Future UDPP Concepts

Context

During the Air Traffic Flow and Capacity Management (ATFCM) tactical phase, when the capacity of an en-route sector or at the destination airport is expected to be exceeded by the demand, flights are delayed at the origin airport and assigned new take-off times through ATFM slots. This results in the so-called ATFM delay, which represents a significant cost for airlines, passengers, and the ATM system as a whole. The policy traditionally used to allocate ATFM delay is the First Planned First Served (FPFS) principle. FPFS minimises the total delay, but due to the cost of delay being highly non-linear and varying from one flight to another, it may not be the optimal solution from the point of view of Airspace User (AU) costs. The possibility of rearranging flight sequences when facing situations of demand-capacity imbalance offers remarkable potential to reduce the impact of ATFM delay. To realise this potential, SESAR developed the User Driven Prioritisation Process (UDPP). Early UDPP developments introduced a number of mechanisms that provide AUs with increased levels of flexibility to prioritise their flights and distribute delay according to its impact on AU operations and costs. However, the level of flexibility provided by these basic UDPP concepts is still relatively low. Other, more flexible flight prioritisation instruments are proposed in the literature, including market mechanisms, but their implementation is hindered by the difficulties to design, test and validate such instruments. Classical modelling approaches from economics and operations research, such as game theory and linear programming, have been used for this purpose, but the strong assumptions behind these approaches, such as agent rationality and perfect information, make such models unrealistic in certain circumstances, which may lead to overlooking the risks and the potential unintended consequences of certain mechanisms when stakeholders’ behaviour departs from these rigid assumptions.

The project

The project “Exploring Future UDPP Concepts through Computational Behavioural Economics” aimed at developing new modelling approaches to enable a rigorous and comprehensive study of highly flexible, advanced UDPP mechanisms. To this end, the paradigm of computational behavioural economics was adopted, as a particularly suitable framework for the representation of features that are not properly captured by classical approaches, such as bounded rationality, evolutionary behaviour, and asymmetric, imperfect and uncertain information.

Goals

The project had the following specific objectives:

1. Develop an assessment framework for the comprehensive evaluation of the impact of flight prioritisation mechanisms on network performance and on ATM stakeholders, including aspects such as their ability to ensure equity and their resilience and robustness in the presence of irrational or strategic behaviour of AUs.
2. Conduct a detailed review of the tactical slot and trajectory allocation mechanisms proposed in the literature and identify the most promising ones to improve UDPP.
3. Create an agent-based model that enables the evaluation of different flight prioritisation mechanisms based on the proposed assessment framework.
4. Run a set of simulation experiments, considering different AUs’ behavioural assumptions, in order to conduct a systematic assessment and comparison of the identified flight prioritisation mechanisms and derive conclusions on the advantages and disadvantages of each of the proposed mechanisms.

Results

The project developed an agent-based model that simulates a day of operations, where the Network Manager takes care of flow management and the airlines make decisions on how to deal with the delays imposed in congestion situations. The main features of the model are summarised in the deliverable D3.1 “Agent-Based Simulation Model for the Analysis of Tactical Slot and Trajectory Allocation Mechanisms”.

The set of scenarios selected for the simulations was the result of the combination of different configuration variables, including the air traffic conditions and the airlines’ behavioural rules. From the analysis of the experimental results, one can see how the different flight prioritisation mechanisms impact the selected KPIs. More information can be found in D4.1 “Results of Simulation Experiments: Comparative Analysis of Different Tactical Slot and Trajectory Allocation Mechanisms”.

Several conclusions were drawn regarding the suggested modelling approach. Emergent and counterintuitive phenomena which would have been ignored otherwise were identified for some scenarios, unveiling the added value of agent-based modelling. The outcomes were noticeably impacted by network effects, underscoring their relevance in assessing prioritisation mechanisms. Conversely, the results displayed a high degree of sensitivity to specific modelling assumptions such as re-routings, airline behaviours, or the implementation of the CASA algorithm. A more extensive investigation of these aspects was conducted in the BEACON project.

Public deliverables

• D1.1 UDPP assessment framework: indicators and metrics.
• D2.1 Tactical slot and trajectory allocation mechanisms: qualitative assessment.
• D3.1 Agent-based simulation model for the analysis of tactical slot and trajectory allocation mechanisms.
• D4.1 Results of simulation experiments: comparative analysis of different tactical slot and trajectory allocation mechanisms.

Journal articles or scientific papers

• Paper SIDs 2020: Evaluation of flight prioritization mechanisms through agent-based modelling.
• ATM Seminar: Exploring Future UDPP Concepts through Computational Behavioral Economics.

Conference talks or Presentations

• Engage Workshop 2019: Exploring UDPP Concepts through Computational Behavioural Economics.