Support to Design for Air Traffic Management: An Approach with Agent-Based Modelling and Evolutionary Search

To enhance Air Traffic Management (ATM) and meet the future traffic demand and environmental requirements, present ATM system is going to be modified (SESAR Joint Undertaking, 2017), designing new services to be integrated in future architecture considering the evolution of present fragmented structure of the airspace and the entanglement of air routes. Such a change process is complicated due to the nature of ATM, which is a large-scale Socio-Technical System (STS), typically involving a complex interaction between humans, machines and the environment. In such kind of systems, managing their evolution is a complex and difficult task since the social and technical implications of any proposed concept should be fully assessed before a choice is made whether or not to proceed with the related development. Often, simulation tools are also used to support the design of the concept itself by enabling what-if-analyses. However, these may be too effort and time consuming due to the exponential growth of the required analysis cases. A quite common mismatch between the performance evaluations in simulated conditions and those achieved in real life is represented by the partial assessment of human aspects that can be performed throughout the new concept lifecycle from its lowest maturity level up to “ready to market”. The proposed work defines an approach to support the design of new ATM solutions, including the evaluation on human behaviour. The approach adopts a combined paradigm, which involves Agent-Based Modelling and Simulation (ABMS) to specify and analyse the ATM models, and Agent-based Evolutionary Search (AES) to optimize the design of the new solutions. A specific case study is used to demonstrate the effectiveness of the proposed approach. Transition from Direct Routing Airspace (DRA) to Free Routing Airspace (FRA), respectively described by Solution #32 and Solution #33 in the SESAR solutions catalogue (SESAR Joint Undertaking, 2017), is used for both validation and experimentation activities. In detail, the proposed experimentation case regards the design of sector collapsing/decollapsing configuration to optimize controller workloads. The achieved results are presented and discussed.