related work
related work
Embedded Files
Strong air traffic demand growth and capacity limitations have created significant challenges for airport planning, management and operations (de Neufville and Odoni, 2013). First, demand-capacity imbalances routinely lead to flight delays and cancellations, and passenger misconnections. Approximately 18% of the flights experience significant delays in the US and in Europe (FAA 2016; Eurocontrol 2016). In 2007, the total impact of delays in the US was estimated at over $32 billion (Ball et al., 2010). Moreover, limited capacity availability at busy airports imposes long-term economic costs of lost demand, high airfares, and restrictions in airlines’ route development. As air traffic is expected to continue growing fast in the next 20 years (4% annually), it will become ever more critical to design mechanisms for managing airport capacity in an efficient, reliable and sustainable manner.
Strong air traffic demand growth and capacity limitations have created significant challenges for airport planning, management and operations (de Neufville and Odoni, 2013). First, demand-capacity imbalances routinely lead to flight delays and cancellations, and passenger misconnections. Approximately 18% of the flights experience significant delays in the US and in Europe (FAA 2016; Eurocontrol 2016). In 2007, the total impact of delays in the US was estimated at over $32 billion (Ball et al., 2010). Moreover, limited capacity availability at busy airports imposes long-term economic costs of lost demand, high airfares, and restrictions in airlines’ route development. As air traffic is expected to continue growing fast in the next 20 years (4% annually), it will become ever more critical to design mechanisms for managing airport capacity in an efficient, reliable and sustainable manner.
Airport capacity management interventions are broadly classified as supply-side or demand-side. On the supply side, capacity expansion can take place through the construction of new landside and airside facilities. Santos and Antunes (2015) present an optimization approach to support aviation authorities at selecting the best long-term expansion interventions in an airport network. However, such interventions are generally very expensive and time consuming. At the operational level, Air Traffic Flow Management (ATFM) measures can minimize the effects of congestion through the adoption of ground-holding programs (Richeta and Odoni 1993; Lulli and Odoni 2007), the rerouting or metering of flights (Bertsimas et al. 2011a), and the selection of airport runway configurations (Bertsimas et al. 2011b; Jacquillat et al. 2017). These measures can mitigate the impacts of congestion, but are insufficient when demand exceeds airport capacity significantly.
Airport capacity management interventions are broadly classified as supply-side or demand-side. On the supply side, capacity expansion can take place through the construction of new landside and airside facilities. Santos and Antunes (2015) present an optimization approach to support aviation authorities at selecting the best long-term expansion interventions in an airport network. However, such interventions are generally very expensive and time consuming. At the operational level, Air Traffic Flow Management (ATFM) measures can minimize the effects of congestion through the adoption of ground-holding programs (Richeta and Odoni 1993; Lulli and Odoni 2007), the rerouting or metering of flights (Bertsimas et al. 2011a), and the selection of airport runway configurations (Bertsimas et al. 2011b; Jacquillat et al. 2017). These measures can mitigate the impacts of congestion, but are insufficient when demand exceeds airport capacity significantly.
On the demand side, airport capacity management involves interventions aimed at limiting the number of flights scheduled at peak hours. First, the airlines themselves can adopt strategies to minimize the negative effects of airport congestion by developing flight schedules that are less prone to flight delays and their propagation (Pita et al. 2012), or that make use of the available airport capacity efficiently (Pita et al. 2014). Second, the airports may implement demand management mechanisms, which provide adjustments to the flight schedules requested by the airlines to limit the number of flights scheduled at peak hours by distributing them more evenly and/or limiting their number over the day (Barnhart et al., 2012).
On the demand side, airport capacity management involves interventions aimed at limiting the number of flights scheduled at peak hours. First, the airlines themselves can adopt strategies to minimize the negative effects of airport congestion by developing flight schedules that are less prone to flight delays and their propagation (Pita et al. 2012), or that make use of the available airport capacity efficiently (Pita et al. 2014). Second, the airports may implement demand management mechanisms, which provide adjustments to the flight schedules requested by the airlines to limit the number of flights scheduled at peak hours by distributing them more evenly and/or limiting their number over the day (Barnhart et al., 2012).
Demand management mechanisms are of two types: economic and administrative. The former ones have attracted much attention from academic researchers (e.g., Fan and Odoni 2002; Ball et al. 2005; Gillen et al. 2016). However, in practice, their application is limited. In contrast, administrative mechanisms are used extensively. The main such mechanism, applied everywhere in the world except in the US, is the IATA slot allocation process (IATA, 2015). The guidelines for this process define the priority rules to be followed by a slot coordination entity when allocating slots to flight arrivals and departures. These guidelines are widely accepted by the airlines and airports, and are flexible enough to encompass airport-specific rules (Ulrich, 2008). Notwithstanding, they suffer from significant criticisms, mainly from some airports and from new airlines, which believe that the current rules impose stringent constraints on business development.
Demand management mechanisms are of two types: economic and administrative. The former ones have attracted much attention from academic researchers (e.g., Fan and Odoni 2002; Ball et al. 2005; Gillen et al. 2016). However, in practice, their application is limited. In contrast, administrative mechanisms are used extensively. The main such mechanism, applied everywhere in the world except in the US, is the IATA slot allocation process (IATA, 2015). The guidelines for this process define the priority rules to be followed by a slot coordination entity when allocating slots to flight arrivals and departures. These guidelines are widely accepted by the airlines and airports, and are flexible enough to encompass airport-specific rules (Ulrich, 2008). Notwithstanding, they suffer from significant criticisms, mainly from some airports and from new airlines, which believe that the current rules impose stringent constraints on business development.
From the literature on administrative demand management mechanisms, we highlight the integer optimization models for airport slot allocation from Zografos et al. (2012), Jacquillat and Odoni (2015) and Pyrgiotis and Odoni (2016). The latter two papers focus on the US context, thus do not consider the IATA slot allocation guidelines. In contrast, Zografos et al. (2012) do consider some of the IATA guidelines. However, it does not take into account important operational and regulatory constraints, including some priorities among slot requests, as well as terminal capacity, apron capacity, and noise restrictions.
From the literature on administrative demand management mechanisms, we highlight the integer optimization models for airport slot allocation from Zografos et al. (2012), Jacquillat and Odoni (2015) and Pyrgiotis and Odoni (2016). The latter two papers focus on the US context, thus do not consider the IATA slot allocation guidelines. In contrast, Zografos et al. (2012) do consider some of the IATA guidelines. However, it does not take into account important operational and regulatory constraints, including some priorities among slot requests, as well as terminal capacity, apron capacity, and noise restrictions.
From a computational standpoint, Zografos et al. (2012) tested their model at a small airport, Iraklion (Greece), which handles less than 50,000 aircraft movements per year. In this case, it could be solved (to exact optimality) using commercial solvers. For midsize airports (e.g., Porto airport, which handles nearly 100,000 aircraft movements per year), mathematical programming solvers will probably also work reasonably well, but formulation and tractability issues will certainly be a concern (Williams 2013). For large airports (e.g., Lisbon and Sao Paulo, with 175,000 and 300,000 movements per year, respectively), the model is expected to be computationally tractable only through the use of specialized methods like hybrid metaheuristics (Talbi 2013; Junqueira et al. 2013; Oliveira et al. 2014). These algorithms combine classical metaheuristics with mathematical programming, constraint programming or dynamic programming, and although they do not guarantee optimal solutions, they will provide near-optimal solutions in efficient computational times if properly built and calibrated.
From a computational standpoint, Zografos et al. (2012) tested their model at a small airport, Iraklion (Greece), which handles less than 50,000 aircraft movements per year. In this case, it could be solved (to exact optimality) using commercial solvers. For midsize airports (e.g., Porto airport, which handles nearly 100,000 aircraft movements per year), mathematical programming solvers will probably also work reasonably well, but formulation and tractability issues will certainly be a concern (Williams 2013). For large airports (e.g., Lisbon and Sao Paulo, with 175,000 and 300,000 movements per year, respectively), the model is expected to be computationally tractable only through the use of specialized methods like hybrid metaheuristics (Talbi 2013; Junqueira et al. 2013; Oliveira et al. 2014). These algorithms combine classical metaheuristics with mathematical programming, constraint programming or dynamic programming, and although they do not guarantee optimal solutions, they will provide near-optimal solutions in efficient computational times if properly built and calibrated.
This discussion highlights that an optimization approach to support slot allocation at major airports, as currently carried out according to the IATA guidelines, does not yet exist. This constitutes a major gap in the state-of-the-art. The development of such an approach will undoubtedly be of great utility to slot coordination entities. Moreover, a detailed assessment of the impacts of possible changes to the IATA slot allocation guidelines is still to be performed. The optimization approach would facilitate this assessment, by enabling to design potential adjustments in the guidelines and compute the best possible corresponding outcomes, and quantify the resulting benefits for the different airport stakeholders.
This discussion highlights that an optimization approach to support slot allocation at major airports, as currently carried out according to the IATA guidelines, does not yet exist. This constitutes a major gap in the state-of-the-art. The development of such an approach will undoubtedly be of great utility to slot coordination entities. Moreover, a detailed assessment of the impacts of possible changes to the IATA slot allocation guidelines is still to be performed. The optimization approach would facilitate this assessment, by enabling to design potential adjustments in the guidelines and compute the best possible corresponding outcomes, and quantify the resulting benefits for the different airport stakeholders.
references
references
Report abuse