Skip to main content
European Commission logo print header

Adaptive self-Governed aerial Ecosystem by Negotiated Traffic

Periodic Reporting for period 4 - AGENT (Adaptive self-Governed aerial Ecosystem by Negotiated Traffic)

Reporting period: 2017-08-01 to 2018-01-31

ATC interventions to preserve safety distances between aircraft have some inherent shortages when scalability problems arise, that can lead to well-known sector saturation and frequent separation infringements in front of an increment of the traffic demand. To address the safety-net drawbacks in future air traffic, the project AGENT claims for a collaborative and proactive separation management system considering the socio-technological aspects in which both human behavior and automation will be aligned. The concept seeks for an extended time horizon in the conflict resolution process to achieve better compatibility among safety nets.
AGENT aims at contributing to enhancement of the overall performances of the air traffic system, mainly targeting the separation management and its connections with collision avoidance. The system performance level shall be primarily increased by maintaining safety, and improving the efficiency for all agents in the system.
AGENT addresses the following key objectives:
• To propose an innovative automation-based future system design supporting a shift from a centrally controlled ATM system to a distributed system, in which aircraft and ATC collaborate to form adaptive aerial ecosystems, to find an optimal compromise in resolution.
• To develop ATC- and Pilot-side decision making tools to transform aircraft into intelligent agents that can communicate with each other using machine-to-machine interfaces.
• To build ontology for machine-to-machine communication between intelligent agents.
• To verify the AGENT tools using real traffic information and validate results by means of the scenarios in a simulated environment.
• To demonstrate and quantify the potential for the innovative ATM design providing benefits in safety, capacity and operational efficiency, and ensuring a wider acceptance of the research results.
WP1 was focusing on definition of the Concept of Operations relying on the spatiotemporal analysis of the aircraft interdependencies, the computation of a set of conflict resolution maneuvers, and an ontology to support an efficient negotiation mechanism for a resolution consensus. A background for the evaluation procedure was comprehensively analyzed, as well as the metrics based on defined procedure. The evaluation was planned as an automatic simulation to support a wide range of scenarios. The main part of the designing procedure was the Open Demonstrator release, comprising definition of its functionalities and services together with the interoperability features, the input and output data structure.
WP2 had elaborated the state space analysis methodology to support different decision-making tools as services within the Open Demonstrator. The efforts had also foreseen creation of the multi-agent simulation framework, its formalization and structural elements. The attention was given to the satisficing method for the proposed agents’ solutions as a driver of the system acceptability, criticality and scalability. The work on the Open Demonstrator release fully covered the generation of trajectories that feed the simulation platform, and characteristics of the scenarios created from extracted traffic. The main part was testing the inputs and outputs, as well as the parameters for the modules execution.
WP3 was realized in two directions: release of the conflict/collision probabilistic forecasting tool and generation of the resolution trajectory algorithm. The work covered the functionalities of the Open Demonstrator modules: conflict detection, ecosystem identification and compulsory resolutions generation. The conflict/collision probabilistic forecasting tool was required to automatically generate input scenarios, output the probabilistic model and datasets integrated into the Open Demonstrator. The resolution trajectory generation algorithm had addressed the applicable requirements to development of the resolution trajectories, gathered results from simulations, and raised conclusions from the performed exercises. The obtained results have shown the expected behavior from the complexity indicator, whereas the system complexity was reduced.
WP4 provided a basic concept of ontology, its types and key aspects to support the future improvement of an agent-based conflict resolution application in which airspace users interact to reach a consensus in resolution. To mitigate the interaction problems while enhancing the agents' negotiation capacity, the proposed ontology provided the baseline to specify the conditions under which the agents' requests can be satisfied during the negotiation. The ontology was an important objective to support the end-users to test their operational preferences.
WP5 was directed to the modules verification and the scenarios validation, tested in the Open Demonstrator. The modules verification considered an exploration of the methods and functionalities, with respect to the input and output data of each module. Within the module layer all modules were executed. In this way the system enabled an individual/separate verification. Regarding the scenarios validation, the validation plan was introduced by describing the process and envisaged parameters.
WP6 elaborated AGENT communication and dissemination policy and strategy, target groups, timeline, metrics, resources and roles, actions and channels for dissemination. The work also considered the exploitation activities, through definition of the exploitation plan, that gathered the project objectives, main outcomes, and what exploitation activities should be developed accordingly.
WP7 was active during the project lifetime. The objective was to ensure an efficient and active coordination of the project through administrative including support in the implementation of recommendations; executing, and post-processing of project meetings; preparation and submission of the management related parts of the reports; management of the financial reports; maintenance of the project intranet and data repository.
In overall, AGENT operational framework supported:
• An automated ATM environment with reduced human interventions, by shifting decision-making process to the airborne agents.
• Human interaction within highly automated system. Ecosystems strive to facilitation of the airborne interactions through development of novel operational procedures.
• State space search for the compromised solutions in the resolution process.
The AGENT impact has been kept viable in the following aspects:
- Demonstration and quantification of the potential for higher automation levels as well as provision of the benefits in safety, capacity and operational efficiency.
- Expression of an improved cost-effectiveness and contributions to decision support techniques for safety-critical applications.
An extent, to which the expected impact on results has been achieved, is considered as two-fold:
- Cooperation between the consortium members in a frame of developed AGENT tools certainly produced an impact on the overall project scope, especially if considering different orientations of industry and scientific communities.
- Collaborative networking between the AGENT consortium and the European ATM institutions (SESAR JU, EUROCONTROL).