Elsa wp e new

University of Palermo

Coordinated by:

Deep Blue

Consortium members:

Scuola Normale Superiore di Pisa

Which are the methods to analyse and monitor the emerging characteristics of the ATM system?

Research Questions Can the application of complex system science to ATM provide means to analyse and monitor the ATM system?

The ELSA project analyse, describes and models the dynamics of the ATM system, especially those concerning the propagation of network disturbances. The analysis will be carried out in the current scenario (based on real data) and in a SESAR scenario (based on an Agent Based simulation).

Research Scope

1

The project aimed to: (i) perform an extensive statistical analysis of ATM system data with Complex Systems theory techniques in order to characterise statistical regularities; (ii) develop an Agent Based Model of increasing complexity and degree of realism, to simulate the emergent properties of the trajectory-based SESAR scenario; (iii) design a prototype of a decision support tool, to monitor, predict, and intervene on the air traffic.

The ELSA Agent Based Model reproduces some of the stylised facts observed in the Air Traffic Management of the European airspace. The model itself has two main parts: (i) The strategic layer, focused on the interaction between the Network Manager and the Airline Operators and (ii) the tactical layer, focused on aircraft and controllers behaviour in Air Traffic Control (ATC) sectors.

The decision support tool describes possible workflow and interactions among the Network Manager, the Flight Operation Centre, the Complexity Manager in a SESAR scenario. Candidate technological solutions, humaninterface elements, and interaction modalities are proposed.

The data analysis was performed on 14 months of ECAC traffic data. The main result is a list of “best analyses”, including the following: 1. Community analysis: seasonal variations 2. Community analysis: impact of the ash cloud 3. Analysis of ATCO management strategies: differences among countries and sectors 4. Analysis of the correlations between STCA events and Complex Network metrics in the Rome FIR 5. Analysis of the most critical navpoints of the Italian airspace

Research Results

2

The preliminary results of the strategic layer of the agent based model show that the “low-cost” and “traditional” Airline Operators have different advantages, depending on the departure patterns and the level of mixture.

Even if it is always better for companies to be surrounded by companies of the other type – hence displaying a “the loner, the better” effect - the dominant strategy might also depend on the distribution of departure times, in a non-trivial way. For instance, on the contrary of the general case, it is sometimes better for “low-cost” companies to have a more uniform distribution. Finally, we find that the advantage of “low-cost” companies over “traditional” companies increases in a constrained environment, leading to a higher resilience for the type of shocks we considered.

The preliminary results for the tactical layer indicate that when shocks in the system are confined in small areas, the interplay between the rerouting and change of flight level strategies may lead to trajectory modifications that give smaller average delays, as long as the number of shocks increases.