DESIGN AND DEVELOPMENT OF AN INTEGRATED COMPUTATIONAL PLATFORM FOR TOTAL AIRPORT PERFORMANCE ANALYSIS By Konstantinos G. Zografos Professor and Michael A. Madas Research Associate Athens University of Economics & Business Department of Management Science & Technology TRANsportation Systems & LOGistics Laboratory (TRANSLOG) Evelpidon 47A & Lefkados 33, Athens, 113 62 Tel. +30 210-82.03.673-5, Fax +30 210-82.03.684, email:
[email protected]
Abstract: The increasing mismatch between airport demand and supply in conjunction with technical, physical, and political constraints in providing sufficient capacity has stimulated vigorous policy discussions towards assessing and closely monitoring the airport performance. Nowadays, there is an urgent need for a decision support system that will allow decision makers and analysts to evaluate the efficiency of the entire airport complex simultaneously by considering implications and interdependencies between various measures of airport effectiveness. The objective of this paper is threefold: i) to introduce the concept of total airport performance analysis, ii) to describe an operational concept for the integration of analytical and simulation models in simultaneously analysing airside and landside, and iii) to present an overview of the capabilities and functionalities of the computational platform. Keywords: airport performance analysis, airport modelling, airport planning.
1. INTRODUCTION & RESEARCH MOTIVATION Among all different transport modes, air transport has shown by far the largest traffic increase and growth potential since the last two decades. Based on anticipated demand figures, an average growth rate of 4.3% per annum is still envisaged for the period 2002-2015 in EU Member States, which will have to be mostly accommodated by a relatively small number of primary hub airports (European Commission, 2001). A direct consequence of the growing demand and the resulted mismatch between demand and supply of air transport (and airport) services is the increase of congestion problems both on the air and on the ground (at airports) with considerable externalities associated with this lack of capacity. In 2001, one flight over four was delayed due to air traffic problems (one flight over six was late in 2000 with an average delay of 22 minutes (European Commission, 2001) with total annual ATFM delay costs for airlines in 2000 amounted between €1.3 and €1.9 billion (EUROCONTROL, 2001). As a matter of fact, a sharp and steady rise increase in delays has taken place during the last few years with the exception of the temporary relief to the congestion problem due to the industry crisis after the tragic events of the 11th of September 2001 terrorist attack. In that respect, passengers anxiously demand for a better quality of service through lower delays and prices and with exceptionally high safety and security standards, airport operators seek for improved airport efficiency through lower delays and expansion of airport infrastructure to tackle with saturated airport capacities, while policy makers stress the need for planning, performance monitoring, and implementing airport environmental endeavours. At present, airport stakeholders lack insight into the integrated set of airport processes, decisions, and their interdependencies in order to cope with the mismatch between demand and supply and effectively tackle with the trade-off’s and interdependencies between the various airport performance metrics (e.g., capacity, delays, level of service, safety / security standards, environmental performance, noise, cost-effectiveness). Therefore, there is an urgent need for a decision support system that will allow decision makers and analysts to evaluate the efficiency of the entire airport complex simultaneously by considering the whole spectrum of measures of airport effectiveness. The objective of this paper is threefold: i) to introduce the concept of total airport analysis, ii) to describe a system architecture and an operational concept proposed for the integration of analytical and simulation models in simultaneously analysing airside and landside operations, and iii) to present an overview of the capabilities and functionalities of the herewith presented computational platform for total airport analysis and optimisation. The remainder of this paper is structured into seven sections. Section two elaborates on the need and motivation for such a decision support tool for total airport performance analysis, and Section three introduces the integrated platform for total airport performance analysis, its overall architectural design, as well as the functionalities and alternative scopes of platform uses by the potential user groups. Section four deals with the conceptual description of the model / tool classification and selection framework, while Section five addresses the model base design along with their associated modules and functionalities. Finally, the paper is complemented by a concluding chapter (Section six), acknowledgements (Section seven), and the list of references (Section eight).
2. STATE-OF-THE-ART AND STATE-OF-PRACTICE REVIEW The comprehensive amount of research that has been devoted on airports and air transport, allows capitalising on existing results as regards: i) well defined requirements and data for airport planning, analysis, and modelling, ii) analytical and simulation models and tools for
the performance assessment and monitoring of airport operations, and iii) advanced methods for planning and optimising the airport system and network (Odoni, 1991; Odoni et al., 1997; Andreatta et al., 1999; TRB 2000; OPAL, 2001a; APRON, 2003). The state-of-the-art and state-of-practice review (OPAL, 2001a; Zografos, 2003; THENA, 2003) clearly suggest that currently available models and tools provide decision support to all levels of decision making (i.e., strategic, tactical, operational) and for all types of airport management decisions (i.e., planning, design, operations) and types of problems / studies (i.e., capacity, delay, noise, security, etc.). Furthermore, existing models and tools successfully capture a number of strategic trade-off’s between the various measures of airport effectiveness with respect to all airport elements (i.e., airside, landside) and flows / entities involved (i.e., aircraft, passenger, baggage, cargo). However, despite the rich experience in both models and tools and their specific applications to support strategic, tactical, and operational decisions for airports, there seems to exist a considerable knowledge gap stemming from the following problems and methodological limitations identified through this comprehensive review (OPAL, 2001a). These problems relate to: The sharp division between available airside and landside models. This means that currently there is no single airport analysis model that can perform analysis of the entire airport complex including airport access, landside, and airside with the combinatorial use of analytical and simulation models / tools of different levels of decision making (i.e., strategic, tactical, operational) and level of aggregation. Recent and on-going research endeavours (e.g., TAPE, OPTAS, LEONARDO) practically attempted the communication and interaction of pre-selected tool configurations / model pairings in order to model and evaluate simultaneously airport airside and landside and assess their interdependencies. In this direction, similar efforts have been documented in the literature (Zografos and Stamatopoulos, 1998; Andreatta et al., 1999) that suggested and introduced an integrated use of airside and landside models for airport strategic planning applications. Nevertheless, the common denominator and weakness of these efforts is the lack of a harmonized, integrated, and fully-automated computing environment needed to execute the various models and report (with post-processing capabilities) their integrated results. The limited integration between models / tools able to provide capacity and delay estimates and models / tools able to provide environmental, and safety analysis. The lack of harmonized / standardized measures of effectiveness addressing the performance of the airside and landside airport elements. The lack of integration of models addressing cargo and passenger handling activities. The limited applicability and usefulness of available models and tools to high-level decision making in airport planning and design, on the grounds that they are quite complicated and rather data-intensive, while necessitating substantial tool familiarity and computational expertise (OPAL, 2001a). As a result, it is urgently needed to bridge this gap by developing a seamlessly integrated computing environment that will address a number of very important airport planning and operations decisions in a user-friendly manner without requiring extensive training or prior user familiarity with the selected tools or tool combinations.
3. DEVELOPMENT OF AN INTEGRATED COMPUTATIONAL PLATFORM In response to the voids and gaps and research needs identified in the previous section, an integrated computational platform was developed and validated within the framework of the “Optimisation Platform for Airports including Landside (OPAL)” research project funded by the European Commission (DG TREN, 2000-2002), whose the major objective was to
provide a generic concept for the development of a decision support system for total airport performance analysis. Within the framework of the OPAL project, a user-friendly, integrated, and distributed computational environment has been developed to enable airport stakeholders to model, evaluate, and optimise the entire flow efficiency by simultaneously modelling all components of the airport complex. It has the form of a computational facility for total airport performance analysis through the integrated use of various analytical and simulation models that allow the user to carry out a diverse range of airport studies exploring the impacts of airport planning decisions on a wide spectrum of airport performance metrics such as capacity, congestion and delay, level of service, safety, cost-effectiveness, and environment. The OPAL platform enables the combined and integrated use of selected (mostly available, but also newly developed) airport analysis tools (i.e., analytical, simulation) based on a distributed computational environment consisted of (Figure 1): i) a Central Database that enables the use of common data elements and the communication between tools integrated into the platform, ii) an overall Human Machine Interface (HMI) that enables the access to and visualisation of the input requirements and output / results, iii) a Model Base that consists of selected models / tools and model / tool configurations capable of performing trade-off analyses between various measures of airport effectiveness, iv) a Diagnostics Tool that enables the analysis of propagated impacts and interdependencies and identifies the overall bottlenecks in a total airport system, v) an Optimisation Tool that enables the user to optimise the airport configuration given a particular scenario, the models used and a specific optimisation criterion, vi) Data Converters that enable the integrated tools to read from and write to the central database of the platform, and vii) a Scenario Manager that executes, monitors, and controls the tools within the platform, integrating them into a harmonized virtual working environment for the platform users.
Figure 1. OPAL Platform Architecture (OPAL, 2001b) Through its airport performance analysis and evaluation capabilities, the OPAL platform supports the decision making of airport authorities / operators, airlines, ATS/ATC providers, as well as governmental and policy / regulatory authorities (e.g., EUROCONTROL, European Commission, Civil Aviation Authorities). The platform supports the search for solutions to problems faced by airports since it allows the user to perform "what-if" studies with respect to “frequently asked airport planning questions” related to the impacts of new operational
procedures, investments in infrastructure expansion, advancement in technological procedures, changes in the traffic profile and volumes and, in general, the implications of alternative “what-if” scenarios on a broad spectrum of airport performance metrics (i.e., capacity and delays, level of service, safety, environment / noise, costs and benefits). A proofof-concept of the OPAL platform has been already reported through the platform validation at six major European airports: Amsterdam-Schiphol, Athens, Frankfurt, Madrid-Barajas, Palma de Mallorca, and Toulouse-Blagnac (OPAL, 2002).
4. MODEL / TOOL SELECTION One of the major challenges that should be tackled in order to develop this integrated computational platform was the design of the appropriate model / tool base along with the associated model / tool configurations. In accomplishing this objective, the design phase should follow the methodological steps presented in Figure 2. User Requirements (i.e., types of airport studies)
State-of-the-Art Review
State-of-Practice Review
Development of an Inventory of Airport Performance Models / Tools
Classification of Models / Tools
Model attributes & characteristics Performance metrics examined Airport elements addressed Level of decision making covered Level of detail / data aggregation
Evaluation of Suitability of Available Tools for Integration into the Platform Integration Requirements
Model / Tool Combinations
Model / Tool Selection
Integration Requirements
Modules' Specification
Design of Model Base
Figure 2. Model Classification & Selection Approach The above presented model classification and selection approach ensures the accomplishment of the identified user requirements by sequentially implementing the following methodological steps: i) to develop an inventory of airport performance models based on a comprehensive state-of-the-art and state-of-practice review, ii) to review and systematically classify the available models and tools on the basis of the particular model attributes and performance metrics captured, iii) to evaluate and determine the most appropriate set of models / tools capable of performing the specified trade-off analyses between the various
performance metrics by addressing the various airport entities involved (e.g., aircraft, passengers, baggage, freight) and airport elements covered, iv) to identify “gaps” and requirements for model enhancements or developments needed to secure that the elicited user requirements have been addressed by the platform under development, v) to define the communication and integration requirements between the models / tools comprising the various model combinations, and vi) to design the overall architecture of the model base in the form of the individual models and tools, the platform modules, as well as their specifications (Figure 2). Based on the previously described process of identification and classification of available models and tools, the inventory of airport performance models was systematically structured and classified on the basis of the classification attributes presented in Figure 2. This classification scheme identifies the prime candidate models to be incorporated into the OPAL model base, and enables the design and specification of the corresponding OPAL modules. Simultaneously, voids and gaps in terms of models / tools capable of performing particular types of analyses were identified. These gaps and inefficiencies identified in the previous stage were the drivers to the design and development of new tools capable of capturing the requirements of particular types of studies not currently addressed by available models and tools. Therefore, two new tools were developed and incorporated in the platform: i) MACS freight model, and ii) the Cost-Benefit Model (CBM). As far as the analytical models are concerned, MACAD tool (Zografos et al., 1998; Stamatopoulos et al., 2003) represents a prime candidate for inclusion in the OPAL model base, as it efficiently deals with each aspect of airport airside capacity. MACAD integrates macroscopic airside models and provides an overall assessment of the capacity and level of service (e.g., delays) of the airside elements and facilities (e.g., apron, runway, taxiway). In general, it provides an overall assessment of the airside operations and can be efficiently integrated with SLAM for total airport analysis. With respect to landside elements, SLAM is the only analytical model for modelling the landside part of an airport at a macroscopic, strategic level by providing estimations of capacity and delays for passengers and baggage in the landside process facilities. In addition, SLAM has documented interaction capabilities with macroscopic models of the airport airside, and particularly MACAD (Zografos and Stamatopoulos, 1998; Andreatta et al., 1999). In conclusion, MACAD-SLAM clearly constitute the most suitable analytical tool combination for the total airport capacity and delays analysis. The Integrated Noise Model (INM) assesses the predicted noise levels around airports during the departure and arrival operations, and it seems the most suitable analytical model for tactical and operational environment analysis. Similarly, ENVIRA is a detailed analytical model that estimates the noise exposure contours around airports, while it can be effectively integrated with the External Safety Analyzer (ESA). Furthermore, the External Safety Analyser (ESA) estimates the external risk due to aircraft accidents in the vicinity of airports and it is the major candidate for the model base. Finally, the Cost Benefit Model (CBM) aims to assess the costs and benefits that arise from various measures in the form of an aggregate airport management information system, while MACS models the freight terminal activity and analytically computes the capacity pertaining to a given freight demand. For tactical or operational level in the airside, the three prime candidates are SIMMOD, TAAM, and Airport Machine. TAAM is the most fully featured, while the most freely available and widely used is SIMMOD, both covering operational and strategic level of decision making with Airport Machine covering mostly operational level of decision making. For tactical / operational landside capacity analysis, POWERSIM seems to be an appropriate choice, while WITNESS-MODA Airport Modelling (including its enhancement), is a microscopic model covering all passenger landside flows and processes, and represents an also quite promising option for the airport terminal capacity and delay analysis. In addition, POWERSIM (System Dynamics) is claimed to be able to capture the ground access elements
of an airport. PAX model describes passenger flows (departure, arrival, transfer) in the terminal, focusing on passport control, while BAX describes, on a high aggregation level, the main part of the baggage handling system from check-in to the laterals. The BAX/PAX models (and concepts) have been validated and used in studies performed in AmsterdamSchiphol Airport. Finally, TOPAZ seems the most appropriate choice for strategic safety assessment in the airport airside (OPAL, 2001b).
5. DESIGN OF THE PLATFORM MODEL BASE Based on the classification and selection of the appropriate set of models capable of fulfilling the elicited user requirements (OPAL, 2001b), the OPAL platform modules and the overall model base architecture were designed. The OPAL model base is consisted of five interacting modules, each of them addressing different areas of study and capturing specific elements of the airport operations. The Capacity and Delays module includes analytical and simulation models, addressing both airside and landside, and provides estimates for the capacity and delays of a total airport. MACAD and SLAM constitute the analytical part of the described module, while SIMMOD, TAAM, POWERSIM, WITNESS-MODA, and PAX / BAX represent the simulation part. SIMMOD and TAAM focus on the airside, while POWERSIM, WITNESS-MODA and PAX / BAX study and analyse the landside elements of the airport. The Environment module incorporates both analytical models (mostly INM) and simulation tools (NOISIM), assessing the predicted noise exposure and levels around airports. The Safety module includes analytical and simulation models, addressing only airside elements of the airport, and provides safety assessment and estimates of the external risk due to aircraft accidents. External Safety Analyser (ESA) represents the analytical part with TOPAZ constituting the simulation part. The Freight module includes only analytical models (i.e., MACS) for freight management. The Cost/Benefit module includes only analytical models and includes the newly developed CBM, which provides assessment of airport costs/benefits.
SAFETY MODULE
ENVIRONMENT MODULE AIRSIDE
AIRSIDE ANALYTICAL ESA
LOCAL DATA BASE
SIMULATION TOPAZ
ANALYTICAL ENVIRA INM
LOCAL DATA BASE
O P A L
SIMULATION NOISIM
COST / BENEFIT MODULE AIRSIDE ANALYTICAL COST/BENEFIT MODEL
GLOBAL COMMON DATA BASE
H M I CAPACITY & DELAYS MODULE
PASSENGER
FREIGHT LANDSIDE MACS
AIRSIDEAIRSIDE LOCAL DATA BASE
ANALYTICAL MACAD
SIMULATION SIMMOD TAAM
LANDSIDE SIMULATION ANALYTICAL POWERSIM SLAM WITNESS PAX/BAX
Figure 3. Overall OPAL Platform Architecture (OPAL, 2001b) All these modules should interact and communicate to each other within the framework of the overall architecture of the model base presented in Figure 3. As it can be inferred from the graphical description of the architecture, each particular model / tool combination has been implemented by means of local databases and data converters that communicate interchangeably the data / input requirements to each tool. Similarly, results and data pertaining to particular tool combinations are made available to other modules and tools through the global database, which represents the common reference and interaction mechanism between all modules and tool combinations. The Capacity and Delays module represents the core module in the model base architecture, which is further decomposed into Passenger and Freight sub-modules. Finally, all three Capacity and Delays, Safety, and Environmental analysis modules "feed" the Cost / Benefit module with input required in order to analyse and assess the net operational costs and benefits for airport and airlines.
6. CONCLUDING REMARKS The boom in air travel and the amplification of the demand for air transport is exacerbating severe problems to the airport system in the form of saturated airport and en route capacity and outstanding delays. The increasing demand and the associated externalities, in conjunction with technical, physical, and political constraints in providing sufficient capacity has stimulated vigorous policy discussions towards assessing and closely monitoring the airport performance with respect to various measures of airport effectiveness. In response to these policy dictates, airport operators urge for improved airport efficiency and expansion / enhancement of the airport infrastructure to tackle with saturated airport capacities, while simultaneously confronting excessive delays, respecting exceptionally high safety operational standards, and promoting airport environmental-driven initiatives. Therefore, there is an urgent need for a decision support system in the form of an integrated platform for total airport performance analysis that will allow decision makers and analysts to evaluate the efficiency of the entire airport complex simultaneously by considering the whole spectrum of measures of airport effectiveness. In response to this need, an integrated computational platform was developed and validated within the framework of the OPAL research project funded by the European Commission, whose the major objective was to provide a generic concept for the development of a decision support system for total airport performance analysis. This platform has the form of a computational facility for total airport performance analysis through the integrated use of various analytical and simulation models that allow the user to perform "what-if" studies in response to “frequently asked airport planning questions”.
7. ACKNOWLEDGMENTS Part of the presented research work has been supported by the “Optimisation Platform for Airports including Landside (OPAL)” research project funded by the European Commission (Directorate General Transport and Energy), 2000-2002. 8. REFERENCES Andreatta G., L. Brunetta, A.R. Odoni, L. Righi, M.A. Stamatopoulos, and K.G. Zografos (1999). A Set of Approximate and Compatible Models for Airport Strategic Planning on Airside and Landside. Air Traffic Control Quarterly, Vol. 7, Number 4, pp. 291-319.
APRON Consortium (2003). Information Requirements’ Analysis. Technical Report, Aviation Policy information Resources based on Observatory Networks (APRON) Research Project funded by the European Commission (Directorate General TRansport and ENergy - DG TREN), European Commission, Brussels, Belgium. European Commission (2001). European transport policy for 2010: time to decide, White Paper, European Commission, Brussels, Belgium. European Organisation for the Safety of Air Navigation (EUROCONTROL). Performance Review Commission (2001). Performance Review Annual Reports: An Assessment of Air Traffic Management in Europe during the Calendar Year 2000, Annual Performance Review Reports, Eurocontrol, Brussels, Belgium. Odoni, A.R. (1991). Issues in Modeling a National Network of Airports. Proceedings of the 1991 Winter Simulation Conference, Barry L. Nelson, W. David Kelton, Gordon M. Clark (editors), Phoenix, AZ, 8-11 December 1991. Odoni, A.R., J. Deyst, E. Feron, R.J. Hansman, K. Khan, J. Knchar and R.W. Simpson (1997). Existing and Required Modelling Capabilities for Evaluating ATM Systems and Concepts, International Center for Air Transportation, Massachusetts Institute of Technology. OPAL Consortium (2001a). Inventory of Airport Performance Models, Technical Report, Optimization Platform for Airports including Landside (OPAL) Research Project funded by the European Commission (Directorate General TRansport and ENergy - DG TREN), European Commission, Brussels, Belgium. OPAL Consortium (2001b). Design of OPAL Model Base, Technical Report, Optimization Platform for Airports including Landside (OPAL) Research Project funded by the European Commission (Directorate General TRansport and ENergy - DG TREN), European Commission, Brussels, Belgium. OPAL Consortium (2002). OPAL Evaluation Report, Technical Report, Optimization Platform for Airports including Landside (OPAL) Research Project funded by the European Commission (Directorate General TRansport and ENergy - DG TREN), European Commission, Brussels, Belgium. Stamatopoulos M.A., K.G. Zografos, and A.R. Odoni (2003). A Decision Support System for Airport Strategic Planning. Transportation Research Part C, (to appear). THENA Consortium (2003). Position Paper on Airport Capacity & Efficiency Issues, Technical Report, Thematic Network on Airport Activities (THENA) Research Project funded by the European Commission (Directorate General TRansport and ENergy - DG TREN), European Commission, Brussels, Belgium. Transportation Research Board (2000). Airport-Airspace Simulations for Capacity Evaluation, TRB 79th Annual Meeting Workshop organized by the TRB Committee on Airfield and Airspace Capacity and Delay (A1J05), Transportation Research E-Circular Number E-C035, Transportation Research Board (TRB), National Research Council, Washington, D.C.. Zografos, K.G., M.A. Stamatopoulos (1998). Using Analytical Models for Evaluating Airport Airside Performance, Proceedings of the 77th Transportation Research Board (TRB) Annual Meeting, A1J05 Committee on Airfield Capacity and Delay, Washington, D.C., January 1115, 1998. Zografos, K.G., M.A. Stamatopoulos, and A.R. Odoni (1998). Modeling Airport Strategic Planning Decisions, Proceedings of the International Conference Air Transport and Airports, Athens, December 3-4, 1998, pp. 123-142. Zografos, K.G. (2003). Research and Development Efforts in the Aviation Sector in Europe, Presentation on the 82nd Transportation Research Board (TRB) Annual Meeting, Session 257 (International Aviation Research) sponsored by A1J05 Committee on Airfield Capacity and Delay, Washington, D.C., January 13, 2003.
