systems (t^ _< t^). Just as the general system identification was expanded to reveal its components or sub-systems, so might the Airport system.  The two  major sub-systems are identified by the kinds of flow elements - passengers and aircraft. Just as the Black Box might be divided into s t i l l smaller, more functionally oriented components, so can each of the major sub-systems of the Airport.  Consider the Passenger Flow System.  There is a three-  fold sub-division of this system in operation at an airport. defined, there are the following sub-divisions: 2) the Baggage Flow System; and is represented in Figure 1.3.  Sequentially  1) the Check-in System;  3) the Gate Procedure.  This division  The three systems interact continuously  to form the Passenger Flow System.  Passengers may either arrive or  depart by plane. The Check-in System is the process of validating tickets, allocating excess baggage charges and checking baggage. part of the Baggage Flow System.  This system initiates  The boundaries of this system are  MAJOR SUB-SYSTEMS AIRCRAFT SYSTEM DEPARTURES  ARRIVALS  PASSENGERS SYSTEM AIRPORT  Time Element  Flow Elements: Passengers and Aircraft  < t. Figure 1.2 The Airport as a Black Box  AIRPORT  AIRCRAFT FLOW SYSTEM  7 GATE & TAXI PROCEDURE  MAINTENANCE  CARGO HANDLING  CHECK-IN  BAGGAGE FLOW  GATE PROCEDURE  Figure 1.3 Systems in Operation at the Airport  PASSENGER FLOW SYSTEM  defined by the state of the passengers.  As an input to the system,  passengers queue at one of the available counters and become outputs when check-in is completed. The Baggage Flow System transforms terminal baggage checked by passengers to on-board cargo and vice versa. The  third system operates at the Departure Lounge of the airport.  Passengers receive a final ticket validation and seat selection is made. The output of the system consists of specific passengers on board specific flights. If  the flow element were defined as an aircraft, then there are  three sub-systems present: Maintenance.  Gate and Taxi Procedure, Cargo Handling, and  These three systems may be described in connection with the  state of the aircraft in the system. Taxi Procedure System.  For example, consider the Gate and  When an aircraft arrives i t must taxi by a definite  path and terminate at a particular gate. reverse procedure is followed.  When the aircraft departs, the  The other systems are as their name would  suggest. The hierarchy of the airport systems may be classified into various strata (see Figure 1.4).  A system which has been identified as  being of the third strata, is of the appropriate size and dimension for a viable system analysis (the Passenger Check-in System, etc.).  The  systems have succinct and readily definable boundaries, objectives, inputs and outputs. If  optimal methods are formulated and implemented for each sub-  system, then the entire system w i l l advance closer to a state of optimality. R. Mountjoy, "Airport Simulation Models," AGIFORS Proceedings, 1969, pp. 619-42.  Strata 1  PASSENGER  AIRCRAFT  CARGO  MAINTENANCE  Strata 2  GATE, TAXI PROCEDURE  CHECK-IN  GATE  BAGGAGE Strata 3  ON-LOAD  STANDBY  REVENUE  OFF-LOAD  Strata 4 Figure 1.4 Strata of the Airport Systems  TICKET  BAGGAGE  The justification for a study of airport systems is quite apparent i f a major a i r l i n e terminal has been observed during periods of peak volume.  The main problem results from simultaneous arrivals and depar-  tures of aircraft.  The resultant volume of passengers and aircraft  tends to 'clog' the systems at the airport. Specifically, there are several reasons for a study to be undertaken. 1. There has been an increased demand for air transportation. Airlines have met this demand with expanded fleets and increased capacity of aircraft.  As a result, greater volumes of passengers must be pro-  cessed through air terminals. 2. There must be more sophisticated technologies employed to co-ordinate, plan and control the volumes of traffic at airports.  The  question remains as to the selection and implementation of these technologies for viable results. . 3. The air transport industry is a highly competitive business. The a i r l i n e which supplies quality service both in the air and on the ground, while maintaining lower costs, w i l l benefit.  This can only be  achieved through efficient use of resources in their systems.  The general direction of this study has been indicated in the previous two sections of Chapter I.  What remains is to define the exact  nature and scope of the thesis as i t w i l l be presented. The systems approach w i l l be applied to a readily available and familiar system - the Passenger Check-in System.  The methodology  used in this thesis w i l l be applicable, in the most part, to any major airline check-in system.  The specific details of the thesis w i l l be  derived from the Air Canada system, Vancouver International Airport, Richmond, British Columbia. Chapter II w i l l present the f i r s t part of the systems approach - the Passenger Check-in System Identification.  The elements, boundaries  and objectives w i l l be discussed to enable a greater understanding of the system's internal operations. Chapter III w i l l present the detailed analysis of the system. As part of the analysis a computer simulation model w i l l be constructed. Experiments w i l l be performed on the model in the area of service policy formulation and evaluation.  Appendix III contains a program l i s t i n g of  the model and Appendix IV, a User's Manual for the model. Chapter IV contains a summary and conclusions.  The conclusions  of the thesis w i l l be pertinent to both the present system and future research on the system.  Appendix V contains a management summary directed  towards the Management of Air Canada. There is a twofold objective of this thesis: 1.  To provide a comprehensive tool that w i l l make possible the testing of alternative mangement policies.  2. To f u l f i l l the requirements for the Master of Business Administration degree at the University of British Columbia.  SYSTEM IDENTIFICATION This chapter w i l l deal primarily with the identification of the Check-in system.  To begin with, a brief review of previous studies  concerning the Passenger Check-in system would be appropriate.  This is  a logical step for any analyst since i t could provide insight into the methodology employed for problematic areas.  A REVIEW The five studies reviewed below were extracted from the AGIFORS Proceedings  (The Airline Group, International Federation of the Operations  Research Society) from 1967 to 1969. 1 1.  Title:  Author:  2.  Optimal Staffing  at Airline  Terminals,  1969, pp. 15-42.  K. C. Khanna and H. Takamori (airline not mentioned).  Objective:  To determine the manpower requirements and shifts for the various functions at the airline terminal.  Technique:  The problem was formulated as an optimization over a large number of small integer problems. Subsequent use of branch and bound methods.  Title:  Author:  Airport  Simulation  Models, 1969, pp. 619-642.  R. Mountjoy, United Airlines.  An annual symposium of the major airlines is held for the purpose of discussing studies on problems encountered and their respective solutions.  Objective:  To handle inter-airline questions concerning (a) runway capacity; (b) taxi times and terminal configurations; (c) a i r l i n e schedules, terminal f a c i l i t i e s and station policies affecting customer service; (d) airport f a c i l i t i e s response to increased traffic forecasts.  Technique:  3.  Title: Author:  Subdivide the Airport system into several simulation models to handle complex inter-relationships.  Airport  Manpower Planning  Systems, 1969, pp. 681-696.  V. K. Wozniuk, Air Canada.  Objective:  to provide management with accurate short-range forecasts (up to two or three years) of manpower requirements at certain airport positions to be used for planning of budgets and agent training. Also to provide management with a means of evaluating the proposed schedules and various standards.  Technique:  Sub-divided AMPS (Airport Manpower Planning System) into three areas: (a) Check-in Counter - arrival stream and its properties; - service time distribution; - queue discipline and system type. (b) Departure Gate - departure requirements for five-minute intervals. (c) Arrival Gate - same as departure gate. Uses classical queueing methods to determine requirements.  4.  Title:  Manpower Planning 245-267.  for Airport  Author: S. O'Broin, Air Lingus.  Operations,  1968, pp.  5.  Objective:  To determine manpower requirements for the Passenger Flow system at various levels and service standards.  Technqiue:  Uses classical queueing theory for requirements. Scheduling is determined by an algorithm.  Title: Author:  Passenger Handling  Model, 1967, pp. 52-80.  W. Gensema, Royal Dutch Airlines (KLM).  Objective:  To determine the number of service counters required for a specific standard of service.  Technique:  The derivation of the model uses both classical queueing theory and Monte Carlo simulation.  The studies that have been cited here, are not exercises in the use of pure theory, but are practical approaches to the determination of viable solutions for problems.  The techniques varied depending upon  the problem, but when queueing situations were encountered, classical queueing theory was utilized extensively.  Unfortunately, the use of  this body of mathematic theory becomes extremely complex, i f not impossible, when applied to large systems.  Many of the objectives of a detailed  systems analysis are unattainable, because of the assumptions and limitations of the theory.  While the theory is applicable for static analysis  of a homogeneous system, the theory does not represent the dynamic behaviour which characterizes a real system.  The queueing behaviour of real and  complex systems is extremely d i f f i c u l t to analyze with classical queueing theory due to the simultaneous  movement of elements in the various queues.  On the other hand, a simulation model of a queueing situation has the ability to handle heterogeneous elements in queues and to reveal the true behaviour of a dynamic system.  THE PASSENGER CHECK-IN SYSTEM The system identification division, as previously stated, contains three main steps: A.  Definition and isolation of the system;  B.  Determination of the boundaries of the system;  C.  Formulation of the objectives of the system.  Each of these steps w i l l be presented and explained in turn as the study is directed towards the application of the systems concept to the Passenger Check-in System.  A.  Definition and Isolation of the System To the casual observer at the airport, the arrival of passengers  for check-in, prior to flight departure, is at times confusing to say the least.  However, there are patterns of behaviour that can be discerned  so that the whole system may be defined.  As stated in Chapter I, the  check-in system may be described generally as a flow system with input elements, a transformation or conversion process, and output elements. The translation of the general system definition to that applicable for the check-in system may be seen in Figure 2.1.  The input elements are  passengers but the process is defined as the arrival stream.  The con-  version of inputs to outputs is the check-in process consisting of baggage and ticket counters.  The output element of the system may be defined  simply as a passenger that has been checked i n . The Arrival Stream.  The arrival or input stream for the check-in  system may be defined by the number of passengers entering the system.  ARRIVAL STREAM  DEPARTURE STREAM  PASSENGER ARRIVAL  INPUT ELEMENT  CHECK-IN  CONVERSION  PROCESS  PROCESS  PASSENGER, CHECKED-IN  Figure 2.1 Translation of General to Specific System  OUTPUT ELEMENT  The passengers for a particular flight usually arrive in the 1 1/2 hour 2 time period prior to departure.  The number of passengers that arrive in  a given time interval within that period is a function of the load factor (ratio of the number of passengers to maximum capacity of the aircraft)and the amount of time until the aircraft departs.  It was observed that a  certain percentage of the passengers arrived in each of the six 15-minute periods.  This percentage was observed to remain approximately constant  for a particular flight even though the load factor varied. For example, i f the arrival pattern for a DC-8L (198 passenger capacity) is such that 30 per cent of the passengers arrive in a certain 15-minute period, then 47 passengers w i l l arrive i f the load factor is 80 per cent and 35 i f the load factor is 60 per cent.  More w i l l be said  about this later. The arrival pattern of passengers for different flights overlap due to the scheduling of departure times.  Thus the pattern might be such  that in a 15-minute time interval 50 passengers for flight X arrive as well as 45 passengers for flight Y.  The arrival stream consists of 95  passengers for the particular 15-minute period. Another characteristic of the arrival stream is that passengers travel in groups.  The groups themselves are independent upon arrival  at the airport i f the assumption is made that the small number of groups that do arrive by bus do not affect the arrival pattern significantly. The group size may range from one to six passengers on the most part. S. O'Broin, "Manpower Planning for Airport Operations," AGIFORS Proceedings, 1968, pp. 245-67.  The occurrence of groups larger than six is a very small proportion of the total group arrivals.  The arrival rate of groups may be approximated  by a Poisson distribution since the group arrivals are independent. The number of group arrivals is then the number of passengers divided by the average group size. The passenger groups arriving at the airport are of two types: independent or dependent.  The distinction is made because of the method  of processing passengers at the check-in counters.  The passengers of a  group may be checked-in independently or dependently.  If the group is  of the independent type, the passengers of that group are processed individually.  A group of size one, w i l l , by definition, be independent.  Another example of an independent group would be two or three business associates travelling together.  When they arrive by taxi, they do so  as a group, but are processed at the counters individually. The dependent groups that arrive at the airport are families. They are processed at the check-in counters as one. Due to the method of processing the passengers, the flow elements w i l l be defined as "passenger units."  A family or one businessman in a  group represents one passenger unit. The final characteristic of the arrival stream is the identification of the passengers travelling at f u l l fare (Revenue) or reduced fare (Standby).  Revenue passengers have confirmed seats while Standby  passengers travel only i f space is available.  This characteristic deter-  mines whether the group w i l l enter the Revenue or the Standby Passenger System.  It was observed that passengers within groups rarely were of both classifications  (Revenue and Standby in one group).  always directed to the Revenue system.  Families were  The flow chart of the arrival  stream appears in Figure 2.2. In summation, the arrival stream may be identified statistically by: 1.  The number of passengers entering the system in a specific period;  2.  The group size and type distribution for a particular airport;  3.  The average number of Revenue passengers as a percentage of total for a particular flight.  The Check-in Process.  In Figure 2.2 i t was noted that a passen-  ger group may enter either the Revenue or Standby System.  The two systems  differ because of the operating policy employed by Air Canada.  The  operating policy for a system is the set of methods used to direct or control passenger movement through the system in a prescribed manner. As this thesis identifies the Revenue and Standby System, these operating rules w i l l be c l a r i f i e d . The Revenue System.  Figures 2.3 and 2.4, the flow charts of  the passenger flow for the Revenue check-in process, indicate the operating rules which are in effect.  As groups of passengers enter the airport,  the f i r s t operating rule is that the passengers must have a ticket before their baggage may be checked. arrive non-ticketed.  A small percentage of Revenue groups  If the group is non-ticketed, they go to the  Revenue Tickety System - another component of the Revenue System. (This system w i l l be explained later.)  GROUP ARRIVAL  Figure 2.2 Arrival Stream  NO M  UPTS  YES -J  RPTS  CURBSIDE CHECK-IN  BAGGAGE COUNTER CHECK-IN  Figure 2.3 Revenue Passenger Baggage System  RPTS = Revenue P_assenger T_icket System  Figure 2.4 Revenue Passenger Ticket System  RPBS = Revenue Passenger Eiaggage System  Continuing the flow, the next operating rule is that i f a passenger group does not have excess baggage, a "red cap" may be employed to tag the baggage and deliver i t to the Air Canada f a c i l i t i e s .  This is  called "Curb-side Check-in" since i t takes place at the terminal sidewalk.  The passengers do not have to be served at the check-in counters,  and may proceed to the appropriate departure lounge or explore the terminal shops.  A small percentage of passenger groups use this check-in procedure. Passengers having a ticket then select an agent for service at  the Revenue baggage counter.  The queueing discipline may be described  as the selection of the shortest queue.  It was observed that business  associates or friends travelling together tended to select a queue as a group.  Families must queue for a single f a c i l i t y as they are processed  dependently.  The changing of lines was minimal because passengers had  to move their baggage. When the baggage agent serves a passenger unit, the baggage is tagged with the appropriate flight number, the ticket is validated and the weight of the baggage is checked.  If the weight exceeds a certain  limit, the passenger unit is defined as having "excess baggage." The next operating rule is that i f there is excess baggage, the passenger unit proceeds to the ticket counter to pay the excess charge. If there is no excess baggage, the passenger unit has completed the check-in process and proceeds to the Departure Lounge. The f a c i l i t y utilization (the percentage of time an agent was utilized for service to the total time interval) is dependent on the f a c i l i t y policy and the demand for service at the counters.  The f a c i l i t y  policy which specifies the number of counters that w i l l be available for passenger service is set by management. The service time for the baggage counters (the length of time an agent is employed in service with a passenger) is dependent upon the size of the passenger unit and whether or not there is excess baggage. As indicated in Figure 2.3, there are two procedures by which a passenger might enter the Revenue Ticket System - a passenger group arriving at the airport non-ticketed or a passenger unit having excess baggage charges to pay.  This distinction is evident in Figure 2.4 as  the passengers flow through the system. The queueing discipline in the Revenue Ticket System is similar to that of the baggage counter.  When a passenger purchases a ticket,  only the Revenue ticket counter may be used.  The Standby ticket counter  may also be utilized for excess payment.  The service times are dependent  upon the nature of the service required.  The mean service time for ticket  purchases was greater than that for excess baggage payments.  Once the  ticket is purchased, the passenger unit proceeds to the Revenue baggage counters.  If excess payment has been made, the passenger unit has com-  pleted the check-in requirements. The Standby System.  The operating policies of the Standby System  are represented in Figure 2.5.  The procedure is similar to that of the  Revenue Passenger System, except there is no option for curbside check-in. Passengers must have times and priorities concerning their Standby status determined at the baggage counter. that of the Revenue System.  The ticket procedure is the same as  Non-ticketed groups proceed to the "Standby  PURCHASE TICKET  BAGGAGE CHECK-IN  PAY EXCESS CHARGE  Figure 2.5 Standby Process Stream  Ticket Purchase" counter.  Passenger units with excess charges to be  paid may u t i l i z e the Revenue ticket counters as well as the Standby ticket counter. To aid in the conceptualization of the system, a floor plan of the f a c i l i t i e s with the flow of passengers superimposed on the floorplan appear in Appendix I. In summation, the Check-in System may be statistically defined by: 1.  The operating policies in effect;  2.  The distribution of service times at the various f a c i l i t i e s and the type of service demanded;  3.  The percentage of passenger units who arrive non-ticketed or use curbside check-in, or have excess baggage;  4.  The number of f a c i l i t i e s that are available for use in the system.  The Output.  The output of the Check-in System are passenger  units that have met a l l the requirements of the check-in procedure. The passenger units then proceed to the departure lounge for a final ticket validation and seat selection.  If time permits, passengers may  browse in the airport shops. Associated with the output, there are system s t a t i s t i c s .  These  statistics describe the flow elements of the system, given the definition of the arrival stream and check-in process.  The transit time s t a t i s t i c  is the time a passenger unit has spent travelling through the system. Another s t a t i s t i c is the aforementioned f a c i l i t y u t i l i z a t i o n . Other statistics may be formulated depending on the objectives of the analysis.  B.  Boundaries of the System The second step of the system identification is the determination  of the boundaries of the system.  The essence of this step is to structure  the system for a concise analysis by defining what is exogeneous to the system when the analysis is undertaken. The input boundary should be considered as the mean arrival rate of passengers.  No attempt w i l l be made to s t a t i s t i c a l l y determine  the functional relationship between the expected and actual passenger arrivals.  This figure - the number of passenger arrivals - w i l l be an  exogeneous variable to the system. The procedures of the agent w i l l be dealt with as just a distribution of service times.  If certain procedural changes were to be  made which would effect this distribution, the system could be restructured to reflect the change and subsequent state of the system. The output boundary is defined by the state of the passengers as they pass through the system.  Once the passengers have completed  the requirements at the check-in counters, they are deemed to be an output element.  No examination w i l l be made of the Departure Lounge at  this time. The results of the systems analysis could be used as inputs to a much larger systems analysis.  C.  Formulation of the Objectives of the System The final step in the identification of the Passenger Check-in  System is the formulation of the objectives.  These objectives w i l l be  vised to measure the performance of the system and to determine the desirability of alternative methods of operating the system. There are three main performance objectives for the Check-in System: 1.  To attain a high utilization of manpower in the system.  2.  To achieve a minimum customer waiting time in the queues.  3.  To check in the passengers properly.  There is undoubtedly a non-linear relationship between manpower u t i l i z a t i o n and customer waiting times. to the utilization of manpower. a high cost of service.  The cost of service is related  A low utilization is representative of  Only when there is a small number of passengers  in a system and the resultant demand for service is low, does this relationship become unrepresentative of variable costs.  The functional relation-  ship w i l l be examined in Chapter III. The strategies or service policies that have been formulated by Air Canada are the means of achieving the systems objectives.  These  service policies are reflected by the operating rules described earlier and are listed below. 1.  Separate the Standby passenger service from that of  the Revenue passenger.  Since Revenue passengers comprise a large per-  centage of travellers and pay f u l l fare for their tickets, they are the most important.  If Revenue passengers become dissatisfied, hopefully  i t w i l l not stem waiting for service while a passenger who has paid a lower fare is being served.  2.  Staff the f a c i l i t i e s such that 85 per cent of the Revenue  passenger units and 75 per cent of the Standby passenger units do not wait longer than 2.5 minutes i n any one queue. of service set by A i r Canada.  This i s the standard  I t should be mentioned that due to the  importance of the Revenue passengers, the 15 per cent who wait longer than 2.5 minutes i s a maximum number.  The Standby passengers are of  lesser importance and the 25 per cent should be considered as a working f i g u r e and not a s t r i c t maximum as i n the case of Revenue passengers. 3.  Use a 15-minute s t a f f i n g horizon to e f f i c i e n t l y  and  e f f e c t i v e l y a l l o c a t e manpower and forecast manpower requirements.  This  strategy i s employed because the demand f o r service fluctuates greatly over the day.  A 15-minute planning horizon w i l l allow f o r adequate  3 s t a f f i n g and transfer of agents to various functions i n other systems. The attainment of the objectives may be measured s t a t i s t i c a l l y by: 1. 2.  Facility utilization; The percentage of passenger units waiting i n queues longer than 2.5 minutes;  3.  The t r a n s i t time i n the system.  Since the system has been defined as to i t s inputs, outputs, conversion process, boundaries and objectives, the second step of the system analysis w i l l be undertaken. study i s of extreme importance.  The d e f i n i t i o n of the system under  I f a v i a b l e analysis i s to be accom-  plished a thorough understanding of the system both d e s c r i p t i v e l y and s t a t i s t i c a l y i s imperative. 3  V. K. Wozniuk, "Airport Manpower Planning Systems," Proceedings, p. 686, 1969.  AGIFORS  SYSTEMS ANALYSIS As stated in Chapter I, the analysis of a system consists of three steps: A.  A statement of the problem and the formulation of the objectives of the analysis;  B.  The design of the analysis;  C.  The understanding and determination of viable alternatives for the problematic system.  The statement which defines the problematic areas of the system is generally derived from the stated objectives of the system.  It should  define the problems encountered in operating the system and the possible source.  The second part of the step is the formulation of explicit objec-  tives for the analysis.  These objectives w i l l define exactly what w i l l  be undertaken in the analysis.  The result of this step w i l l be the  methodology used to rectify the problematic system. The second step is a direct result of the previous section. The design of the analysis is the mode or vehicle by which the analytical objectives w i l l be achieved.  This is the construction of a laboratory  tool that can be used for experimental analysis under controlled conditions. A c r i t i c a l part of this step w i l l be the verification of the modelled system.  The third step is the analysis of the system as set out in the objectives.  An important part of this analysis is the interpretation  and presentation of the results.  The ultimate user of the results must  be kept in mind since the analysis is undertaken for the decision-makers who control the system.  (The results of this step is the determination  of viable policies for the real system.) The three steps could be considered as a system, the inputs being the statements of the problem and the objectives, and the conversion process being the modelled system.  The output would be the  ability to test for and determine viable solutions for the problems of the system.  A.  Statement of the Problem The purpose in defining the objectives of the Passenger Check-  in System in Chapter II was to provide a means of evaluating the performance of the system.  Alternatively stated, i t is desired to determine  the degree of attainment of the stated objectives.  Policies are formu-  lated and implemented in the system as a means of achieving a desired state.  The inherent problem of operating a real system is the measure-  ment of the effectiveness  of implemented policies and the determination  i f proposed policies w i l l be viable in the system. The measurement of existing policies to determine i f they do in fact achieve their objective is a very real problem in a dynamic system. It is imperative for the management to know i f their policies achieve their purpose.  Efficient control of the system can then be maintained.  If new policies are formulated, then i t is necessary to forecast their feasibility in the system. avenues to follow.  The decision-makers have two  One is to formulate the policy and test for feasibility  by implementation in the real system. i f the hypothesis is false.  This could be costly and disruptive  The other avenue is to construct a model  of the system so that controlled testing of the policy may take place prior to the implementation. emulate the system.  The model used could either simulate or  A justification for this w i l l be presented shortly.  Specifically, Air Canada's management needs to know the u t i l i zation of f a c i l i t i e s and the associated percentage of passengers waiting in queues longer than 2 1/2 minutes.  If alternative methods of handling  passengers are formulated, what are the effects on the system?  What is  the processing capacity of the existing f a c i l i t i e s at the Vancouver Airport?  And f i n a l l y , in the present system, has Air Canada an optimal  f a c i l i t y policy for the various arrival rates encountered in a 15-minute time interval? These questions w i l l form the basis for the objectives of the systems analysis.  Objectives of the Analysis The objectives of the analysis w i l l be used to direct the analysis of the system.  If experiments are performed on the modelled system, the  answers to questions relating to the system w i l l be known.  Also, the  relationship of the dynamic elements in the system w i l l be revealed. The f i r s t objective is to provide a laboratory tool for the management of Air Canada that w i l l enable the evaluation of policies,  both implemented and proposed. l a t i o n model.  This t o o l w i l l be a computer-based simu-  I t w i l l be presented i n the second step of the systems  analysis - the design of the a n a l y s i s . The second objective i s to determine the r e l a t i o n s h i p between c e r t a i n variables or systems s t a t i s t i c s .  This w i l l a i d i n the formu-  l a t i o n of future p o l i c i e s and the understanding of present  ones.  This  part of the analysis w i l l examine the e f f e c t s of the a r r i v a l rates on passenger unit t r a n s i t time (as determined by the percentage of passenger units waiting i n queues longer than 2 1/2 minutes) and on f a c i l i t y  utili-  lization. The t h i r d objective of the analysis i s to provide a f a c i l i t y p o l i c y (P^) for the various a r r i v a l rates ( A ^ ) . This p o l i c y w i l l a t t a i n the system's objectives as stated i n Chapter I I . The fourth objective i s to determine, with the use of the simul a t i o n model, the maximum capacity of the e x i s t i n g f a c i l i t i e s system objectives.  given  When forecasts exceed this capacity, plans must be  made to adjust for the volume. The f i n a l objective of the systems analysis i s to determine i f there are a l t e r n a t i v e methods of handling passengers i n the system. Thus new operating rules w i l l be formulated, tested, and compared to present  operating  policies.  Since the statement of the problem and the objectives of the analysis has been made, the next step i n the systems analysis w i l l be presented.  B.  Design of the The  The in  first the  design  of  part w i l l  a n a l y s i s of  describe tified  Analysis  how  the  the  'simulate'  the  Part  s t a t i s t i c a l d e f i n i t i o n s and  known as  real  1 - Why  The  converted  In the  r e v i e w of p r e v i o u s  it  pointed  out  h a n d l e c o m p l e x and  the  defined  operations  research  third  fact  technique  t e c h n i q u e overshadow t h a t  the  the  Simulation  user to be  explicitly.  The  flow  elements or passengers  has  be  type.  the  encountered,  The  very  nature  Each  relation-  the model i s  system i s  to  realized. nature  classified  system.  as  through the  a  "flow"  system.  For  I.B.M. c o m p u t e r  of  this  P a s s e n g e r C h e c k - i n S y s t e m i s r e a d i l y a d a p t a b l e t o many case the  of  constructed  s i m u l a t i o n i s the very  'flow'  languages, i n t h i s  mathe-  this  the a b i l i t y  analytical.  Thus e v e n b e f o r e  s t r u c t u r e of  of  presented i n Chapter I I ,  i n such a system.  second reason f o r u s i n g  simulation  will  iden-  The  used i n a systems a n a l y s i s of  dynamic s t a t e s  s y s t e m u n d e r s t u d y , w h i c h may  able  the  is limited.  the  reason, the  r u l e s as  t h a t when c o m p l e x q u e u e i n g s y s t e m s a r e  a sound u n d e r s t a n d i n g of The  operating  computer model does i n  s t u d i e s w h i c h was  techniques forces  s h i p must be  2.  c o u l d be  queueing theory  simulation  the  for using  a n a l y t i c a l powers of  techniques that  classical  second p a r t  i s threefold.  matical  was  The  model  Simulation  justification  The  a simulation  i n t o a s i m u l a t i o n model.  that  parts.  system.  simulation  1.  for using  Passenger C h e c k - i n System.  the v e r i f i c a t i o n  the  presented i n three  present a j u s t i f i c a t i o n  i n Chapter I I are  p a r t w i l l be  a n a l y s i s w i l l be  avail-  simulation  language - GPSS/360 (General Purpose Simulation System). 3.  The third reason is that once the model is constructed i t  becomes a very useful management tool to control the system.  Relatively  inexpensive experiments may be made under controlled conditions.  If  a policy is to be tested, the decision maker simply specifies the i n i t i a l state of the system or flow charts the new path of the passengers in the system.  The analysis is then easily performed.  Part 2 - Construction of the Model The construction of the model w i l l be divided into two parts. An outline of each part w i l l be given below and then discussed in detail. (a)  General information w i l l be discussed concerning the  nature of the model, the prime elements of the computer language, and the model structure. (b)  In Chapter II, the systems identification was divided  into three areas - the arrival stream, the check-in process and the output.  Each area had certain s t a t i s t i c a l or quantitative  definitions.  This section w i l l convert those definitions into operational elements of the model. (a) General Information The computer language used to code the model was GPSS/360. During the late spring of 1970, the University of British Columbia Computing Centre obtained a new compiler called GPSS-V: OS Version. A few changes have been made in the compiler to enable the undertaking of extremely large and complex simulation models.  The computer  facilities  at U.B.C. consist of an IBM 360/67 Duplex u t i l i z i n g MTS (Michigan Terminal System).  The general principles of the GPSS/360 program have been extracted from the User's Manual and appear below. Block diagrams or flow diagrams are widely used to describe the structure of systems. They consist of a series of blocks, each of which describes some step in the action of the system. Lines which join the blocks indicate the flow of traffic through the system, or describe the sequence of events to be carried out. Alternative courses of action that arise in the system are represented by having more than one line leaving a block. Conversely, one block may have several lines entering i t to represent the fact that this block is a common step in two or more sequences of events. The choice of path, where an alternative is offered, may be a probabilistic event or a logical choice, depending upon the state of the system at the time of the choice. Both of these methods of selection can be used in the GPSS/ 360 program. The units of traffic that move through the system depend upon the system being simulated. Units might be messages in a communication system, electrical pulses in a d i g i t a l c i r c u i t , work items in a production line, or any number of other units. These units upon which the system operates in the GPSS/360 program w i l l be called "transactions." The GPSS/360 program also has various other entities ( f a c i l i t i e s , storages, queues, tables, etc.) whose attributes are changed by the movement of trasactions through the various block types. Although a block diagram is a commonly used means of describing a system, the notation used in normal block diagrams depends upon the system and the person who is describing the system. For the purpose of the GPSS/360 program, certain conventions and systems concepts have been defined, each corresponding to some basic action or condition that generally occurs in systems. Statist i c a l variations may be introduced in the block diagram, and many s t a t i s t i c a l sampling procedures are provided. Levels of priority may be assigned to transactions and complex logical decisions may be made during the simulation. The GPSS/360 program operates by moving transactions from block to block of the simulation model in a manner similar to the way in which the units of traffic they  represent progress i n the r e a l system. Each such movement i s an event that i s due to occur at some point i n time. The GPSS/360 program maintains a record of the times at which these events are due to occur, then proceeds by executing the events i n t h e i r correct time sequence. When transactions are blocked and cannot move at the time they should, the program moves them as soon as the blocking condition or conditions change. In order to maintain the events i n the correct time sequence, the GPSS/360 program simulates a clock that i s recording the instant of time that has been reached i n the model of the r e a l system. The number shown by this clock at any instant i s referred to as the "absolute clock time." Another clock time, the " r e l a t i v e clock time" i s one of the Standard Numerical Attributes which can be externally addressed by the analyst. A l l times i n the simul a t i o n model are given as i n t e g r a l numbers. The unit of system time which i s represented by a unit change of clock time i s implied by the user, who enters a l l data r e l a t i n g to times i n terms of the time unit he has selected. Whatever unit of time i s chosen, such as m i l l i s e c o n d or tenth of an hour, i t must be used consistently throughout a simulation model. The GPSS/360 program does not simulate the system at each successive i n t e r v a l of time. Instead, i t updates the absolute clock to the time at which the next . most imminent event i s to occur. The c o n t r o l l i n g factor i n the amount of computing time that i s used by the program i s , therefore, the number of events to be simulated, not the length of real-system time over which the simulation i s being made.^  The time units of the Check-in model are seconds.  This unit  was used because of the nature of the s e r v i c e time d i s t r i b u t i o n and length of simulated time.  Detailed flowcharts of the system appear  i n Appendix I I .  IBM User's Manual GPSS/360, #GH20-0362-3, IBM Corporation Technical Publication Dept., 1970, pp. 4-5.  (b) Conversion to GPSS The Arrival Stream.  As stated in Chapter II, the arrival stream  of the Passenger Check-in system may be defined statistically by: 1.  The number of passengers who enter the system in a certain time interval;  2.  The type and group composition distribution for a particular airport;  3.  The ratio of Revenue passengers to total arrivals in a certain time interval;  4.  The operating rules in effect at the airport and the percentage of passenger groups who are effected.  Each of these quantitative definitions w i l l be discussed and illustrated. Each departing flight has an associated arrival pattern of its passengers.  The distribution defining this pattern is the percentage  of the load factor (L.F.) arriving in a particular period relative to time of departure.  The '% of L.F.' was observed to remain approximately  constant even though the L.F. changed.  If the 15-minute time intervals  prior to departure are defined as t ^ / i = 1 ,  6; (t^ = 0 - 1 5  minutes  prior to departure) then the % of L.F. arriving in t^ for flight X may be defined. If flight X has a capacity of 198 passengers and i f L.F. = 80% for a particular sample Then i f 50 passengers arrive in t^, the % of L. F. would be 50 198 x .80  x 100% = 32% of L.F.  Assume the distribution of % of L.F. for flights X and Y were: Time Period t.  1  % of L.F. X Y  t  1  10  5  t  2  20  30  t  3  35  40  t. 4  20  20  t  10  5  _5 100%  _0 100%  5  t 0  If the scheduling of flights i s such that X departs 15 minutes before Y, the distributions of % of L.F. i s depicted in Figure 3.1. The load factor depends upon the time of year, day of the week, destination of the flight and time of departure. In a discrete time period of the day T^, assume the departure time of flight X i s T. and flight Y i s T, , .. J 3+1 into the system w i l l depend on t^(X) + t^  +  Then the l a t a l arrivals  ^(Y) , where t^(X) and t^(Y)  are defined in the obvious manner. If the load factor of X i s 80% and Y i s 60% and T  -  t (X) + 2  t (Y) 3  Then the arrival rate of T  w i l l be 20% of (.80 x 198) + 40%  of (.60 x 198) = 79 passengers. Thus, this variable, the total number of passengers who enter the system in a 15-minute time period T^ w i l l be the exogenous to study. The thesis w i l l not attempt to determine the expected number of arrivals at the Vancouver airport but justifies how this variable may be obtained.  % of L.F. 40 30 20 10  35 20 "  5  % of L.F.  10  10 t  t  5  t  4  40 30 20 10  t±  2  30  Flight Y L.F. = 60%  20  's  4  fc  j  T A  I  t  3  40  h  t. = 15 minutes  Flight X L.F. = 80%  A r r i v a l s of  fc  3  |  i - 1  2  T i  = t ^ of Y + t  2  ' l  fc  i  T"  !  l + l  of X  T. = 15 minutes l  Figure 3.1 D i s t r i b u t i o n of Load Factors  T  i  ± + 2  X  The next s t a t i s t i c a l definition of the arrival stream is the distribution of the group type and composition. The group type distribution is used to distinguish between independent and dependent groups.  Independent groups may be groups of size 1  (a person travels alone) or a group wherein the members are processed independently at the check-in counters.  The distribution appears in  Table 3.1. 2  TABLE 3.1 GROUP TYPE DISTRIBUTION VANCOUVER AIRPORT, 1967 %  Group Type  Cumulative %  1.  Independent - Alone Others  16.0 13.5  16.0 29.5  2.  Dependent  70.5  100.0  - Family  It should be noted that 70.5% of groups departing on A i r Canada flights were family units. Table 3.2 is the distribution of group size.  3  A. T. Wiley, Director, Marketing Intelligence, Air Canada. A study was performed in 1967 to determine group composition and size for Vancouver Airport. A. T. Wiley,  ibid.  GROUP SIZE DISTRIBUTION VANCOUVER AIRPORT, 1967  Group Size  %  Cumulative %  1  16.0  16.0  2  34.9  50.9  3  15.4  66.3  4  17.2  83.5  5  8.9  92.4  6  7.6  100.0  The next step i s to c a l c u l a t e the a r r i v a l rate.  As stated  previously, since the groups a r r i v e independently of one another, the a r r i v a l rate may be approximated by a Poisson d i s t r i b u t i o n . If the mean number of passengers a r r i v i n g i s denoted by A^, the mean group a r r i v a l rate A i s A  P  AG - y  G  y« = mean group s i z e (2.9) passengers as calculated from CJ  Tables 3.1 and 3.2. The i n t e r a r r i v a l rate, or time between a r r i v a l s i s approximated by a negative exponential function and i s calculated as follows: Time Interval Mean Group A r r i v a l Rate  X Negative Exponential Function  — A  x ln (R.N.)  G  In (R.N.) = natural log of a Random number. The percentage of Revenue and Standby groups (passengers) can  1  be calculated from the flight manifests. The fourth s t a t i s t i c a l definition is derived from the allocation process in the arrival stream.  The flowchart of this process appears  in Figure 3.2. The process deals with the total group arrivals in the system and subsequent determination of Revenue or Standby groups.  As groups  flow through the system, only non-family groups may fly Standby.  Thus  70 per cent of groups arriving bypass this allocation process to Revenue or Standby systems. The arrival stream at the airport may be explicitly specified by the analyst to reflect changes in the composition of passenger arrivals or operation characteristics. The Check-in Process.  The discussion of the quantitative defini-  tions of the check-in process w i l l centre on: 1.  The allocation process in the Revenue and Standby systems;  2.  The f a c i l i t i e s and their characteristics;  3.  The queueing behaviour at the f a c i l i t i e s . The allocation process within the Revenue and Standby systems  is graphically displayed in Figure 3.3. be added to the description.  A quantitative dimension w i l l  The allocation process within the two  systems acts upon either groups or passenger units.  The percentage speci-  fications are exogenous to the system and must be defined as follows.  GROUP ARRIVALS  S  Family Bypass  GROUP A >v FAMILY 1^-»T  *•»  r\ a.  No = 30% of a l l Groups XH30 = % Standby  70% of Total Group Arrivals  RPS  =  Revenue Passenger System  SPS  =  Standby Passenger System  Figure 3.2 Arrival Stream Specification  XH = Halfword Save Value to Specify Percentage (Yes answers) in Parts per Thousand. XH31, 150 or 15% of Revenue Groups Arrive Non-ticketed  Figure 3.3 - Check-in Process Specifications  * % of groups using curbside check-in (XH33) % of passenger units with 'excess baggage (XH34) STANDBY SYSTEM % of groups arriving non-ticketed (XH32) % of passenger units with excess baggage (XH35) *  (XH31) is the name of the variable for allocation process.  **  Note:  groups that arrive non-ticketed are not allowed to use the curbside check-in services.  These percentages may be the averages for a l l flights in a certain period of the year or by the month.  The averages do fluctuate  because more excess baggage is incurred in the winter months due to the weight of winter clothes.  A l l specifications except curbside check-in  are percentages of the total.  If curbside users are defined as a per-  centage of the total, allowance must be made for the fact that nonticketed passenger groups are not allowed to use curbside check-in. The f a c i l i t i e s are the baggage and ticket agents'check-in counters. The number of f a c i l i t i e s that are made available for passenger service is given by the f a c i l i t y policy. is a 15-minute interval.  The staffing horizon used by Air Canada  This policy must be specified in the model,  section 1.1, by the following method. XH1 = number of Revenue baggage counters open XH2 = 16 - the number of Revenue ticket counters open XH3 = 16 + the number of the Standby baggage counters open.  only one is operational at this point in time. The service times of the f a c i l i t i e s may be approximated by the  4 Erlang distribution (K = 3 or 4).  At the baggage counter, the mean  service time was dependent on the size of the passenger unit and whether or not the passenger unit had excess baggage.  In the model, exponential  service times were used to impute a conservative element into times.  the service  The mean service time was dependent upon the size of the group  and i f excess baggage was encountered, the mean increased by 30 seconds. Table 3.3 reveals the mean service times by groups. At the ticket counter, the service times were dependent upon the type of service and not the group size.  Data collection problems were  encountered when trying to identify the passenger unit size.  Therefore  two probability functions were formulated - one for ticket purchases, the other for payment of excess baggage.  The distributions are in  Figure 3.4. TABLE 3.3 MEAN SERVICE TIMES FOR BAGGAGE COUNTERS Group Size  Mean Service Time (seconds)  1 2 3 4 5 6  V. K. Wozniuk,  45 60 90 125 160 180  ibid.  The queueing discipline at the check-in was on a first-come, first-served basis.  When passenger groups arrived at the counters,  the shortest queue was selected as a group.  If groups ' s p l i t ' into  passenger units, those passenger units would follow the same queueing behaviour.  This behaviour was programmed into the model as follows. Let  = current length of queue i in passenger units.  Let BV. = a boolean variable such that BV. = 1 1 l i f the f a c i l i t y ' i ' is in use, and '0' i f i t is not. Let the number of counters open be n. Therefore as groups arrive, the queue size (L^) is L. = Q + BVJ . l i i J  L^ MIN. is selected as the queue for the group starting from LI. (If LI = L2 = MIN then LI is selected; also, i f Q1 = Q2 = 0 and BV± = 1 and BV2 = 0, then L2 is selected.) The Output.  The output unit of the Passenger Check-in system  was stated as being a passenger unit which has completed check-in. Associated with this flow element is a group of statistics which describe the state of the system.  This set may be used to compare the  effects of various policies used in the system. The f i r s t statistic is the average transit time per passenger unit.  The transit time of a particular passenger unit is the summation  of  its waiting times in a l l queues and the service times encountered  while checking i n .  No consideration has been given to walking time  to and from the various counters because of the sampling problems i n volved and the lack of useful information that would be derived. The  next statistic deals with the f a c i l i t i e s .  The time that an  agent is engaged in service as a ratio to total time available is the utilization of that f a c i l i t y .  If the average utilization of a set of  f a c i l i t i e s is high and service standards maintained, the cost of service is low because the manpower of the system is being used.  More w i l l be  said about this in the analytical experiments. The  statistics which indicate the service standard is the per-  centage of passenger units who wait in lines longer than 'x' minutes. Air  Canada has stated that 85% of the Revenue passenger units must not  exceed 2.5 minutes waiting time in any one queue; 75% of the Standby passengers would be a desirable amount to meet this same standard. In  summary, then, the state of the system may be defined as: 1.  Transit time per passenger unit;  2.  Utilization of f a c i l i t i e s ;  3.  % of passenger- units exceeding 2.5 minutes waiting time in queues.  Part 3 - Verification To simulate, is to duplicate the essence of the system or activity without actually attaining reality itself.r  in  George W.  Morgenthaler, "Theory and Application of Simulation  Operations Research," in Progress  in Operations  Research,  Vol. I,  Russell L. Ackoff, ed. (New York: John Wiley & Sons, Inc., 1961), pp. 367, cited in Meier, Robert C , Newell, William T . , Pazer, Harold, Simulation in Business and Economics, (Englewood C l i f f s , New Jersey, Prentice-Hall Inc., 1969), p. 2.  The verification that a computer based model i s , in fact, "duplicating the essence" of the system is of extreme importance. The requirement that a complete validation of the simulation model take place, is that the results of experimentation w i l l be used as inputs for the decision-making process that controls the real system. Thus, verification must proceed the utilization of the model. The methodology for verification consists of four sequential steps.  The result of each step w i l l be the foundation or assumptions  of the next step.  The four steps are described below and w i l l be pre-  sented in turn. Step 1:  Logic verification.  The logic of-the model w i l l be  verified to ensure that i t represents the system correctly.  This is a  fundamental step in the verification since further verification is based on the assumption that the logic has been modelled. Step 2:  Several runs w i l l be made to determine the length of  simulated time necessary for the model to become stable, and thus, representative. Step 3:  Since the logic and the time necessary to stabilize  the system are known, a comparison of observed data with simulated results w i l l be made.  This is a crucial step since i t w i l l determine  i f the system has in fact been simulated. Step 4:  The primary objective of Step 4 is to determine the  feasibility of using independently determined f a c i l i t y (15 minutes) policies for planning over a longer range.  Before this can be achieved,  optimal f a c i l i t y policies must be known for given arrival rates.  Step 1.  As previously stated, the logic of the system must coincide  with that of the simulation model.  The operating rules of the system  were verified by the personnel who work in and control the system. Flow charts of the system similar to those appearing in Appendix II were shown to the agents.  Thus, i t may be stated that the basic model  represents the logic of Air Canada's present Check-in Systems at the Vancouver International Airport, Richmond, B. C. The next step in the logic verification is testing the model to ensure that the parameter specification properly allocates passengers to the subsystems.  The parameters used for verification steps 1, 2  and 4 are as follows: Revenue Stream: 85% of total Passengers are Revenue 15% of Revenue groups arrived at the Airport without a ticket 10% of Revenue groups used curbside check-in. Standby System: 15% of total passengers 15% of Standby groups arrived without a ticket 15% of Standby passenger units have excess baggage. The arrival and group composition were as previously defined for Vancouver. The model reacted as prescribed and allocated the passenger groups or units to the various sub-systems.  An indication of the accuracy of this  allocation process is presented in Table 3.4.  It should be noted that  close approximations are acceptable due to the random assignment of passengers and the length of time simulated.  TABLE 3.4 PARAMETER ALLOCATION  Specified  Parameter % Revenue Passengers  Allocation  Simulated  85.0%  86.3%  15.0% 10.0%  14.2% 10.3%  15.0%  15.1%  15.0%  16.0%  15.0%  14.0%  % Revenue groups: - non-ticketed - using curbside  r  % Revenue passenger units - having excess baggage % Standby groups - non-ticketed Standby passenger units - excess baggage  Simulated time was 120 minutes with an arrival rate of 100 passengers per 15-minute period. The percentage "specified" are the parameter values that should be approximated by the column marked "simulated" - the resultant statistics obtained from the output. The essence of this step is to confirm that the logic and allocation of passengers in the system is taking place within reasonable limits.  Step 2.  The objective of Step 2 is to determine the length of simulated  time necessary for the system to stabilize or reach a "steady state." The steady state of a simulated system is reached when fluctuations in system statistics from one time period to the next become nominal. Only when the system is at steady state are the statistics representative. Stochastic variation of a simulated system is the variation in a s t a t i s t i c at one particular point in time resulting from natural variations in system flows under normal operating conditions. Thus, when the perturbations associated with bringing a system from an "unloaded" state to normal operating levels (people in queues, etc.) are reduced and the statistics converge to an acceptable range about a single value, the system that is being simulated has stabilized. Representative statistics of the system's state may now be obtained. The length of simulated time to steady state w i l l be utilized in Steps 3 and 4 when representative statistics of the system must be known for verification. The method used to determine the steady state occurrence of a A^/P^ (an arrival rate and the associated f a c i l i t y policy P^) was as follows: 1.  For a given date ( A . ) and an associated f a c i l i t y policy (P.), the system was simulated for four hours.  2.  At intervals of 60 minutes, the statistics were collected.  3.  Steps 1 and 2 were repeated twice using new random number seeds.  4.  A different A . / P . was selected and Steps 1, 2 and 3 were repeated.  The statements in the GPSS program used in verification appear in Appendix IV. The results were plotted and appear in Figures 3.5 and 3.6. The average transit time in the system per Revenue passenger unit and the Revenue baggage counter utilization were selected to provide a succinct view of the system as simulated time progresses.  These statistics reflect  the parameter specification and system performance.  The stochastic  variation tends to smooth over time, as expected. In summation then: 1.  The steady state of the computer model is reached in the 180  to 240 minute range of simulated time.  In this range, the statistics  are representative of the state of the system.  Thus for future v a l i -  dation and experimentation, the statistics should be collected after 180 minutes of simulated time. 2.  If Figures 3.5 and 3.6 are compared, i t is noted that when the  arrival rate is high (X = 100) the statistics tend to smooth more quickly than at lower arrival rates (X = 50).  There is probably an inverse  relationship (non-linear) between the volume of arrivals and the length of time to steady state. Step 3.  Since the optimal time to simulate the system is known, the next  step in the verification is to determine i f the model does in fact simulate the real system.  This step is of great importance because further analysis  of the system is based on the assumption that the model does 'simulate' the system.  REVENUE PASSENGER TRANSIT TIME  2.8 MIN.  2.6 2.4 2.0.  60  AVERAGE REVENUE BAGGAGE COUNTER UTILIZATION  55 %  50 UTILIZATION 45 40  60  120  Figure 3.5 - Search for Steady State  1W  240"  Simulate Minutes A = 100 Passenger/15 Minutes  3.0  4  2.8 2.6  REVENUE PASSENGER TRANSIT TIME .  2.4 2.0  -  40  AVERAGE REVENUE BAGGAGE COUNTER UTILIZATION  35 UTILIZATION  30 25 20  60  120  Figure 3.6 - Search for Steady State  180  2%0  Simulated Time X = 50 Passengers/15 Minute Period  It should be noted that the human systems are less predictable than mechanical systems.  This is due to the fact that humans control  their movements in the system.  A mechanical sys em is very ordered  because the machine speeds and movements of parts have a very small variance in their descriptive mean times.  When simulated, the results  more closely approximate or predict the state of the real ystem. The methodology of this step is as follows. 1.  Observe the system and obtain sample statistics which reflect the state of the system.  2.  The \ ^ and that was observed in the real system were used as parameters in a simulation. Simulated system statistics were then calculated. Various random number seeds were used to obtain averages.  3.  A comparison of observed results with that of the simulated statistics was made.  It should be noted that observed statistics have been used in the simulation model.  First there are the service times.  These were  collected by observations and the mean times varied directly (non-linear) with the number of people per passenger unit.  The type and group com-  positions were obtained from Air Canada and are representative of the average arrival stream.  Thus any or part of the variation in the simu-  lated results from that of the observed might stem from the fact that the type and group composition of the sampled time interval was not the same as the data obtained from Air Canada (Tables 3.1 and 3.2). Another source of variation could be the use of a mean arrival rate. In the real system, the arrival rate changes continuously.  The model,  however, uses a constant average arrival rate for a given 15-minute  The validation run was constructed as follows: 1.  The arrival rate at the Revenue and Standby baggage  counters was used.  This was done because  1) curbside check-in does  not affect the counter operations; and because  2) a l l non-ticketed groups  must proceed to the baggage counters via the ticket counters.  Thus the  total arrivals would be the sum of the arrivals at the baggage counters plus the number of passengers using curbside check-in, or in this case, X = 58 passengers per 15 minutes 2.  The f a c i l i t y policy that was in operation was: Revenue baggage ticket  4 3  Standby baggage ticket  1 1  The comparison of the observed arrival stream to the simulated appears in Table 3.5.  The variation comes from the fact that the model  produced 2.8 people per Revenue passenger unit while the observed was 2.6 people.  However there were the same number of Revenue passenger  units in the system. more people.  Thus the simulated system handled eight per cent  Therefore, the system statistics should be slightly higher  than the observed.  Statistics  Observed  Simulated  Total number at baggage counters Number of Revenue passengers Standby passengers Revenue as a per cent of total Number of Revenue passenger units Standby passenger units Passengers per passenger unit (Revenue)  58  64 (+8%)  46 12  51 13  79%  80%  18 12  18 13  2.6  2.8  (+8%)  The data collection in the stimulation for the system statistics occurs as follows. At intervals of 60 seconds, the number of passenger units in the queues was counted.  Thus, the average as well as the maximum number  of passenger units in queues at the Revenue and Standby counters may be calculated.  At the same time, i f the f a c i l i t y was in use, i t was  counted as ' 1 ' , i f i t was not in use, then a ' 0 ' .  Therefore the approxi-  mate average utilization for a set of f a c i l i t i e s would be the sum of the ' l ' s divided by the number of observations.  It is felt that these  three statistics should determine whether or not the model does in fact simulate the real system.  The comparison is given in Table 3.6.  TABLE 3.6 COMPARISON OF SYSTEM STATISTICS  Observed  Simulated  Revenue baggage ticket  4 2  4 3  Standby baggage ticket  5 3  3 5  Revenue baggage ticket  2.1 1.8  2.6 2.3  Standby baggage ticket  4.4 2.0  3.7 2.5  Statistics Maximum number of passenger units in queues  2.  3.  Average number of passenger units in queues  F a c i l i t y utilization Revenue baggage ticket  47% 40%  48% 39%  Standby baggage ticket  73% 33%  65% 34%  Variation could be caused from method of data collection.  Thus, i t may be concluded that the simulation model as  constructed,  does in fact, represent the system at the Vancouver Airport. Step 4.  The fourth step in verification of the model is to confirm that  independently determined optimal f a c i l i t y policies w i l l be compatible  from one period to the next. large and A  ^ of period t + 1 is small, the f a c i l i t y policy determined  +  by simulation for A A  When the arrival rate A of period t is  +  ^ may no longer be optimal.  The magnitude of  may be such that passengers are left in the system and thus w i l l  affect the performance of the system in period t + 1. Thus, the methodology for this step is as follows: 1.  Determine an optimal ^ / P ^  2.  Set A./P. at the optimal level and simulate until a steady state is reached. This signals the end of the f i r s t time period (t).  3.  The passengers in the system (queues and f a c i l i ties) w i l l remain, but a l l other system statistics w i l l be set to zero at the start of t + 1.  4.  A./P. w i l l be changed to A_/P„ and simulated for 15 minutes. The statistics w i l l be collected at this point in time.  r o r  ^ = 100, 50, 25.  The results were as follows: In the determination of a f a c i l i t y policy for each arrival rate, the passengers must be processed within acceptable norms of system performance; namely, 85% of the Revenue passenger units must not wait in queues longer than 2.5 minutes; 75% of Standby passengers must meet the same conditions. Thus, by a search technique of setting a A_^ and varying the P^, the following policies were obtained: X 1 = 100  = 5, 3/2, 1  A2 =  50  P 2 = 4, 2/1, 1  A3 =  25  P 3 = 2, 1/1, 1  The notation of  = X^, X 2 / X 3 > X^ is defined as:  X^ = no. of Revenue baggage counters open; X 2 = no. of Revenue ticket counters open; X^ = no. of Standby baggage counters open; X^ = no. of Standby ticket counters open. There is an approximate P^ for each A^ that achieves the service standard or norms of system performance.  Steady State statistics for  A^/P^ (optimal) appear in Table 3.7 and w i l l be used in the remainder of Step 4.  TABLE 3.7 AVERAGES OF SYSTEM STATISTICS Statistic  2.  X. = 100  50  25  Transit Time: Per Revenue Passenger Unit Passenger  2.6 min. 2.7  2.4 2.5  2.8 3.1  Standby Passenger Unit (in minutes)  2.9  3.1  2.1  F a c i l i t y Utilization (%) Revenue baggage counters ticket counters  57.5% 48.2%  30.7 30.0  30.0 20.9  Standby baggage counters ticket counters  31.2 49.9  32.4 18.6  16.6 10.2  20 sec. 30  20 sec. 20  30 sec, 10  80 6  50 20  Average time in queues per passenger unit (nearest 10 sec.) Revenue baggage counters ticket counters Standby baggage counters ticket counters 4.  30 170  % of Passenger Units In Queues Longer than 2.5 min. at: Revenue baggage counters ticket counters  5.2% 6.4  Standby baggage counters ticket counters  -°* 30.0 6  3.2% 4.1 23.0 16.7  6.5% 0 13.5 9.0  The change-over from  """n  Run 1  changed to  Run 2  X  l = 100  P  l  A  l = 100  P  Run 3  X  P  =  fc t o  ^2^2 ^ ^  5, 3/2, 1 changed to  2 = 4, 2/1, 1  +  ^  W a S  a  S  2  =  50  2  =  4, 2/1, 1  X  P  2 = 5, 3/2, 1 l = 50  n  2 = 25  X  • P 2 = 2, 1/1, 1 changed to  2  =  25  2  =  2, 1/1, 1  A  P  The s t a t i s t i c a l results are presented in Table 3.8.  As could be expected,  the average transit time for the second period increased compared to the transit times when the policies were evaluated individually.  This  is due to the number of people remaining in the system from the previous period.  The percentage of passenger units waiting in the queue at the  Standby ticket counters was greater than the acceptable mean standard of 25% (A = 100 -> 25, 29.6% and A = 100 -> 50, 28.2%).  However, the  stated service policies derived in the earlier part of Step 4 have remained optimal, since  1) Air Canada does not have the capacity to increase  Standby Ticket purchase counters to two positions; and gers are of lesser importance than Revenue passengers.  2) Standby passenThe degree of  change in optimality of the t + 1 period policies depends also on the amount of slack existing at the previously optimal steady state policy.  RUN I:  A = 100  50  Statistic  Averages Steady State  1.  2.  Transit time (minutes) per:  A = 50  Change-over A = 100 -* 50  Revenue passenger unit  2.4  2.6  Revenue passenger  2.5  2.7  3.1  3.2  3.2%  3.5%  4.1%  14.0%  23.0%  4.8%  16.7%  28.2%  Standby passenger unit % of Passenger units in Queues longer than 2.5 minutes at: Revenue Baggage counter Ticket counter Standby Baggage counter Ticket counter  RUN 2: A = 100 -> 25 RUN 3: A =  50 -»• 25 A = 25  1.  2.  A = 100 -> 25  A = 50 -*• 25  Transit time (minutes) per: Revenue passenger unit  2.8  3.4  3.2  Revenue passenger  3.1  3.5  3.3  Standby passenger unit % of Passenger units in Queues longer than 2.5 minutes at:  2.4  3.0  3.3  Revenue Baggage counter  6.5  6.6  5.0  2.3  15.0  4.5  13.5  18.5  22.5  9.0  29.6  23.3  Ticket counter Standby Baggage counter Ticket counter  In summation to the verification of the computer model:  C.  1.  The representation of the system has been captured and computerized;  2.  The results of further experimentation w i l l be plausible since the basic system has been simulated correctly;  3.  Steady state conditions occur within approximately 180 to 240 minutes of simulated time;  4.  Optimal f a c i l i t y policies determined independently are compatible with one another.  Analysis of the System In summary, the system approach has been applied to the Passenger  Check-in system: 1.  The system was identified as to its inputs, conversion process and outputs in Chapter II;  2.  The system was defined quantitatively modelled and verified thus far in Chapter III.  Thus what remains is the analysis of the system.  There are two  objectives of this analysis: 1.  To provide a tool so that experimentation may take place;  2.  To be able to formulate and evaluate policies.  Since the model has been developed for experimental policy evaluation, the f i r s t objective is deemed completed.  The capabilities of the  model are vast. For instance, due to the method of construction, very slight changes in the program would allow the model to simulate a 24 hour period with fluctuating (continuous) arrival rates and f a c i l i t y policies.  More specifically, the model was designed for the evaluation of policies (service, f a c i l i t y and operating) prior to implementation. There is no need to implement new policies in the real system in order to determine their v i a b i l i t y ; the model w i l l provide a faster and less costly evaluation of these new policies. The remainder of the systems analysis w i l l consist of three steps.  They are briefly described below and w i l l then be presented in  their entirety. 1.  To determine the exact behaviour of the system and its impli-  cations on policy formulation.  This step w i l l , in effect,  allow policies  to be formulated in a more rational manner. 2.  To determine the f a c i l i t y policy that should be implemented  to achieve the required service standard at various arrival rates. The maximum capacity of the f a c i l i t i e s w i l l also be determined. 3.  To formulate alternative operating policies and to determine  i f they are viable in the system.  This w i l l be a multi-step process.  As in the verification, there are certain exogenous variables that must be specified.  In the experimental analysis, a l l steps w i l l  have the same set of variables.  They are:  Revenue System: 85% of total passengers 15% of Revenue groups arrive without tickets 10% of Revenue passenger groups use curbside check-in 15% of Revenue passenger units have excess baggage.  Standby System:  15% of Standby passenger units have excess baggage. Step 1.  This part of the analysis w i l l explore the behaviour of the  system to determine i f the service policy has been properly formulated. The service policy is as follows: To allow no more than 15% of Revenue passenger units to wait in a queue more than 2.5 minutes; to allow an average of 25% of the Standby passenger units to wait more than 2.5 minutes. The main elements which determine the state of the system are the arrival rate, X^, and the f a c i l i t y policy, P^.  The behaviour or  state of the system is reflected by three statistics:  the average transit  time per passenger unit, the average utilizatization of the various f a c i l i t i e s and the percentage of passenger units waiting in a queue longer than 2.5 minutes.  If the f a c i l i t y policy (P^) is held constant,  denoted as P^, then the effect of the arrival rate on the subsequent behaviour of the system may be determined. The methodology of this experiment is as follows. 1.  Set P^ = 4,2/1,1 and simulate with \ ^ = 40 until a steady  state is reached. 2.  Collect the system statistics and repeat step 1 using a  new random number seed. 3.  Repeat steps 1 and 2 until X^ = 160 using increments of 20.  The s t a t i s t i c a l results have been tabulated and appear in Table 3.9.  The selection of the Revenue Baggage Ticket System statistics for  presentation w i l l be sufficient to reveal the justification of the service policy. The relationship between the arrival rate and the average transit time per passenger unit when the f a c i l i t y policy is held constant, may be seen in Figure 3.7. When  The interpretation of this curve is as follows:  is in operation and X_^ is in the range of 40 to 90  passengers per 15 minutes, the average transit time is between 2 and 3 minutes.  The system is thus relatively stable or insensitive to the  number of passengers in the system. When zontal.  is greater than 90 passengers the curve becomes hori-  This implies that the system becomes unstable and is  to the number of passengers in the system.  sensitive  For example when X ^ = 100  goes to X ^ = 120 the transit time goes from 3.3 to 6.5 minutes per passen ger unit.  This is a 20 per cent increase in volume and results in 100  per cent increase in transit time. Thus, i f a curve was derived for each P. at the various arrival I  rates, the vertical portion of the curve would indicate the range of X ^ such that P^ produces stability in the system. Figure 3.8 is a graph of the average utilization of the Revenue baggage counters as X ^ varies.  Note that the curve is almost linear in  the range of X = 40 to X = 120. becomes vertical in nature. would approach 100%.  When X ^ increases past 120, the curve  At a very high arrival rate, utilization  Also from the previous figure, there would be an  extremely long transit time.  Thus, the average utilization for the  Revenue baggage counters w i l l be in the range of 30 to 65% when 40 _< X _< and P = 4, 2/1,1.  SYSTEM STATISTICS FROM ANALYSIS: STEP 1 REVENUE PASSENGER SYSTEM  Arrival Rate  A.1 =  Transit Time Per Passenger Unit (Min.)  Revenue Passenger Baggage System Counters: Average Utilization  % of Passenger . Units exceeding 2.5 minutes waiting time  Revenue Passenger Ticket System Counters Average Utilization  % Passenger Units exceeding 2.5 min. waiting time  40  2.3  25.%  2.3%  26.4%  4.2%  60  2.5  41.6  3.0  42.1  9.2  80  2.6  54.0  6.6  50.0  5.5  100  3.3  66.4  13.4  62.5  16.1  120  6.5  87.3  49.2  68.5  20.8  140  7.6  93.1  59.5  91.0  52.9  160  12.4  94.3  73.0  92.6  61.1  NO  0  10  20  3(!>  40  50  60  70  §5  90  100  Average Utilization in % Revenue Baggage Counter Figure 3.8 - Arrival Rate and Corresponding Utilization  If the average service time is 1 minute and average time in the system is 3 minutes, then the average waiting time in queues would be 2 minutes. Figure 3.9 is a plot of the annual rates versus the percentage of Revenue passenger units waiting longer than 2.5 minutes at the baggage counter.  This curve has a vertical and horizontal section.  When X. l  is in the range of 40 to 100 passengers, the Revenue baggage counter is able to process the passengers at such a rate that the 15 per cent rule is not violated.  In the range above X = 100, the curve becomes  horizontal in nature.  The change-over occurs when approximately 13%  to 15% of the passenger units exceed 2.5 minutes of waiting time. This curve then justifies the current service policy.  If the  percentage is increased, the system becomes very sensitive to the amount of passengers in the system for  (at X = 120 and P^, 49% of Revenue  passenger units exceed 2.5 minutes).  The vertical portion of the curve  represents the range within which P^ produces stability in the system. In Figure 3.10 the utilization and corresponding percentage of passenger units exceeding 2.5 minutes waiting time has been graphed (Figure 3.8 + Figure 3.9). For policy formulation, the effect of increasing the percentage of passenger units allowed to exceed 2.5 minutes waiting time on the f a c i l i t y utilization may be determined from the graph. gives the trade-off of opposing system objectives.  Alternatively, th  If the percentage was  increased to 25%, the average utilization that could be expected would  Arrival Rate X=  160 •  140 •  1  120 •  i  ^  —  100 .  1  P^ held constant  80 .  60 .  |  Present Policy 15%  40 .  1  0  10  20  30  40  Figure 3.9 •- Arrival Rate and % Passenger Units Exceeding 2.5 Minutes  50  60  70  80 %  % passenger units exceeding 2.5 minutes waiting time  Utilization %  P. Held Constant and Arrival Rate 1 Varies  100 . |  90 .  ^  '  RPBS  80 . 70 •  |  60 50  1  Service Policy 15% of the Revenue Passenger Units are Allowed to Exceed 2.5 Minutes  1 |  40 30 . 20  Revenue Passenger Baggage System Counters = 4  1 ,  10 . 0  5  10  15  20  30  40  50  60 70 80 % Passenger Units ^_ 2.5 Minutes Waiting Time in Queues  Figure 3.10 - Average Utilization and % Passenger Units > 2.5 Min. Waiting  Thus, we have developed in this step a methodology for determining the service policy.  Once a change has occurred in the system,  this step should be repeated to determine i f the service policy should be altered. Step 2.  The second step is to determine the f a c i l i t y policy that should  be implemented into the system to achieve the stated service policy. As part of this step, the maximum capacity of the system w i l l be determined.  Although very high arrival rates may not be encountered at the  present time, they could be forecast for a few years hence.  Thus, the  result of this step could be an input for planning Air Canada's operations in the future. The methodology of this step was as follows. 1.  Set X.. l  2.  Vary the P^ such that the service policy is achieved.  Simulate for 180 minutes.  3.  Increase X_^ by 25 passengers and search for a new P.. l  4.  Repeat steps 2 and 3 until the system can no longer achieve the service policy.  It should be noted that when X. is increased, the number of l f a c i l i t i e s open in the new P^ w i l l be equal to or greater than the previous policy.  This in effect limits the search to only increases  in P i . The results of the analysis appear in Table 3.10 and Table 3.11.  FACILITY POLICY FOR X. l  Revenue Baggage (X ) ][  Standby Ticket ( X ) 2  Baggage (X^)  Ticket ( X ) 4  EX.  l  25  2  1  1  1  5  50  4  2  1  1  8  75  4  2  2  1 r  i  9  100  5  3  2  1  11  125  6  4  2  1  13  150  7  5  2  1  15  175  8  5  1  16  200  9  1  17  i  5  i  2  i  2 i  225  ,  Service policy unattainable due to limited number of f a c i l i t i e s  Passenger Composition  85% Revenue Passengers  15% Excess Baggage  10% Use Curbside  15% Non-ticketed  TABLE 3.11 FACILITY POLICIES AND SYSTEM STATISTICS  Baggage U%  Revenue  Ticket U% %>2.5  T. T.  Baggage U% X 3  4  Ticket U%  0  2.1  1  16.6  13.5  1  10.2  9.0  30.0  4.1  3.1  1  32.4  23.0  1  18.6  16.7  2  43.8  4.2  3.3  2  25.6  4.0  1  42.1  15.0  5.2  3  48.2  6.4  2.9  2  31.2  6.0  1  49.9  30.0  62.4  4.7  4  53.3  11.9  3.2  2  51.0  17.3  1  55.7  32.4  7  64.5  6.5  5  48.9  3.5  3.3  2  58.2  22.3  1  56.1  32.8  2.9  9  57.4  4.3  5  58.0  4.4  3.6  2  62.5  20.8  1  73.1  36.8  3.0  9  63.6  6.6  5  63.7  10.5  5.5  2  74.4  37.2  1  76.5  49.0  T. T.  X  25  2.8  2  30.0  6.5  1  20.9  50  2.4  4  30.7  3.2  2  75  2.9  4  50.3  9.8  100  2.6  5  57.5  125  2.8  6  150  2.7  175 200  X. X  =  l  %>2.5  Standby  X  2  T. T. = Transit Time X  %>2.5 = % of Passenger Units exceeding 2.5 Minutes Waiting Time in Queues  X  %>2.5  l = No. of Revenue Baggage Counters Open  X  U% = Utilization of X.  %>2.5  X  2  = No. of Revenue Ticket Counters Open = No. of Standby Baggage Counters Open  X  3  X  4 = No. of Standby Ticket  Counters Open oo o  Table 3.10 gives the f a c i l i t y policy  = x^/x^,x^ that should  be implemented into the system i f the service standard is to be maintained. It is also representative of the capacity of the system.  In each column  there is a vertical line; this indicates the capacity of the system which results from the limited number of counters available.  At X = 100,  our f i r s t partial capacity is reached at the Standby ticket counter. When X = 175, the Standby Baggage counter just remains optimal but care must be used since passengers experience some delay at the ticket counter. The capacity of the Revenue counters is met at X = 200 per 15-minute period. The total agent requirements are also listed, and is computed as the E X . in the P. notation (P. = X n , X „ / X 0 , X . ) . l l l 1 2 3 4 The application of this step is very useful indeed.  The result  of Step 4 allows the use of independently determined f a c i l i t y policies to be used (15-minute horizon) for daily planning.  Suppose the X_^ for  the average 15-minute period t_^ was as follows:  l  =  2  =  t  fc  x  ioo 50  X  3 = X75 Then the P. for each t, would be as follows: l 1 fc  t ^ P 1 = 5, 3/2, 1  ZX  t 2 = P 2 = 4, 2/1,1  ZX  t 3 = P 3 = 4, 2/2, 1  ZX± = 9  ±  ±  = 11 = 8  The 15-minute f a c i l i t y policy that w i l l achieve the stated service policy is available from the model.  The question then arises as to the hourly  f a c i l i t y policy which is complicated by unknown contractual agreements. Although this is outside the scope of the thesis, the results of this step, as contained in Table 3.11 could be used as inputs for schedule determination. Step 3.  The analysis that w i l l be undertaken in Step 3 is directed  towards the operating policy of Air Canada. A few years ago, the operating policy was changed such that the baggage operations and ticket purchases of the Revenue passengers was separated from that of the Standby passengers.  Revenue passengers  paying excess charges were allowed to use the Standby ticket counter and Standby passengers under the same circumstances were allowed use of the Revenue ticket counter. In Part A of Step 3 the impact of complete separation of Revenue and Standby ticket operations is described.  The implementation of this  policy should produce a higher utilization of f a c i l i t i e s and greater percentage of passenger units waiting longer than 2.5 minutes in queues. This is so because the Revenue passengers are limited to one fewer counters for service. The system, which was simulated for multiple 180-minute runs, produced higher system statistics than when the integrated policy was operational.  (See Table 3.12).  If the decision is made that complete  separation is necessary, the f a c i l i t y p o l i c i es as  stated in Table 3.10  should produce stability in the system for the Revenue system.  REVENUE/STANDBY TICKET OPERATIONS IN FULL SEPARATION Integrated  Revenue Transit Time per Passenger Unit Average Utilization % of Passenger Units Waiting Longer than 2.5 Minutes  Separated  2.6 Min.  2.85 Min.  28.2%  49.9%  6.4%  13.8%  Standby Transit Time Average Utilization % of Passenger Units Waiting Longer than 2.5 Minutes  3.75 Min.  2.9 Min.  62.1%  49.2%  ^30%  54.9%  k  25% + 5% would be acceptable but since there is only one f a c i l i t y , a decision must be made to either expand or revise the service policy at the Standby Ticket counter.  In Part B of Step 3 we examine the operating policy at the Revenue Ticket counter more closely.  Since the separation as in Part A, produced  higher system statistics, the queueing discipline w i l l be changed to determine i f the statistics can be lowered while maintaining separated operations. Specifically, instead of using multiple queues or one queue for each agent, a single queue w i l l be used.  A l l Revenue passengers  purchasing tickets or paying excess charges line up in one queue.  When  they reached the front of the line, they wait until any one of the agents becomes available. in Figure 3.11.  Only then do they leave the line.  This is depicted  The results of the simulations combined with the results  of Part A for comparative purposes are in Table 3.13. The transit time per passenger unit is lower and is comparable to the integrated operation.  The utilization is lower as well as the %  of Revenue passenger units exceeding 2.5 minutes of waiting time than the integrated.  Thus i f this particular method of queueing was adopted,  separation of the Revenue/Standby counter could be achieved with viable results. In Part C of this step in the analysis we examine the method of checking in Standby passengers.  Since there has been a large number of  passengers exceeding the 2 1/2 minute of waiting time, the combination of ticket and baggage operations at one counter w i l l be analyzed. one agent would become both a ticket and baggage agent.  In effect,  When the passenger  unit approaches the service counter, a single agent w i l l issue tickets, check the baggage and collect the excess charges.  Since this operation  QUEUE 1  QUEUE 2  QUEUE 3  V AGENT 1  AGENT 2  V AGENT 3  REVENUE TICKET COUNTER  MULTIPLE QUEUES  GOES TO ANY AGENT  MOVEABLE GUIDE RAILS  AGENT 1  AGENT 2  AGENT 3  REVENUE TICKET COUNTER SINGLE QUEUE Figure 3.11 - Multiple or Single Queues  COMPARATIVE STATISTICS FOR REVENUE TICKET COUNTER UNDER VARIOUS OPERATING POLICIES  Standby Excess Allowed  Revenue Ticket Transit Time per Passenger Unit Utilization % P. U. > 2.5'  2.6 Min.  Revenue Standby Separated  2.85 Min.  One Line at Revenue Ticket Counter  2.75 Min.  48.2%  49.9%  42.9%  6.4%  13.8%  4.5%  * Percentage of Revenue Passenger Units exceeding 2.5 Minutes Waiting Time  multiple operations and is not at present in effect the correct distribution of service times must be hypothesized.  It would probably be  approximated by an Erlang distribution with K = 4 or 5.  In the simu-  lation, the service times w i l l be drawn from the already existing distributions for service and baggage check.  The total service time vary  from the present baggage service time only i f there are tickets to be purchased or excess charges to be paid.  The estimated total service  time w i l l probably be high which in this case is conservative since we are testing a completely new agent procedure.  The flowchart of the  operation for a single agent is depicted in Figure 3.12.  The system  statistics are given in Table 3.14. Thus i f this operational procedure is adopted, the transit time (average) of Standby passengers remains relatively the same as when the Revenue/Standby ticket operations were integrated.  However, the  percentage of passengers who exceed 2.5 minutes waiting time is reduced to 28.3%.  Thus the service policy is maintained through the total system  at X = 100 and the aforementioned parameter specifications. These are just a few of the experiments that can be performed on the model.  The experiments presented hopefully have hit on the major  areas of interest and demonstrated the vast power of simulation as an airport management tool.  PURCHASE TICKET  <— ,  CHECK-IN BAGGAGE  - Denotes Service Time Figure 3.12 - 1 Standby Agent Performs Both Ticket and Baggage Operation  TABLE 3.14 STANDBY OPERATIONS COMPARED  Revenue/Standby Separated  Integrated  Statistic  2 Baggage Agents 1 Ticket Agent 2.9 Min.  Transit Time  2 Baggage 1 Ticket 3.75 Min.  Each Performing Baggage and Ticket Operations 3.0 Min.  Baggage 31.2%  30.3%  6.0%  5.8%  Average Utilization  49.9%  62.1%  % Passengers Exceeding 2.5 Min. Waiting Time  30.0%  54.9%  Average Utilization % Passengers Exceeding 2.5 Min. Waiting Time Ticket  A = 100 passengers over 15-minute period  Combined Utilization 44.7% % < 2.5 Min.28.3%  CHAPTER IV CONCLUSIONS In presenting the conclusions of the thesis, Chapter IV w i l l be concerned with three areas: 1.  The conclusions derived from the application of the  'systems approach' concept.  As part of this area, the strengths of  using simulation as an analytical technique w i l l be summarized. 2.  The conclusions of the system analysis of Air Canada's  Passenger Check-in System, Vancouver Airport, Richmond, British Columbia. 3. The proposal of future areas of research in the Airport system.  THE SYSTEMS APPROACH CONCEPT The 'systems approach' concept has been applied to the Passenger Check-in System for the purpose of evaluating and formulating policies to ensure the attainment of the system's objectives. The relative strength of using this methodology is that i t forces the analyst to be pragmatic.  The system must be defined e x p l i c i t l y ,  with logical and quantitative relationships carefully defined.  Once  this has been done, the sub-systems may be isolated and analyzed. The systems approach is readily applicable whenever a phenomena may be described as having an input element, a conversion or transformation  process,  and an output element.  The l e v e l o f d e t a i l o f t h e systems  a n a l y s i s i s d i c t a t e d by t h e r e q u i r e m e n t s o f the u l t i m a t e user  of the  r e s u l t s and t h e s i z e o f t h e system. The  methodology c o n t a i n e d  C h e c k - i n systems o f t h e major The  strengths  i n t h e t h e s i s i s a p p l i c a b l e t o most  airlines.  o f u s i n g s i m u l a t i o n as an a n a l y t i c a l  technique  a r e summarized below. 1. permitting  The t e c h n i q u e i s r e a d i l y a p p l i c a b l e t o a 'flow  observation 2.  system'  o f t h e dynamic b e h a v i o u r .  When t h e r e a r e no p r a c t i c a l a n a l y t i c a l approaches  t h e o r y ) a v a i l a b l e f o r complex systems, s i m u l a t i o n can be u t i l i z e d plify  (queueing to sim-  the a n a l y s i s . 3.  either before  A s i m u l a t i o n model p e r m i t s t h e e v a l u a t i o n o f a p o l i c y  or a f t e r i m p l e m e n t a t i o n s i n c e i t i s b a s i c a l l y a l a b o r a t o r y  t o o l t h a t can be used under c o n t r o l l e d c o n d i t i o n s . 4.  It facilitates  a means o f s t u d y i n g  t h e management o f a system by p r o v i d i n g  cause-effect r e l a t i o n s h i p s inherent  i n t h e system.  Thus t h e o b j e c t i v e o f t h i s study i s t o p r o v i d e management  with  a comprehensive t o o l t h a t w i l l make p o s s i b l e t h e t e s t i n g o f a l t e r n a t i v e management p o l i c i e s ,  f u l f i l l e d by p r o v i d i n g a s i m u l a t i o n model o f t h e  C h e c k - i n system.  CONCLUSIONS OF THE SYSTEMS ANALYSIS The  systems a n a l y s i s , a s . p r e s e n t e d  I I I was d i r e c t e d towards the p r o b l e m a t i c  i n t h e l a t t e r p a r t o f Chapter areas o f system management.  1.  The evaluation of a policy after i t has been implemented.  2.  The determination of a f a c i l i t y policy such that the  service policy is achieved. 3.  The formulation and evaluation of an operating policy  prior to implementation. These three analytical areas have provided a demonstration of the value of a simulation model and the merits of a systems analysis. The conclusions of each area are presented below. 1.  The service policy that a maximum of 15% of the Revenue  passenger units be allowed to exceed 2.5 minutes of waiting time has been formulated correctly. 2.  The nature of the system is such that greater u t i l i z a t i o n  of f a c i l i t i e s w i l l not be achieved by a nominal increase in the allowable percentage of passenger units exceeding 2.5 minutes. 3.  The f a c i l i t y policy and associated procedures have been  formulated so that the objectives of the system w i l l be attained. 4.  The maximum capacity of the system is exceeded when the  arrival rate is in excess of 200 passengers per 15 minute period. 5.  The use of a single queue at the Revenue ticket counter  w i l l ensure greater attainment of the system objectives than the use of multiple queues. 6.  The combination of the baggage and ticket operations  at one counter is a viable alternative in the present system.  AREAS OF FUTURE RESEARCH As a by-product of the Passenger Check-in System's analysis, there are four areas of possible research. 1.  They are:  A study c a r r i e d out to determine the capacity of the  baggage flow system. point of this system.  The a r r i v a l of passengers i s the i n i t i a l i z a t i o n Simulation  could be an e f f e c t i v e t o o l f o r this  study. 2.  A study of the a r r i v a l rates as a function of the reser-  vations at various times p r i o r to f l i g h t departure. contained  i n Reservac I I (Computerized Reservations  The information System) would be  most u s e f u l i n t h i s study. 3.  A s i m i l a r study to the one completed i n t h i s thesis  could be made of the Departure Lounge.  Various operational p o l i c i e s  could be tested to determine the most s u i t a b l e departure gate p o l i c i e s for various t r a f f i c volumes and a i r c r a f t . 4.  I f the on-line reservation system i s i n s t a l l e d at the  baggage counters,  the service times w i l l probably change.  of the simulation model contained  With the use  i n t h i s thesis, f a c i l i t y p o l i c i e s  could be determined p r i o r to this procedural change.  Gensema, W. "Passenger Handling Models," AGIFORS Proceedings. York: Printed at American Airlines, 1967. Hare, Van Court. Systems Analysis. World, 1967.  New  New York: Harecourt Brace and  IBM Application Program. "General Purpose Simulation System/360, Introducing User's Manual," GH20-0304-4, IBM Corporations Technical Publications Department, 19 70. IBM Application Program. "General Purpose Simulation System/360, User's Manual," GH20-0326-3, IBM Corporation Technical Publications Department, 1970. Khanna, K. C . , Takamori, H. AGIFORS Proceedings. 1969.  "Optimal Staffing at Airline Terminals," New York: Printed at American Airlines,  Meir, Robert C , Newell, W. T . , Pazer, H. L. Simulation in Business and Economics. Englewood C l i f f s , New Jersey: Prentice Hall Inc., 1969. Mountjoy, R. "Airport Simulation Models," AGIFORS Proceedings. Printed at American Airlines, 1969.  New York:  O'Broin, S. "Manpower Planning for Airport Operations," AGIFORS Proceedings. New York: Printed at American Airlines, 1968. Wiley, A. T. "Vancouver Passenger Market Survey." Air Canada, Montreal, 1967.  A private study,  Wozniuk, V. K. "Airport Manpower Planning Systems," AGIFORS Proceedings. New York: Printed at American Airlines, 1969.  APPENDIX I AIR CANADA FACILITIES  AIR CANADA FACILITIES VANCOUVER INTERNATIONAL AIRPORT  5 4 3 2 1 Standby ticket  Revenue Ticket  1  2  3  4  5  7  8  9  Revenue Baggage Counters  Offices  Airport Shops  6  12 3 Standby Baggage  To Departure Gates  Enter System  Revenue Standby Ticket and Excess payment  FLOWPATTERN OF PASSENGERS AT THE AIR CANADA FACILITIES  Revenue Baggage Check-in  Standby Baggage Check-in  APPENDIX II FLOWCHARTS OF THE PASSENGER CHECK-IN SYSTEM  FLOWCHARTS OF THE PASSENGER CHECK-IN SYSTEM  Start or end  A process - queue - service - set up characteristics  'Go to' or 'come from' depending on arrow. Section name and number given.  Collect statistics  SECTION 3.1  START OF SYSTEM GENERATE GROUPS  i  ASSIGN PASSENGER CHARACTERISTICS  0  SECTION 3.11 & 3.12 DETERMINATION OF GROUP SIZE & TYPE  \TYPE  >r  >  DEPENDENT FAMILY  INDEPENDENT GROUPS  SINGLE PASSENGER  SECTION 3.2 REVENUE PASSENGER SYSTEM SECTION 3.21 ALLOCATION TO REVENUE SUBSYSTEMS  SECTION 3.22  REVENUE PASSENGER BAGGAGE SYSTEM  GROUP SELECTION OF QUEUE  PASSENGER UNIT SELECTION OF QUEUE  QUEUE AGENT  YES  ADJUST SERVICE TIME  NO CHECK-IN BAGGAGE  CHECK-IN BAGGAGE  SECTION 3.23  REVENUE PASSENGER TICKET SYSTEM  *  SELECT QUEUE AS A GROUP FOR TICKET PURCHASES  1  SELECT QUEUE AS PASSENGER UNIT FOR EXCESS PAYMENT  QUEUE AGENT  SECTION 3.31  STANDBY PASSENGER SYSTEM  ALLOCATION TO STANDBY SUBSYSTEMS  YES  SECTION 3.32  STANDBY PASSENGER BAGGAGE SYSTEM  SPLIT OFF GROUP  SELECT SHORTEST QUEUE  QUEUE FOR AGENT  ©  © ADJUST SERVICE TIME  CHECK IN BAGGAGE  CHECK IN BAGGAGE  G>  QUEUE ONE LINE SPLIT GROUP  >  SELECT QUEUE FOR EXCESS PAYMENT  QUEUE AGENT  0  YES  PAY EXCESS CHARGE  SEND (3.4)  NO  PURCHASE TICKET  SPBB (3.32)  SECTION 3.4  END OF CHECK  END  IN  APPENDIX III PROGRAM LISTING AND SAMPLE OUTPUT  $RUN **  *GPS5  PAR=SIZE=C  SIMULATE  ** #* **  SECTION jj» 9|C ijs  1.1  1.2  ONE  -  J^C Sjj 5jC *J5 5}S *jc  INI TILIZATION 5,C 5jC  *jc ij? 5jc 5jS  3jC ?JC 5jC «JC «{C  FACILITY POLICY: N O . OF C O U N T E R A V A I L A B L E F O R P A S S E N G E R PROCESSING. X H 1 ,4 IN I T I A L X H 1 = NO. OF C O U N T E R S O P E N (RPBS] INITIAL XH2,14 XH2 = 1 6 - NO. OF C O U N T E R S O P E N (RPTS) INI TIAL XH3 ,K17 X H 3 = 16 + NO. OF C O U N T E R S O P E N ( SPBS ] **  ARRIVAL  INITIAL 1.3  *|C 3jC  **  XH30,670 XH31,150 XH32,150 XH33,118 XH34,150 XH35,150  •A,  TWO  ^t. •JV vO 0< ~t** v*r* Jfr N**  MINUTE  XH7 VALUES -  PARAMETER  SECTION  **  15  PER  XH7,50  INITIAL INITIAL INI TIAL INITIAL INI TIAL INITIAL  5j5 3|«  RATE  <JU> >j*.  PERIOD.  = NO.  PARTS  OF  PER  REVENUE REVENUE STANDBY REVENUE REVENUE STANDBY  ARRIVALS  THOUSANDS  GROUPS {%) GROUPS WITHOUT T I C K E T S . GROUPS WITHOUT T I C K E T S . GROUPS U S I N G C U R B S I D E C H E C K - I N . U N I T S WITH E X C E S S BAGGAGE. UNITS WITH EXCESS BAGGAGE.  v**-  • V «Y>  ** **  2.1  **  FUNCTIONS  FUNC TION 0 .1 .915 .7 2.3 .92 3.9 .99  R N l , C24 .104 .2 1.2 .75 2.52 .94 4.6 .995  NEGATIVE .222 .3 1.38 .8 2.81 .95 5.3 .998  2 .16  FUNCTION 10 .30  RN2,03 14 1.0  20  3 . 42  FUNC T I O N 2 .60  RN3,D5 3 .81  4  4 0 1.0  FUNCTION 60 .12 360  R N 4 ,C7 120 .39  SERVICE 150 .58  T I M E TO 180  PURCHASE TICKET .85 240 .92  300  5  FUNC TION 35 .30  RN5,C6 60 .58  90  SERVICE .75  T I M E TO 120  PAY .85  180  FUNC TION 60 3  P40,05 90 4  BAGGAGE 125 5  COUNTER 160  SERVICE TIME 6 180  1 0 .6 .90 . 98  **  ** 0  ** 6 2  GROUP  EXPONENTIAL .355 .4 1.6 .84 2.99 .96 6.2 .999  ARRIVAL  FUNCTION .509 .5 1.83 .88 3.2 .97 7.0 .9997  .69 2.12 3.5 8.0  COMPOSITION  GROUP S I Z E D I S T R I B U T I O N .91 5 1.0 6  EXCESS 150  CHARGE 1.0 FOR  GROUPS  ** ** ** ** **  2.2  **  VARIABLES  2.21  **  REVENUE  FOR  BAGGAGE  1 2 3 4 5 6 7 8 9 **  BVARIABLE BVARI ABLE BVARIABLE BVARIABLE BVARIABLE BVARIABLE BVARIABLE BVARI ABLE BVARIABLE  Fl F2 F3 F4 F5 F6 F7 F8 F9  1 2 3 4 5 6 7 8 9 **  VARIABLE VARIABLE VARIABLE VARIABLE VARIABLE VARIABLE VARIABLE VARIABLE VARIABLE  BV1+Q1 BV2+Q2 BV3+Q3 BV4+Q4 BV5+Q5 BV6+06 BV7+Q7 BV8+Q8 BV9+Q9  2.22 11 12 13 14 15 16 ** 11 12 13 14 15 16  ** **  ** 17 18 17 18 ** 10  **  REVENUE  **  TICKET  VARIABLE VARIABLE VARIABLE VARIABLE VARIABLE VARIABLE  BV11+011 BV12+Q12 BV13+Q13 BV14+014 BV15+Q15 BV16+016  STANDBY  COUNTERS  ,BV(J)  ARITHMETIC VARIABLES VARIABLE V ( J ) WILL CALCULATE THE L E N G T H OF Q U E U E 0 ( J ) FOR FACILITY SELECTION PROCESS  COUNTERS  *  F l l F12 F13 F14 F15 F16  **  SELECTION  BOOLEAN V A R I A B L E S IF F A C I L I T Y F ( J ) I S I N USE W I L L = 1 , I F NOT BV(J)=0.  BVARIABLE BVARIABLE BVARIABLE BVARIABLE BVARIABLE BVARIABLE  2.23  FACILITY  • . • • •  SIMILAR  TO  SECTION  2.21 BOOLEAN,  *  SIMILIAR  TO  SECTION  SIMILIAR  TO  SECTIONS  2.21  ARITHMETIC.  COUNTERS  BVARIABLE BVARIABLE VARIABLE VARIABLE  F17 F18 BV17+017 BV18+018  FVARI ABLE  (900*29)/(XH7*10)  COMPUTES  MEAN  GROUP  2.21 + 2 . 2 2 .  INTERARRIVAL  SECTION  GROUP  THREE  ARRIVALS  -  THE  CHECK-IN  SYSTEM  ** **  -  GENERATE  E A C H P A S S E N G E R G R O U P W I L L H A V E A N S E T ** OF C H A R A C T E R I S T I C S U P O N A R R I V A L A T T H E * * ** AIRPORT V 1 0 , F N 1 f , , , 5 0 , H C R E A T I O N OF G R O U P S  MARK ASSIGN ASSIGN ASSIGN ASSIGN ASSIGN ASSIGN  10 30,XH30 31,XH31 32,XH32 33,XH33 34,XH34 35,XH35  1 1 * *  TO  *  COMPUTE  TIME  IN  SYSTEM  • • A S S I G N M E N T OF P A R A M E T E R P ( J ) • OF H A L F W O R D S A V E V A L U E X H ( J ) . •  VALUE  *  DETERMINATION  OF  GROUP  AND  TYPE  TRANSFER  FN2  ASSIGN ASSIGN ASSIGN TRANSFER  40, 1 42 , 1 7 ,45 , AAA  P40=l PASSENGER A R R I V E S ALONE NO. OF P A S S E N G E R S P E R G R O U P M E A N S E R V I C E T I M E AT B A G G A G E C O U N T E R TRANSFER TO•ALLOCATION• S E C T I O N (AAA)  ASSIGN ASSIGN ASSIGN ASSIGN ASSIGN TRANSFER  41 ,FN3 42,P41 41-,1 40, 1 7, 4 5 AAA  G R O U P A R R I V A L OF I N D E P E N D E N T PASSENGERS P 4 1 = NO. I N GROUP P 4 0 = S E R V E D I N D E P E N D E N T OF G R O U P P7= MEAN S E R V I C E T I M E AT BAGGAGE COUNTEr  ASSIGN A S S IGN ASSIGN TRANSFER  40,FN3 42,P40 7 ,FN6 , A AA  F A M I L Y GROUP A R R I V A L P 4 0 = S I Z E OF F A M I L Y FN6= MEAN S E R V I C E T I M E  12  **  ALLOCATION  FN2=  SIZE  D I S T R I B U T I O N OF  TRANSFER  OF  GROUPS  SAVEVALUE SAVEVALUE TEST E  2 0 + ,1 21+,P42 P40,l,RPS  TRANSFER  . * 3 0 , R P S ,SPS  TO  GROUP  T O ' A L L O C A T I ON•  TYPE  SECTION  BY  GROUP  (AAA)  SIZE  SUB-SYSTEMS  C O U N T NO. OF G R O U P A R R I V A L S - T O T A L C O U N T N O . OF P A S S E N G E R S -TOTAL F A M I L I E S TRAVEL R E V E N U E ONLY I F P 4 0 = l , T R A N S A C T I O N I S NOT A F A M I L Y I D E N T I F I C A T I O N OF R E V E N U E PASSENGER GROUPS RPS= REVENUE P A S S E N G E R SYSTEM SPS= STANDBY  **  **  ** REVENUE PASSENGER SYSTEM *# 3, 2 ** A L L O C A T I O N TO R E V E N U E SUBSYSTEMS ** 3, 2 1 * * ** SAVEVALUE 8+ ,l COUNT = TOTAL REVENUE GROUPS RPS SAVEVALUE 9+,P42 COUNT = TOTAL REVENUE PASSENGERS TRANSFER . * 3 1 , C U R B , RP TS I D E N T I F I C A T I O N OF N O N - T I C K E T E D GROUPS RPTS= REVENUE P A S S . T I C K E T SYSTEM SAVEVALUE 24+,1 COUNT = TOTAL REVENUE GROUPS T I C K E T E D CURB SAVE VALUE 25+,P42 COUNT = TOTAL REVENUE P A S S . T I C K E T E D TRANSFER . * 3 3 , R P B S ,CEND I D E N T I F I C A T I O N OF R E V E N U E G R O U P S U S I N G CURBSIDE CHECK-IN. ** RPBS = REVENUE P A S S . BAGGAGE SYSTEM ** C E N D = E N D OF C H E C K - I N FOR C U R B S I D E  **  **  **  3.22  RPBS **  RPSS  RPBA  RPBB  **  RPBS  -  REVENUE  S E L E C TMIN  1,1,XH1,,V  MARK SPLI T TRANSFER  21 P41,RPBA , RPBA  S E L E C TMIN  1 , 1 , X H 1 , ,V  PASSENGER  BAGGAGE  SYSTEM  S E L E C T I O N OF S H O R T E S T Q U E U E . Q U E U E NO. IS PLACED IN PARAMETER P 1 . X H 1 = NO. OF 1 COUNTERS OPEN TO C O M P U T E T I M E I N Q U E U E A T R P B S S E P A R A T E G R O U P S OF I N D E P E N D E N T PASS.  S E L E C T I O N OF S H O R T E S T Q U E U E FOR P A S S . RPTS - T I C K E T PURCHASE TO C O M P U T E T I M E I N Q U E U E S A T R P B S  1 F 1 ROM  MARK  21  QUEUE SEIZE DEPART MARK  PI PI PI 22  TABULATE ASSIGN TEST G SAVEVALUE SAVEVALUE SAVEVALUE SAVEVALUE  1 22-,P21 P22,150,RPBB 21+,P40,H 28+,1 22+,P40,H 29+,1  TRANSFER  . * 3 4 , R P B C ,RPBD  .  WAIT  TIME  IN QUEUE  PASS.  FOR  LEAVES  SERVICE  QUEUE  FOR  SERVICE  TIME IN S Y S T E M C A L C U L A T E D TIME.G.2.5 MIN. YES PROCEED. iWAS iC O U N T = N O . P A S S . L O N G E R T H A N 2 . 5 M I N . lC O U N T = NO. P A S S . U N I T S L O N G E R T H A N 2.5 COUNT= TOTAL PASSENGERS iC O U N T = T O T A L P A S S . U N I T S  EXCESS  RPBC  ADVANCE RELEASE TRANSFER  P7 ,FN1 PI , END  BAGGAGE-RPBC; NO E X C E S S -RPBD: EXCESS S E R V I C E T I M E =P7 * F U N C T I O N 1 SERVICE IS COMPLETED GO TO END SECTION 3.34  RPBD  ASS IGN SAVEVALUE SAVEVALUE ADVANCE RELEASE ASSIGN TRANSFER  7 + ,30 13+,P40 30 + , 1 P7,FN1 PI 3,1 , R P TB  A D J U S T S E R V I C E T I M E FOR E X C E S S B A G G A G E C O U N T = NO. P A S S E N G E R S W I T H E X C E S S . C O U N T = NO. P A S S . U N I T S W I T H E X C E S S S E R V I C E T I M E =P7 * F U N C T I O N 1 S E R V I C E I S COMPLETE I D E N T I F Y P A S S . U N I T WITH E X C E S S : P 3 = l GO TO R E V . P A S S . T I C K E T 'B«  ** #*  3.23  RPTS R P TA  **  SPL I T S E L EC TM I N  P41,RPTA 2,XH2,15,,V  TRANSFER  »RP TE  RPTB  S E L EC T M I N  2 , X H 2 , 1 6 , ,V  R P TE  MARK QUEUE SEIZE DEPART MARK  24 P2 P2 P2 25  TABULATE ASSIGN TEST G SAVEVALUE SAVE VALUE SAVEVALUE SAVEVALUE  2 25-,P24 P 2 5 , 1 5 0 , R P TC 24+,P40,H 31 + , 1 2 5 + , P 4 0 ,H 32+, 1  TIME IN WAS TIME COUNT = COUNT = COUNT = COUNT =  TEST  P 3 , K 1 , R P TD  IF  ADVANCE RELEASE TRANSFER  1 ,FN5 P2 , END  . S E R V I C E FOR * GO TO 'END'  ADVANCE RELEASE ASSIGN TRANSFER  1 ,FN4 P2 6+, 1 ,RPSS  * S E R V I C E FOR P U R C H A S E OF T I C K E T I D E N T I F Y P A S S . WITH T I C K E T P6=l GO TO REVENUE PASS. BAGGAGE'S ,SEC.3.22  R P TC  R P TD  E  D I V I D E INTO PASSENGER UNITS S E L E C T SHORTEST QUEUE(XH2-15) PLACE NO. I N P A R A M E T E R 2.NON-TICK ETED PASS, GO TO R P T E S E L E C T I O N OF S H O R T E S T Q U E U E FOR P A S S , WITH E X C E S S BAGGAGE ONLY TO C O M P U T E T I M E I N Q U E U E S A T R P T S * . W A I T I N Q U E U E FOR S E R V I C E * T I M E P A S S . L E A V E S Q U E U E FOR S E R V I C E  SYSTEM CALCULATED .G. 2 . 5 M I N . Y E S PROCEED P A S S . W A I T I N G LONGER THAN PASS. UNITS LONGER THAN TOTAL P A S S . I N RPTS TOTAL P A S S . U N I T S I N RPTS  P A S S . HAS  EXCESS  P3=l  ,  2. 2.  PROCEED  EXCESS PAYMENT.GO S E C . 3.4  TO  •END'  1  ** **  3.  **  STANDBY  ** **  3. 3 1  **  ALLOCATION  SPS  ** **  ASSIGN SAVEVALUE SAVE VALUE TRANSFER  PASSENGER TO  SYSTEM  SUB-SYSTEMS  5+,l I D E N T I F Y STANDBY PASSENGER P5=l 39+,P42 COUNT = TOTAL STANDBY PASSENGERS 40+, 1 COUNT = TOTAL STANDBY GROUPS . # 3 2 , S P B S , S P T S I N D E N T I F Y STANDBY GROUPS NON-TICKETED  SPBS - STANDBY PASSENGER BAGGAGE SYSTEM 3.32 ** ** S E P A R A T E G R O U P S OF I N D E P E N D E N T PASSENGER P41 ,SPBB SPL I T SPBS S E L E C T I O N OF S H O R T E S T Q U E U E . Q U E U E NO. 4,17,XH3,,V S E L EC T M I N SPBB I S P L A C E D I N P A R A M E T E R P 4 . X H 3 -16 =NU. ** OF C O U N T E R S O P E N . TO C O M P U T E T I M E I N Q U E U E S A T SPS-BAGGAGb 27 MARK * P4 QUEUE P4 SEIZE . W A I T I N G I N Q U E U E FOR S E R V I C E P4 DEPAR T 28 MARK TIME PASSENGER L E A V E S QUEUE  **  SPBC  T A B U L A TE ASSIGN TEST G SAVE VALUE SAVEVALUE SAVEVALUE SAVEVALUE  3 28-,P27 P 2 8 , 1 5 0 ,S P B C 27+,P40, H 41 + , 1 28+,P40, H 42 + , 1  TRANSFER  .*35,SPBD,SPBE  ADVANCE RELEASE TRANSFER  P7,FN1 P4 , S END  S E R V I C E T I M E TO C H E C K - I N LEAVE COUNTER GO TO 'END' S E C . 3 . 4  ASSIGN SAVEVALUE SAVEVALUE ADVANCE RELEASE ASSIGN TRANSFER  7+,30 17+,P40 43 + , 1 P7,FN1 P4 3,1 ,SPTB  A D J U S T MEAN F O R ' E X C E S S • T I M E C O U N T = NO. P A S S E N G E R S W I T H E X C E S S C O U N T = NO. P A S S E N G E R U N I T S E X C E S S S E R V I C E T I M E TO C H E C K - I N WITH EXCESS L E A V E COUNTER I D E N T I F Y PASSENGERS WITH EXCESS GO TO S T A N D B Y P A S S . T I C K E T » B ' S E C . 3 . ^  ** ** SPBD  SPBE  ** 3 ** SPTS  SPTA ** SPTB ** SPTC  SPTD  SPTE  13  SPTS  -  STANDBY  TIME IN WAS TIME COUNT = COUNT = COUNT = COUNT =  SYSTEM CALCULATED GREATER THAN 2.5 M I N . - S P S C ( N G ) TOTAL PASSENGER OVER 2.5 MIN T O T A L P A S S . U N I T S OVER 2.5 MIN T O T A L S T A N D B Y P A S S . AT B A G G A G E TOTAL STANDBY P A S S . U N I T S BAGG.  IDENTIFY  PASSENGER  P A S S . WITH  EXCESS  -  SPBE  TICKET SYSTEM  *  ASSIGN SAVEVALUE SAVEVALUE SPLI T TRANSFER  2+,16 33+,P42 34 + , 1 P41 ,SPTA ,SPTC  P A S S . Q U E U E A T # 1 6 TO B U Y A T I C K E T C O U N T = NO. P A S S . N O N - T I C K E T E D C O U N T = NO. P A S S . U N I T S NON-TICKETED S E P A R A T E G R O U P S OF I N D E P E N D E N T PASS. GO TO'SPTC  SELECTMIN  2 ,XH2 , 1 6 » t V  SELECT  MARK QUEUE SEIZE DEPART MARK  24 P2 P2 P2 25  T A B U L A TE ASSIGN TEST G SAVEVALUE SAVEVALUE SAVEVALUE SAVEVALUE  4 25-,P24 P 2 5 , 1 5 0 , SPTD 35+,P40 36 + , 1 37+,P40 38 + , 1  C A L C U L A T E TIME I N QUEUE WAS TIME GREATER THAN 2.5 MIN.-SPTD C O U N T = NO. P A S S . G R E A T E R T H A N 2 . 5 M I N . C O U N T = NO. P A S S . U N I T S G R T T H 2 . 5 M I N . C O U N T = T O T A L P A S S E N G E R S AT SPTS C O U N T = T O T A L P A S S . U N I T S AT S P T S  TEST  P3,1,SPTE  IF  ADVANCE RELEASE TRANSFER  1 t FN5 P2 , S END  *  ADVANCE RELEASE TRANSFER  1 ,FN4 P2 ,SPBB  SERVICE  E  TO  * .  *  COMPUTE WAIT  TIME  GO  GO  SHORTEST  PASS.  IN  TO  'SEND' TO  P2  SERVICE  PAYMENT PAY  NO.IN  QUEUE  -  CONTINUE  EXCESS  SEC.  PURCHASE  STANDBY  PLACE  QUEUES  FOR  LEAVES  EXCESS  SERVICE  TO  TIME  I N QUEUES  FOR  TO  QUEUE  PASS.  3.4 TICKET BAGGAGE  • B  1  SEC.3.<  ** ** ** END  SAVEVALUE SAVEVALUE MARK SAVEVALUE SAVEVALUE TERM INA TE  1+,P40 44+, 1 11 2+,V51 45+,V38  COUNT = NO. REV. PASSENGERS CHECKED-IN COUNT = NO. REV. PASS. UNITS CHKD IN COMPUTE TIME LEAVE SYSTEM CALCULATE=TIME * NO.PASSENGERS =AVG.T.1 CALCULATE = TI ME FOR AVERAGE TRANSIT TH END  CEND  SAVEVALUE SAVEVALUE TERM INA TE  11+,P40 48 + ,1  COUNT = NO. CURBSIDE PASSENGERS COUNT = NO. CURBSIDE PASSENGER UNITS  SEND  SAVEVALUE SAVEVALUE MARK SAVEVALUE SAVE VALUE TERMINATE  3+,P40 46 + , 1 11 4+,V51 47+,V38  30 31 32 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54  SAME AS 'END',ONLY THIS IS FOR STANDE PASSENGER CALCULATIONS  4.0  **  SECTION FOUR - CALCULATIONS AND CONTROL  4.1  **  VARIABLES FOR STATISTICAL CALCULATIONS CALCULATIONS ARE ASSIGNED TO SAVEVALUES IN SECTION 4.2* CORRESPONDING SAVEVALUES APPEAR ON THE RIGHT.  FVARIABLE FVARIABLE FVARIABLE FVARIABLE FVARIABLE FVARIABLE FVARIABLE VARIABLE FVARIABLE VARIABLE VARIABLE FVARIABLE VARIABLE VARIABLE VARIABLE VARIABLE VARIABLE VARIABLE FVARIABLE FVARIABLE VARIABLE FVARIABLE FVARIABLE FVARIABLE  (X28*1000)/X29 RPBS-P.U. .G.2.5 (X55) (X56) (X31*1000)/X32 RPTS-P.U. .G.2.5 (X57) (X41*1000)/X42 SPSB-P.U. .G.2.5 (X36*1000)/X38 SPST-P.U. .G.2.5 (X59) (X8*1000 J/X20 % REV. GROUP ARRIVALS (XH42) 1000-V35 % STB. GROUP ARRIVALS (X60) % REV. PASSENGERS TO TOTAL (X49) (X9*1000)/X21 PASS. UNIT TRANSIT TIME P11-P10 X8-X24 REV. GROUPS NON-TICKETED (X50) NO. REV. TICKET COUNTERS (XH40) 16-XH2 XH3-16 NO. STB. BAGG. COUNTERS (XH41) (V39*1000)/X8 % REV. GROUPS NON-TICKETED (X51) FR1+FR2+FR3 + FR4+FR5+FR6 + FR7+FR8 + FR9 REV.BAG.UT ILIZ V43/XH1 AVG. UTILIZATION-RPBS (XH44) FR11+FR12+FR13+FR14+FR15 (XH46) V45/V40 AVG. UTILIZATION-RPTS FR18+FR17 (XH48) AVG. UTILIZATION-SPSB V47/V41 NO. PASS.UNITS NON-TICKETED ( X5 ) X9-X25 % PASS. UNITS NON-TICKETED (X6) {V49*1000)/X9 TRANSIT TIME- REV.PASSENGERS (P11-P10 )=.P40 % GROUPS CURBSIDE CHECK-IN (X7) (X48*1000)/X8 % PASS. CURBSIDE CHECK-IN (Xll*1000)/X9 ( X10) (X39*1000)/X21 % STANDBY PASSENGERS OF T0TAL(X12)  FVARIABLE FVARIABLE FVARIABLE FVARIABLE FVARIABLE FVARIABLE FVARIABLE FVARIABLE FVARIABLE FVARIABLE VARIABLE  55 56 57 58 59 60 61 62 63 64 65 **  ** ** **  *#  4.2  (X34*1000)/X40 (X33* 1000)/X39 (X30*1000 )/X29 (X13*1000)/XH22 (X43*1000)/X42 (X17*1000)/XH28 (X45/X44./6 (X2/XD/6 (X47/X46)/6 (X4/X3)/6 FR16  STB. GROUPS NON-TICKETED-^ % STB . PASS. NON-TICKETED % REV .PASS.UNITS EXCESS % REV .PASS. EXCESS % STB . PASS.UNITS EXCESS % STB . PASS.EXCESS AVG. TRANSIT TIME - REV.PU. AVG. TRANSIT TIME - REV P. AVG. TRANSIT TIME - STB PU. AVG. TRANSIT TIME - STB P. STB. UTILIZATION -SPST  (X14) (X16) (X18) (X19) (X53) (X54) (X61) (X62) ( X63) (X64) (X65)  ASSIGNMENT OF VARIABLE VALUES FOR OUTPUT VALUES OF SECTION 4.1 ASSIGNED TO SAVEVALUES IN 4.2 FOR PRINTING IN REQUESTED FURMAT OF OUTPUT GENERA TE SAVEVALUE SAVEVALUE SAVEVALUE SAVEVALUE SAVEVALUE SAVEVALUE SAVEVALUE SAVEVALUE SAVEVALUE SAVEVALUE SAVEVALUE SAVEVALUE SAVEVALUE SAVEVALUE SAVEVALUE SAVEVALUE SAVEVALUE SAVEVALUE SAVEVALUE SAVEVALUE SAVEVALUE SAVEVALUE SAVEVALUE SAVEVALUE SAVEVALUE SAVEVALUE SAVEVALUE SAVEVALUE SAVEVALUE SAVEVALUE TERM I NA TE  300 40,V40,H 41 ,V41,H 42,V35,H 44,V44 ,H 46,V46,H 48, V48,H 5 ,V49 6, V50 7, V52 10,V53 12,V54 14,V55 16,V56 18 ,V57 19,V58 49, V37 50, V39 51, V42 53, V59 54, V60 55, V30 56, V31 57, V32 59, V34 60, V36 61, V61 62, V62 63, V63 64, V64 65, V65  ** ** 1 2 3 4 ** ** **  TABLE TABLE TABLE TABLE 4.4  MP21,0,30,14 MP24,0,30,14 MP27,0,30,14 MP24,0,30,14 CONTROL  G E N E R A TE SAVEVALUE TERMINATE START  OF  60  9+,l,H 1 120  PASS. UNIT WAITING IN QUEUES. T I M E C A L C U L A T E D I N I N T E R V A L S OF 3 0 S E C T A B L E N O . R E F E R S TO ' T A B U L A T E CARD I N MAIN PROGRAM S E C T I O N 3. 1  SIMULATED  TIME  GENERATE EVERY 60 SECONDS C O U N T = NO. OF S I M U L A T E D T I M E : M I N U T E S END = ONE MINUTE S TAR T= A M O U N T OF T I M E TO S I M U L A T E  S E C T I O N F I V E - REQUESTED OUTPUT $$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$  REPORT EJECT SPACE TEXT THE M E A N A R R I V A L R A T E WAS # X H 7 , 2 / X X X X # P A S S E N G E R S 12 31 TEXT PER 15 M I N U T E P E R I O D . SPACE 1 10 TEXT REVENUE PASSENGER SYSTEM 1 SPACE TEXT T R A N S I T T I M E I N THE S Y S T E M 12 TEXT AVERAGE TIME PER P A S S E N G E R UNIT #X61,2/1LXXX.X#MIN. 12 TEXT A V E R A G E TIME PER P A S S E N G E R #X62,2/1LXXX.X#MIN. 12 SPACE 1 TEX T 12 REVENUE PASSENGER BAGGAGE SYSTEM TEXT 12 N U M B E R OF C O U N T E R S O P E N #XH1,2/XXX#. A V E R A G E U T I L I Z A T I O N OF F A C I L I T I E S #XH44,2/1LXXX.X#% 12TEXT TEXT A V E R A G E W A I T I N G T I M E I N Q U E U E S # T 1 , 3 / X X X X # S E C . P E R P£ 12 S S E N G E R UN I T % OF P A S S E N G E R U N I T S I N Q U E U E S O V E R 2 . 5 M I N . WAS #X55 12 TEXT , 2 / 1 L X X X ,X#Sg SPACE 12 TEXT REVENUE PASSENGER TICKET SYSTEM N U M B E R OF C O U N T E R S O P E N #XH40,2/XXXX#. 12 TEXT A V E R A G E U T I L I Z A T I O N OF F A C I L I T I E S # X H 4 6 , 2 / 1 L X X X . X # % ( 12 TEXT COUNTERS 1 1 - 1 5 ONLY) A V E R A G E W A I T I N G T I M E I N Q U E U E S # T 2 , 3 / X X X X # S E C . P E R P/! 12 TEXT SSENGER U N I T (11-15). % OF P A S S E N G E R U N I T S I N A L L T I C K E T OR E X C E S S Q U E U E S C 12 TEXT V E R 2 . 5 M I N . WAS # X 5 6 , 2 / 1 L X X X . X # % SPACE STANDBY PASSENGER SYSTEM 10 TEXT 1 SPACE T R A N S I T T I M E I N THE S Y S T E M 12 TEXT #X63,2/lLXXX.X#MIN. AVERAGE TIME PER P A S S E N G E R UNIT 12 TEXT #X64,2/1LXXX.X#MIN. AVERAGE TIME PER P A S S E N G E R 12 TEXT 1 SPACE STANDBY PASSENGER BAGGAGE S Y S T E M 12 TEXT NUMBER UF C O U N T E R S O P E N #XH41,2/XXXX# 12 TEXT A V E R A G E U T I L I Z A T I O N OF F A C I L I T I E S # X H 4 8 , 2 / 1 L X X X . X#% 12 TEXT A V E R A G E W A I T I N G T I M E I N Q U E U E S # T 3 , 3 / X X X X # S E C . P E R Pt 12 TEXT SSENGER U N I T . 12 TEXT % P A S S E N G E R U N I T S I N Q U E U E S O V E R 2 . 5 M I N . WAS #X57,2/ lLXXX.X#5g SPACE 12 TEXT STANDBY PASSENGER TICKET SYSTEM 12 TEXT A V E R A G E U T I L I Z A T I O N OF F A C I L I T I E S # X 6 5 , 2 / 1 L X X X . X # % 12 TEXT AVERAGE W A I T I N G TIME IN QUEUES #T4,3/XXXX# S E C . PER . ASSENGER UNIT. 12 TEXT * P A S S E N G E R U N I T S I N Q U E U E S O V E R 2 . 5 M I N . WAS # X 5 9 , 2 / 1LXXX.X#%  **  APPENDIX I I I SAMPLE OUTPUT  REVENUE PASSENGER SYSTEM TRANSIT TIME IN THE SYSTEM AVERAGE TIME PER PASSENGER UNIT AVERAGE TIME PER PASSENGER  2.8 MIN. 3.1 MIN.  REVENUE PASSENGER BAGGAGE SYSTEM NUMBER OF COUNTERS OPEN 5. AVERAGE UTILIZATION OF FACILITIES 59.6% AVERAGE WAITING TIME IN QUEUES 21 SEC. PER PASSENGER UNIT % OF PASSENGER UNITS IN QUEUES OVER 2.5 MIN. WAS 5.5% REVENUE PASSENGER TICKET SYSTEM NUMBER OF COUNTERS OPEN 3. AVERAGE UTILIZATION OF FACILITIES 41.8% (COUNTERS 11-15 ONLY) AVERAGE WAITING TIME IN QUEUES 32 SEC. PER PASSENGER UNIT (11-15). % OF PASSENGER UNITS IN ALL TICKET OR EXCESS QUEUES OVER 2.5 MIN. WAS 9.1% STANDBY PASSENGER SYSTEM TRANSIT TIME IN THE SYSTEM AVERAGE TIME PER PASSENGER UNIT 2.3 MIN. AVERAGE TIME PER PASSENGER 2.3 MIN. STANDBY PASSENGER BAGGAGE SYSTEM NUMBER OF COUNTERS OPEN 2 AVERAGE UTILIZATION OF FACILITIES 34.1% AVERAGE WAITING TIME IN QUEUES 41 SEC. PER PASSENGER UNIT % PASSENGER UNITS IN QUEUES OVER 2.5 MIN. WAS 10.7% STANDBY PASSENGER TICKET SYSTEM AVERAGE UTILIZATION OF FACILITIES 45.5% AVERAGE WAITING TIME IN QUEUES 104 SEC. PER PASSENGER UNIT. % PASSENGER UNITS IN QUEUES OVER 2.5 MIN. WAS 27.0%  APPENDIX IV USER'S MANUAL  The user's manual for the simulation model (Appendix III)  will  be presented in five sections: I.  Reference Material.  Basic support material w i l l be  provided concerning the language and computer time used for simulating the system. II.  Identification of the Simulation Model.  Working  closely with the program l i s t i n g , the basic sections of the model w i l l be identified.  No programming  knowledge is expected in this section. III.  A Program for Analysis.  A step by step approach w i l l  be presented to enable the user to carry out the first two steps of the systems analysis as presented in Chapter III.  Only a very basic knowledge of computer  use is expected. (  IV.  The Verification Steps.  The adjustment of the main  program for the verification steps contained in Chapter III w i l l be presented. V.  Analytical Steps.  Similar to Section IV, the adjust-  ments made for the analytical steps w i l l be presented.  on the GPSS VERSION V c o m p i l e r .  Most  computer  companies have s i m i l a r l a n g u a g e s . O p e r a t i n g Equipment:  The U n i v e r s i t y o f B r i t i s h Columbia has  an IBM360/67 Duplex computer w i t h a v a r i e t y o f software. Time Usage:  The computer time usages (assembly and e x e c u t i o n ) f o r a s i m u l a t i o n run o f 180 minutes i s a p p r o x i m a t e l y .7 minutes.  IBM Manuals: 1.  G e n e r a l Purpose S i m u l a t i o n System/360 User's  Manual R e f e r e n c e Number GH20-0326-3. 2.  GPSS/360 I n t r o d u c t o r y User's Manual R e f e r e n c e  Number H20-0304-0.  II.  IDENTIFICATION  OF THE SIMULATION MODEL  The program l i s t i n g sections.  i n Appendix I I I i s s u b d i v i d e d i n t o  The u s e r s h o u l d f o l l o w a l o n g w i t h the l i s t i n g  h i m s e l f w i t h t h e computer  five  to f a m i l i a r i z e  'deck' make-up.  I n t r o d u c t o r y Statements 1.  Each l i n e i n the l i s t i n g  r e p r e s e n t s 1 computer c a r d .  Each  c a r d c o n s i s t s o f 80 spaces o r columns. 2.  Each a l p h a b e t i c o r n u m e r i c a l c h a r a c t e r o r b l a n k  1 space o r column. 3.  Column '1' i s on the l e f t hand  side.  represents  4. or l i n e .  An ' * ' asterik in column '1' represents a comment card It may be used to comment on the operation of the deck or  as a 'space' card. 5.  A l l other cards are operational.  The two cards prior to Section 1 are control cards and must start each deck. The first card i s : Column  Controls  1 2-4 6 7-10 13-22  $ RUN * GPS5 PAR=SIZE=C  The second card i s : Column 8-15  SIMULATE  The program has been called and activated. Section 1.  Section one is the i n i t i a l i z a t i o n of the system or the  setting of the parameters. Section 1.1.  It consists of three sub-sections:  F a c i l i t y Policy.  The number of counters  that are available are stated here. Section 1.2.  Arrival Rate.  The number of passengers  that arrive in a 15-minute period is stated in this section. Section 1.3.  The percentage of passengers having spec i f i c characteristics  (non-ticketed,  etc.)  is stated here. The above three sub-sections w i l l be explained under Program  for Analysis.  The essence of this section is to set the i n i t i a l  conditions of the system Section 2.  Section two contains two sub-sections.  The f i r s t  is  the specifications of the functions used to describe the service times, and type and group composition.  The second sub-section  contains the variables for the f a c i l i t y selection process described in Chapter III. Section 3.  Section three contains the program of the passenger flow  through the various systems.  The subsections are:  3.10  Arrival of groups  3.11  Determination of Group size and type  3.12  Allocation of groups to sub-systems (the Arrival Stream)  3.2  Revenue Passenger System  3.21  Allocation to Revenue Subsystems  3.22  Revenue Baggage System  3.23  Revenue Ticket System  3.3  Standby Passenger System  3.31  Allocation to Subsystems  3.32  Standby Baggage System  3.33  Standby Ticket System  3.4  End of Flow for Passengers  On the right hand side starting in Column 33 there is a narrative of the process performed by each line.  This thesis suggests  that an attempt be made to read through the program. Section 4. This section is for calculations and control of the simulation. 4.1  Calculation of system statistics.  4.2  Transfer of statistics to 'save values' for reference.  4.3  Table identification for calculation of transit times.  4.4  Section 4.4 is very important as i t controls the length of simulated time.  Both this section ( 4 . 4 )  and Section 1 w i l l be discussed in Program for Analysis. Section 5.  This section is the request for the output to be printed  in a certain manner.  If this section is omitted, GPSS/360 has an  automatic output. After Section 5, there is another control card: Columns 8-10  END  If Section 5 is omitted, the card is entered after Section 4.  Ill  PROGRAM FOR ANALYSIS This section w i l l enable the user to carry out an analysis  similar to Analysis Step I or Step II as cited in Chapter III.  Step 1.  (a)  Select the f a c i l i t y policy that is in  operation—the number of counters open for the Revenue baggage and ticket counter and the Standby baggage counter. (b)  Remove the three cards in Section 1.1.  (c)  Revenue Baggage counters open.  Columns 8-14 Columns 19-22  INITIAL XHI,  Column  print the number of counters open.  23  For example, i f 5 were open the card would read: INITIAL  XHI, 5  (d) Revenue Ticket Counters open. Columns 8-14 INITIAL Columns 19-22 XH2 Columns 23-24  print the value of 16—no. of counters open  For example, i f 3 ticket counters were open, the card would read: INITIAL  XH2, 13  (e) Standby Baggage Counters Columns 8-14 INITIAL Columns 19-22 XH3 Columns 23-24  print the value of 16 + number of Standby  counters open. For example, i f two baggage Standby counters were open, the card would read INITIAL (f)  XH3, 18  Replace Section 1.1 with the three new cards.  our example, Section 1.1 would read: INITIAL XHI, 5 INITIAL XH2, 13 INITIAL XH3, 18  Step 2. period.  (a)  The a r r i v a l r a t e o f passengers p e r 15 minute  Remove the c a r d i n S e c t i o n 1.2 t h a t i s marked  INITIAL.  (b)  S e l e c t the a r r i v a l r a t e f o r a 15 minute  (c)  Columns Columns Columns 24 &  8-14 INITIAL 19-22 XH7 23, 25 the a r r i v a l  period  rate.  F o r example, i f t h e a r r i v a l r a t e was 100 p a s s e n g e r s p e r 15 minute  p e r i o d , the c a r d would INITIAL  Step 3.  read: XH7, 100  (d)  Replace the new c a r d i n S e c t i o n 1.2  (a)  Section. 1.3 c o n t a i n s 6 c a r d s marked  ( s t a r t i n g i n column '8') and s p e c i f y a r r i v e a t the a i r p o r t w i t h c e r t a i n Start The  INITIAL  the p e r c e n t a g e o f p a s s e n g e r s  that  characteristics.  i n column '19' t h e r e i s an  'XH—,'  two numbers a f t e r the 'XH' r e f e r e n c e c e r t a i n  passengers  characteristics. 'XH30,' = the p e r c e n t a g e o f Standby groups t o t o t a l i n d e pendent groups. Remember t h a t 70% o f group a r r i v a l s a r e f a m i l i e s and t h e r e f o r e dependent. 'XH31,' = the p e r c e n t a g e o f Revenue groups w i t h o u t  tickets  'XH32,' = the p e r c e n t a g e o f Standby  tickets.  groups w i t h o u t  'XH33,' = t h e p e r c e n t a g e o f revenue groups t o t a l a l l o w e d ( i . e . minus n o n - t i c k e t e d groups) t h a t use c u r b side service. 'XH34,' = the p e r c e n t a g e o f Revenue passenger u n i t s w i t h excess baggage.  'XH35,1 = the percentage of Revenue passenger units with excess baggage. The percentages must be specified in parts per thousand.  Thus  i f 15% of Standby passenger units had excess baggage, the card would read: INITIAL  XH35, 150  (b)  Remove cards in Section 1.3  (c)  Column 8-14 INITIAL Column 19-20 XH Column 21-22 the reference number Column '23' print ' , ' Column '24''26' the percentage (in parts per thousand) of the specification  (d) Replace new cards in Section 1.3. Step 4.  There must be six cards.  The step has to do with the length of simulated time.  (a) In section 4.4 has four cards. printed in Columns 8 - 1 2 .  The fourth card has START  Remove this card.  (b) Then Column '8'-'12'  START  Column '19'-'21'  the desired simulated time in minutes.  For example, i f the simulated time was to be 180 minutes, the card would read: START  180  Step 5.  Compile the Complete Deck.  Step 6.  Place deck in card reader with appropriate command  cards surrounding the deck. Step 7. of Appendix III.  The output w i l l be similar to that on the last page  This section w i l l contain the program statements used in the verification steps, Chapter III. A l l the statements or words on the left start in column ' 8 ' , and the statements on the right start in column '19.' line.  Only change the sections indicated.  Verification Step 1.  Section 1.1 should read: INITIAL INITIAL INITIAL  XH1, 5 XH2, 13 XH3, 18  Section 1.2 should read: INITIAL  XH7, 100  Section 1.3 should read: INITIAL INITIAL INITIAL INITIAL INITIAL INITIAL  XH30, XH31, XH32, XH33, XH34, XH35,  670 150 150 118 150 150  Section 1.4 should read: GENERATE SAVEVALUE TERMINATE START Verification Step 2.  60 9+, 1, H 1 120  Section 1.1 should read: INITIAL INITIAL INITIAL  XH1, 5 XH2, 13 XH3, 18  Section 1.2 should read: INITIAL  XH7, 100  One card per  Section 1.3 should read the same as Verification Step 1. Section 4.4 should read: GENERATE SAVEVALUE TERMINATE START CLEAR START CLEAR START CLEAR INITIAL INITIAL INITIAL START CLEAR START CLEAR START Verification Step 3.  60 9+, 1, H 1 240,60 XH1-XH3, XH7, XH30-XH35 240,60 XH1-XH3, XH7, XH30-XH35 240,60 XH30-XH35 XHI, 4 XH3, 17 XH7, 50 240,60 XH1-XH3, XH7, XH30-XH35 240,60 XH1-XH3, XH7, XH30-XH35 240,60  Section 1.1 should read: INITIAL INITIAL INITIAL  XHI, 4 XH2, 13 XH3, 17  Section 1.2 should read: INITIAL  XH7, 64  Section 1.3 should read: INITIAL INITIAL INITIAL INITIAL INITIAL INITIAL  XH30, 890 XH31, 120 XH32, 120 XH33, 118 XH34, 130 XH35, 100  Section 4.4 should read: GENERATE SAVEVALUE TERMINATE START CLEAR START  60 9+, 1, H 1 180 XH1-XH3, XH7, XH30-XH35 180  Section 1.1 should read for INITIAL INITIAL INITIAL  = 100  XH1, 5 XH2, 13 XH3, 18  Section 1.2 should read: INITIAL  XH7, 100  Section 1.3 should read the same as in Verification Step 1 and 2. Section 4.4 should read for a l l three runs: GENERATE SAVEVALUE TERMINATE START CLEAR START Part  60 9+, 1, H 1 180 XH1-XH3, XH7, XH30-XH35 180  (b):  Sections 1.1 and 1.2 w i l l have the f a c i l i t y policy and associated arrival rate as determined in Verification Step 4. V.  ANALYTICAL STEPS Step 1.  Analysis of Service Policy Section 1.1 should read: INITIAL INITIAL INITIAL  XH1, 4 XH2, 14 XH3, 17  Section 1.2 should read: INITIAL  XH7, 40  Section 4.4 should read: GENERATE SAVEVALUE TERMINATE START CLEAR  60 9+, 1, H 1 180 XH1-XH3, XH7, XH30-XH35  INITIAL START CLEAR START CLEAR INITIAL START CLEAR START CLEAR INITIAL START CLEAR START CLEAR INITIAL START CLEAR START CLEAR INITIAL START CLEAR START CLEAR INITIAL START CLEAR START  XH7, 60 180 XH1-XH3, 180 XH1-XH3, XH7, 80 180 XH1-XH3, 180 XH1-XH3, XH7, 100 180 XH1-XH3, 180 XH1-XH3, XH7, 120 180 XH1-XH3, 180 XH1-XH3, XH7, 140 180 XH1-XH3, 180 XH1-XH3, XH7, 160 180 XH1-XH3, 180  XH7, XH30--XH35 XH30, XH35 XH7, XH30--XH35 XH30-XH35 XH7, XH30--XH35 XH30-XH35 XH7, XH30--XH35 XH30-XH35 XH7, XH30--XH35 XH30-XH35 XH7, XH30-•XH35  determine f a c i l i t y policy Sections 1.1,  1.2 and 1.3 must be specified by the user in the  same format that has been presented up to now.  The essence of Step 2  was to set an arrival rate \ ^ and search for a f a c i l i t y policy P^. Thus, set section 1.2 and vary section 1.1 until the service policy achieved. Section 4.4 should read: GENERATE SAVEVALUE TERMINATE START CLEAR START  60 9+, 1, H 1 180 XH1-XH3, XH7, XH30-XH35 180  is  Section 1.1 should read: INITIAL INITIAL INITIAL  XH1, 5 XH2, 13 XH3, 18  Section 1.2 should read INITIAL  XH7, 100  Section 1.23 - RPTS has a minor adjustment RPTB  SELECTMIN  2, XH2, 16, V  Columns '25 and '26 should have a '15' RPTB  SELECTMIN  2, XH2, 15, V  Section 3.33 - SPTS The card marked SPTB  SELECTMIN  2, XH2, 16, V  should be changed to: Column 2-5, SPTB Column 8-13 ASSIGN Column 19-22 2, 16 Section 4.4 should read the same as Step 1. Step 3.  Part (b) Sections 1.1, 1.2, 1.3, 4.4 should be the same as  Step 1. Section 3.23 - RPTS = REVENUE PASSENGER TICKET SYSTEM uses a single queue for a l l ticket counters, when an agent becomes free, a passenger unit w i l l leave the line and go to the agent that is  free.  RPTS Split  P41, RPTA  Mark  25  Substitute with following:  RPTB  RPTS Split Transfer Mark Queue TEST L SELECTMIN SEIZE DEPART MARK  P41, RPTB RPTB 24 10 V19, V40 2, XH2, 15, F P2 10 25  Add i n Section 2.22 19 40  Variable Variable  BV11 + BV12 + BV13 + BV14 + BV15 16 - XH2  Part (c) Sections 1.1, 1.2, 1.3 and 4.4 same as Part A. Remove Sections 3.31, 3.32 and 3.33.  Substitute with following.  (Starting columns are 2, 8 and 19). SPS SPSB SPSA SPSC  SAVEVALUE 39+,P42 SAVEVALUE 40+,1 TRANSFER .*32,SPSA,SPSB ASSIGN 5,1 SAVEVALUE 33+,P42 SAVEVALUE 34+,l SELECTMIN 4,16,XH3,,V SPLIT P41,SPSC MARK 27 MARK 24 QUEUE P4 SEIZE P4 DEPART P4 MARK 28 MARK 25 TABULATE 3  SPSD  SPSF SPSE SPSH  SPSI  SPSG  TABULATE ASSIGN TEST G SAVEVALUE SAVEVALUE SAVEVALUE SAVEVALUE TEST E ASSIGN TEST G SAVEVALUE SAVEVALUE SAVEVALUE SAVEVALUE ADVANCE TRANSFER ASSIGN ADVANCE SAVEVALUE SAVEVALUE TEST G SAVEVALUE SAVEVALUE SAVEVALUE SAVEVALUE ADVANCE RELEASE TRANSFER ADVANCE RELEASE TRANSFER  3 28-,P27 28,150,SPSD 27+,P40,H 41+, 1 28+,P40,H 42+, 1 P5,1,SPSE 25-,P24 P25,150,SPSF 35+.P40 36+,1 37+.P40 38+, 1 1,FN4 •*35,SPSG,SPSH 7+,30 P7,FN1 17+,P40 43+,1 P25,150,SPSI 35+.P40 36+, 1 37+.P40 38+,1 1,FN5 P4 ,SEND P7,FN1 P4 , SEND  APPENDIX V REPORT TO AIR CANADA  The p r i m a r y o b j e c t i v e o f t h i s s t u d y i s t o p r o v i d e a compreh e n s i v e management t o o l t h a t w i l l a i d i n p o l i c y f o r m u l a t i o n and e v a l u a t i o n f o r t h e P a s s e n g e r C h e c k - i n System. The A i r Canada C h e c k - i n System a t t h e v a r i o u s a i r p o r t s can be d e f i n e d by t h e a r r i v a l o f p a s s e n g e r s and t h e i r and, t h e f a c i l i t y  characteristics,  and s e r v i c e p o l i c i e s i n o p e r a t i o n .  L i s t e d below a r e t h e t h r e e performance o b j e c t i v e s o f t h e Check-in  System. 1.  To o b t a i n a h i g h u t i l i z a t i o n o f c h e c k - i n f a c i l i t i e s and thus reduce t h e cose o f s e r v i c e p e r passenger.  2.  To a c h i e v e a minimum p a s s e n g e r w a i t i n g time i n queues and t h u s enhance customer  3.  relationships.  To c h e c k - i n t h e p a s s e n g e r s i n accordance w i t h procedural policy.  In  order that these three o b j e c t i v e s are f u l f i l l e d ,  p o l i c i e s have been implemented 1.  i n t h e system.  They a r e :  S e p a r a t e t h e Standby p a s s e n g e r s e r v i c e  (baggage  c h e c k - i n and t i c k e t purchase) from t h a t o f t h e Revenue P a s s e n g e r .  S i n c e Revenue p a s s e n g e r s  comprise t h e l a r g e s t p e r c e n t a g e o f t r a v e l l e r s  several  and pay f u l l - f a r e f o r t h e i r t i c k e t s , they are the most important.  I f they become d i s s a t i s -  f i e d with service, hopefully i t w i l l not stem from waiting i n queues while a Standby passenger i s being served. 2.  Staff the f a c i l i t i e s such that 85% of the Revenue and 75% of the Standby passenger units (a passenger unit i s equivalent to a single person t r a v e l l i n g i n a group or a family unit) do not exceed 2^ minutes waiting time i n any one queue.  3.  Use a 15-minute s t a f f horizon to e f f e c t i v e l y and e f f i c i e n t l y allocate manpower i n the system.  This  p o l i c y has been implemented because the demand for service (the a r r i v a l rate) fluctuates greatl y during the day and the mobility of agents to other functions i n the a i r p o r t . The effectiveness of a policy may be measured by: 1.  The average u t i l i z a t i o n of the various check-in facilities  2.  (baggage and t i c k e t counters).  The percentage of passenger units who exceed 2% minutes of waiting time i n queues.  3.  The average t r a n s i t time per passenger unit (the t o t a l time spent i n waiting i n queues and being checked-in at the counters).  The aforementioned tool that w i l l aid in policy formulation and evaluation is a computer simulation model.  The Check-in system  at the Vancouver International Airport was studied as to its stat i s t i c a l and operational characteristics,  then modelled.  The tool  was verified that i t did in fact simulate the real system. In order to evaluate a policy, the user of the model simply specifies certain starting conditions, they are: 1.  The number of passengers arriving in the system per 15 minute period.  2.  The Passenger Composition: Revenue Passengers: % of total passengers % using Curbside Check-in % arriving non-ticketed % having excess baggage Standby Passengers: % arriving non-ticketed % having excess baggage  3.  The Facility Policy in Operation The number of Revenue baggage and ticket counters open. The number of Standby baggage and ticket counters open.  What the model w i l l do is define the beahviour of the system using the three previously mentioned statistics for the specified conditions. To demonstrate the capabilities of this model, an analysis was undertaken in three areas of policy management. 1.  The areas are:  To determine the implications of the behaviour of the Passenger Check-in System on policy formulation.  This analysis, w i l l , in effect, de-  termine i f the present service policy is formulated correctly. 2.  To determine the f a c i l i t y policies that should be implemented in order to achieve the present service policy.  As part of this analysis the  maximum capacity of the system w i l l be determined. 3. To formulate alternative operating policies and test for v i a b i l i t y prior to implementation. The results of the analysis are presented sequentially. 1.  Service Policy Evaluation The present service policy for the Revenue passengers is as  follows:  to allow no more than 15% of the Revenue passenger units  to wait in any one queue longer than 2% minutes. The main elements which determine the state pf the system are the arrival rate and the f a c i l i t y policy.  If the f a c i l i t y policy  (4 Revenue baggage counters and 2 Revenue ticket counters) is held  constant, then the effect of the arrival rate on the subsequent behaviour of the system may be determined. Figure 1 depicts the effect of the arrival rate on the average transit time per Revenue passenger unit. When the stated f a c i l i t y policy is in operation and arrival rate is in the range of 40 to 90 passengers per 15 minutes, the average transit time is between 2 and 3 minutes.  The system is thus  relatively stable or insensitive to the number of passengers in the system. When the arrival rate is greater than 90 passengers the curve becomes unstable and is sensitive to the number of passengers in the system.  For example, when the arrival rate of 100 goes to 120 passen-  gers per 15 minute period, the average transit time goes from 3.3 to 6.5 minutes per Revenue passenger unit.  This is a 20 per cent i n -  crease in volume and results in 100 per cent increase in the average transit time per passenger unit. Thus, i f a curve was derived for each f a c i l i t y policy at the various arrival rates, the vertical portion of the curve would indicate the range of arrival rates such that the f a c i l i t y policy produces stability in the system.  ARRIVAL RATE 160 PASSENGERS PER 15 MIN.  EFFECT OF VARIOUS ARRIVAL RATES ON THE AVERAGE TRANSIT TIME NO. OF FACILITIES OPEN HELD CONSTANT  140  120  100 4 REVENUE BAGGAGE COUNTERS INSENSITIVE RANGE  80  60 ]  40  —i  i  .i  2 INSENSITIVE RANGE  4  6  8 10 12 AVERAGE TRANSIT TIME IN MINUTES PER REVENUE PASSENGER UNIT  Figure 2 is a graph of the average utilization of the Revenue baggage counters as the arrival rate varies.  Note  that the curve is almost linear in the range of 40 to 120 passengers per 15 minutes.  When the arrival rate increases  past 120, the curve becomes vertical in nature.  At a very  high arrival rate, average utilization of the Revenue Baggage Counters would approach 100%.  Also from the previous figure,  there would be an extremely long transit time.  Thus, the  average utilization for the Revenue baggage counters w i l l be in the range of 30% to 65% when the arrival rate is in the range of 40 to 90 passengers and 4 counters open. Figure 3 is a plot of the arrival rate versus the percentage of Revenue passenger units waiting longer than 2.5 minutes at the baggage counter. tal section.  This curve has a vertical and horizon-  When the arrival rate is in the range of 40 to 100  passengers, the Revenue baggage counter is able to process the passengers at such a rate that the 15 per cent rule is not violated.  In the range above 100 passengers the curve becomes  horizontal in nature.  The changeover occurs when approximately  13% to 15% of the passenger units exceed 2^ minutes of waiting time.  ARRIVAL RATE (PASSENGERS PER 15 MIN.)  EFFECT OF VARIOUS ARRIVAL RATES ON THE REVENUE BAGGAGE COUNTER UTILIZATION FACILITY POLICY CONSTANT  160  140  120  100 80  60  40  10  20  30  40  50  60  70  80  90  100  % AVERAGE UTILIZATION REVENUE BAGGAGE COUNTERS  EFFECT OF VARIOUS ARRIVAL RATES ON THE % OF REVENUE PASSENGER UNITS EXCEEDING 2% MINUTES WAITING TIME IN A QUEUE  ANNUAL 160 RATE (PASSENGERS 140 • PER 15 MIN.) 120 100 .  80 PRESENT SERVICE POLICY 15% 60  40  1  0  i 10  15  i  20  30  40  50  60  70  80  % REVENUE PASSENGER UNITS EXCEEDING 2h MINUTES WAITING TIME IN QUEUES  This curve then justifies the current service policy.  If  the percentage is increased, the system becomes very sensitive to the amount of passengers in the system for the f a c i l i t y policy (at arrival rate of 120 passengers, 49% of Revenue passenger units exceed 2.5 minutes).  The vertical portion of the curve represents  the range within which the f a c i l i t y policy produces stability in the system.  These results are consistent with the other graphs.  In Figure 4 the utilization and corresponding percentage of passenger units exceeding 2.5 minutes waiting time has been graphed (Figure 2 and Figure 3). For policy formulation, the effect of increasing the percentage of passenger units allowed to exceed 2.5 minutes waiting time on the f a c i l i t y utilization may be determined from the graph. Alternatively, this gives the trade-off of opposing system objectives. If the percentage was increased to 25%, the average utilization that could be expected would be 75%. 2.  Facility Policy Determination The second area of the analysis is to determine how the  model is able to formulate an optimal f a c i l i t y policy for a given set of conditions.  The given set was:  AVERAGE UTILIZATION 100%.  AVERAGE UTILIZATION AT THE REVENUE BAGGAGE COUNTERS VERSUS THE % OF REVENUE PASSENGER UNITS EXCEEDING 2h MINUTES WAITING TIME IN QUEUES FACILITY POLICY HELD CONSTANT  90 ' 80 70 60 50 40  SERVICE POLICY 15% REVENUE PASSENGER UNITS ALLOWED TO EXCEED 2\ MINUTES  30 20 10 50  60  70  80  % REVENUE PASSENGER UNITS EXCEEDING 2% MINUTES WAITING IN QUEUES  2.  The a r r i v a l r a t e s a r e i n the range o f 25 p a s s e n g e r s to the y e t undetermined  maximum c a p a c i t y o f the  system. 3.  The Passenger  c o m p o s i t i o n was  Revenue passengers 85% o f t o t a l passenger  arrivals  15% o f a r r i v e d n o n - t i c k e t e d (group) 10% used c u r b s i d e c h e c k - i n (group) 15% had excess baggage (passenger u n i t s )  Standby  passengers  15% a r r i v e d n o n - t i c k e t e d (groups) 15% had e x c e s s baggage (passenger u n i t s ) The  result  o f t h i s a n a l y s i s appear i n T a b l e 1 which  the f a c i l i t y p o l i c y f o r v a r i o u s a r r i v a l r a t e s . of the i n d i v i d u a l counters i s also  3.  reflects  The maximum c a p a c i t y  indicated.  O p e r a t i n g P o l i c y F o r m u l a t i o n and E v a l u a t i o n The a n a l y s i s undertaken here i s d i r e c t e d towards the o p e r a t i n g  p o l i c i e s o f A i r Canada f o r t h e t i c k e t c o u n t e r s , b o t h Revenue and Standby A few y e a r s ago, t h e o p e r a t i n g p o l i c y was changed such  that  baggage o p e r a t i o n s and t i c k e t purchases o f the Revenue passengers was  s e p a r a t e d from t h a t of t h e Standby  passengers.  Revenue passengers  FACILITY POLICY FOR THE VARIOUS ARRIVAL RATES ARRIVAL RATE TOTAL PASSENGERS PER 15 MIN. PERIOD  REVENUE COUNTERS  STANDBY  BAGGAGE  TICKET  25  2  1  1  1  50  4  2  1  1  75  4  2  2  100  5  3  2  125  6  4  150  7  5  175  8  5  200  9  5  CAPACITY REACHED PASSENGER COMPOSITION  BAGGAGE  CAPACITY. REACHED  CAPACITY REACHED  TICKET  1  9  1  11  2  1  13  2  1  15  2  1  16  1  17  CAPACITY REACHED 85% 10% 15% 15%  ,  MANPOWER REQUIREMENTS TOTAL  OF TOTAL ARE REVENUE OF REVENUE USED CURBSIDE HAD EXCESS BAGGAGE WERE NON-TICKETED ON ARRIVAL  paying excess charges were allowed to use the Standby ticket counter and Standby passengers under the same circumstances were allowed to use the Revenue ticket counters. Part A of this analysis, focuses on the impact of complete separation of Revenue and Standby ticket operations.  The  implementation of this policy should produce a higher u t i l i z a tion of f a c i l i t i e s and greater percentage of passenger units waiting longer than 2.5 minutes in queues.  This is so because  the Revenue passengers are limited to one fewer counters for service. The simulation model produced higher system statistics than the integrated policy was operational (see Table 2).  For  example the average utilization of the Revenue ticket f a c i l i t i e s increased from 28.3% to 45.5%. plete separation is necessary,  If the decision is made that comthe f a c i l i t y policies as stated  in Table 1 should produce stability in the system for the Revenue system. In Part B of this analysis we examine the operating policy at the Revenue ticket counter more closely.  Since the separation  as in Part A produced higher system statistics, the queuing d i s c i pline w i l l be changed to determine i f the percentage of Passenger units exceeding 2\ minutes waiting time can be lowered while maintaining separated operations.  REVENUE  INTEGRATED  T r a n s i t Time p e r passenger Unit Average  Utilization  SEPARATED  2.6 M i n . 28.2%  % o f Passenger U n i t s w a i t i n g l o n g e r than 2.5 minutes  2.85 Min. 49.9%  6.4%  13.8%  2.9 Min.  3.75 Min.  STANDBY T r a n s i t Time Average  utilization  49.2%  62.1%  % o f Passenger U n i t s w a i t i n g l o n g e r than 2.5 minutes  30%*  54.9%*  * 2 5 % + 5% would be a c c e p t a b l e but s i n c e t h e r e i s o n l y one Standby t i c k e t f a c i l i t y , a d e c i s i o n must be made t o e i t h e r expand o r r e v i s e the s e r v i c e p o l i c y a t the Standby t i c k e t counter.  Specifically, for  i n s t e a d o f u s i n g m u l t i p l e queues o r one queue  each agent, a s i n g l e queue w i l l be used.  A l l Revenue passengers  p u r c h a s i n g t i c k e t s o r p a y i n g excess charges l i n e up i n one queue. When they r e a c h the f r o n t o f the l i n e , the  agents becomes a v a i l a b l e .  T h i s i s d e p i c t e d i n F i g u r e 5.  they w a i t u n t i l  any one o f  Only then do they l e a v e the l i n e . The r e s u l t s o f the s i m u l a t i o n s com-  b i n e d w i t h the r e s u l t s o f P a r t A f o r comparative purposes a r e i n T a b l e 3.  COMPARATIVE STATISTICS FOR REVENUE TICKET COUNTER UNDER VARIOUS OPERATING POLICIES STATISTIC  STANDBY EXCESS ALLOWED (INTEGRATED)  REVENUE STANDBY SEPARATED MULTI-QUEUES SINGLE QUEUES  AVERAGE TRANSIT TIME PER REVENUE PASSENGER UNIT  2.6 MIN.  2.85 MIN.  2.75 MIN.  AVERAGE UTILIZATION (%)  48.2%  49.9%  42.9%  % PASSENGER UNITS EXCEEDING 2h MINS. WAITING TIME IN QUEUES OF TICKET COUNTER  6.4%  13.8%  4.5%  FIGURE 5 S I N G L E OR M U L T I P L E QUEUES AT T H E REVENUE T I C K E T COUNTER  CM  <^QUEUE  .  AGENT  AGENT  AGENT  1  2  3  •  .  <^QUEUE  ^/QUEUE  CO  REVENUE T I C K E T  M U L T I P L E QUEUE  S I N G L E QUEUE  GOES TO ANY AGENT  COUNTER  OPERATION  OPERATION  MOVEABLE GUIDE  \  V AGENT  AGENT  AGENT  1  2  3  REVENUE T I C K E T  COUNTER  RAILS  The transit time per passenger unit is lower and is comparable to the integrated operation.  The utilization is lower as  well as the percentage of Revenue passenger units exceeding 2.5 minutes of waiting time than the integrated.  Thus i f this par-  ticular method of queuing was adopted, separation of the Revenue/ Standby counter could be achieved with viable results. In Part C of this step in the analysis we examine the method of checking in Standby passengers.  Since there has been  a large number of passengers exceeding the 2% minutes of waiting time, the operating policy of the ticket and baggage operations at one counter w i l l be analyzed.  In effect, one agent would become  both a ticket and baggage agent.  When the passenger unit approaches  the service counter, a single agent w i l l issue tickets, check the baggage and collect the excess charges.  Since this question i n -  volves multiple operations and is not at present in effect the correct distribution of service times must be hypothesized.  The esti-  mated total service time w i l l probably be high which in this case is conservative since we are testing a completely new agent procedure and want to be very careful.  The flowchart of the operation  for a single agent is depicted in Figure 6. are given in Table 4.  The system statistics  YES  NO PURCHASE TICKET  CHECK IN BAGGAGE  PAY EXCESS CHARGE  ONE STANDBY AGENT PERFORMS BOTH BAGGAGE AND ' TICKET OPERATIONS  TABLE 4  VARIOUS OPERATING POLICIES AT STANDBY COUNTERS STATISTIC  TRANSIT TIME PER STANDBY PASSENGER (MIN.  INTEGRATED TWO BAGGAGE AGENTS ONE TICKET AGENT  2.9 Min.  REVENUE/STANDBY SEPARATED TWO BAGGAGE AGENTS THREE AGENTS EACH ONE TICKET AGENT PERFORMS BOTH TICKET AND BAGGAGE OPERATION 3.75 Min.  3.0 Min.  BAGGAGE COUNTER AVERAGE UTILIZATION (%) % PASSENGERS EXCEEDING 2h MIN. WAITING TIME IN QUEUES (%)  31.2%  30.3%  6.0%  5.8%  49.9%  62.1%  TICKET COUNTER AVERAGE UTILIZATION (%) % PASSENGERS EXCEEDING 2h MIN. WAITING TIME IN QUEUES  ) ) ( ) ( ) ( ) )  30.0%  54.9%  ( )  COMBINED UTILIZATION 44.7% % PASSENGERS EXCEEDING 2% MIN. 28.3%  If this operational procedure is adopted, the transit time (average) of Standby passengers remains relatively the same as when the Revenue/Standby ticket operations were integrated.  However,  the percentage of passengers who exceed 2.5 minutes waiting time is reduced to 28.3%.  Thus the service policy is maintained through  the total system at an arrival rate of 100 passengers and the aforementioned parameter specifications. These are just a few of the experiments that can be performed on the model.  The experiments presented hopefully have hit  on the major areas of interest and demonstrated the vast power of simulation as an airport management tool. A sample of the output from the computer model is given in Figure 7. the user.  This may be varied depending upon the requirements of  FIGURE 7 OUTPUT FROM COMPUTER THE MEAN ARRIVAL RATE WAS  100 PASSENGERS PER 15 MINUTE PERIOD.  REVENUE PASSENGER SYSTEM TRANSIT TIME IN THE SYSTEM AVERAGE TIME PER PASSENGER UNIT 2.7 MIN. AVERAGE TIME PER PASSENGER 2.9 MIN. REVENUE PASSENGER BAGGAGE SYSTEM NUMBER OF COUNTERS OPEN 5. AVERAGE UTILIZATION OF FACILITIES 60.1% AVERAGE WAITING TIME IN QUEUES 25 SEC. PER PASSENGER UNIT % OF PASSENGER UNITS IN QUEUES OVER 2.5 MIN. WAS 5.6% REVENUE PASSENGER TICKET SYSTEM NUMBER OF COUNTERS OPEN 3. AVERAGE UTILIZATION OF FACILITIES 50.3% (COUNTERS 11-15 ONLY) AVERAGE WAITING TIME IN QUEUES 31 SEC. PER PASSENGER UNIT (11 % OF PASSENGER UNITS IN ALL TICKET OR EXCESS QUEUES OVER 2.5 MIN. WAS 9.5% STANDBY PASSENGER SYSTEM TRANSIT TIME IN THE SYSTEM AVERAGE TIME PER PASSENGER UNIT 3.9 MIN. AVERAGE TIME PER PASSENGER STANDBY PASSENGER BAGGAGE SYSTEM NUMBER OF COUNTERS OPEN 2 AVERAGE UTILIZATION OF FACILITIES 38.6% AVERAGE WAITING TIME IN QUEUES 24 SEC. PER PASSENGER UNIT % PASSENGER UNITS IN QUEUES OVER 2.5 MIN. WAS 6.8% STANDBY PASSENGER TICKET SYSTEM AVERAGE UTILIZATION OF FACILITIES 70.8% AVERAGE WAITING TIME IN QUEUES 307 SEC. PER PASSENGER UNIT % PASSENGER UNITS IN QUEUES OVER 2.5 MIN. WAS 61.1%  

Cite

Citation Scheme:

        

Citations by CSL (citeproc-js)

Usage Statistics

Share

Embed

Customize your widget with the following options, then copy and paste the code below into the HTML of your page to embed this item in your website.
                        
                            <div id="ubcOpenCollectionsWidgetDisplay">
                            <script id="ubcOpenCollectionsWidget"
                            src="{[{embed.src}]}"
                            data-item="{[{embed.item}]}"
                            data-collection="{[{embed.collection}]}"
                            data-metadata="{[{embed.showMetadata}]}"
                            data-width="{[{embed.width}]}"
                            async >
                            </script>
                            </div>
                        
                    
IIIF logo Our image viewer uses the IIIF 2.0 standard. To load this item in other compatible viewers, use this url:
http://iiif.library.ubc.ca/presentation/dsp.831.1-0101735/manifest

Comment

Related Items