Modern ATM via Human/Automation Learning Optimisation

Livrables

D3.1 will document the development and tuning of the component SL and RL models, their integration into a hybrid ML system, the synthetic traffic generator capability, HITL runs, and analysis of both ML system performance and participant subjective data.

Analysis & report, Sim 1

A full report on the setup and results of Sim 1

Concept report

The D2.2 report will present the MAHALO ATM concept of operations, including anticipated human roles and information requirements, and candidate test protocols.

Final Project Results Report

In addition to documenting project activities developments results and conclusions this final report will also provide a framework of specific guidelines for designing future automation and a set of recommendations for followup dissemination and exploitation especially with respect to relevant other domains

E-UI validation report

This report will document the low fidelity validation of the E-UI.

Machine Learning demonstrator

A related but standalone deliverable D32 which presents the software realization of the hybrid ML modelSince this will be intended for demonstration purposes only it is anticipated that this will be a playback only noninteractive presentation

Integrated State-of-the-Art Review

This deliverable will present an integrated report on the output of the human performance- and ML reviews, highlighting the latest theoretical and empirical work into each. T1.1 and T1.2 effort will also feed directly into the production of D2.2

1st Workshop report

Reports will be issued within one month of completion of each of the two workshops Contents would cover at a minimum agenda participants links to any public access documents or materials discussed Minutes of the workshop and a summary of action items

2nd Workshop report

Reports will be issued within one month of completion of each of the two workshops Contents would cover at a minimum agenda participants links to any public access documents or materials discussed Minutes of the workshop and a summary of action items

Field simulation report

A full report on the setup and results of Sims 2A and 2B

Experimental design document

Documenting the experimental design variables test procedures methods materials and analysis routines for the closed loop simulation 2A Notice that amendments might have to be made to simulation 2B design based on the outcome of simulation 2A

E-UI design document and demonstrator

This report will document design choices, implementation details, and also include a non-interactive video playback of the E-UI, for demonstration purposes.

Analysis & report, integration

Documenting the integration activities of WP5

Project website

A project website will be established within one month of the kickoff meeting, and will contain at a minimum links to news items of relevance, project results, project calendar, related project links, and demonstration media (images and or video links).

Publications

Human-interpretable input for Machine Learning in Tactical Air Traffic Control

Auteurs: Tiago Miguel Monteiro Nunes, Clark Borst, Erik-Jan van Kampen, Brian Hillburn, Carl Westin
Publié dans: 2021
Éditeur: 11th SESAR Innovation Days Conference
DOI: 10.5281/zenodo.7304699

Building Transparent and Personalized AI Support in Air Traffic Control

Auteurs: Calr Westin, Brian Hilburn, Clark Borst, Erik-Jan van Kampen, Magnus Bang
Publié dans: 2020
Éditeur: 39th Digital Avionics Conference

Machine Learning for Air Traffic Management: guidelines for future AIsystems

Auteurs: Cocchioni Matteo, Bonelli Stefano, Westin Carl, Borst Clark, Bang Magnus, Hilburn Brian
Publié dans: 2022
Éditeur: EASN Conference, Barcelona, 2022

Balancing transparency and conformance of an AI conflict detection and resolution support system

Auteurs: Carl Westin, Tiago Monteiro Nunes, Clark Borst, Stefano Bonelli
Publié dans: 2020
Éditeur: 10th SESAR Innovation Days

Personalized and transparent AI support for ATC conflict detection and resolution: an empirical study

Auteurs: Westin Carl, Borst Clark, van Kampen Erik-Jan, Nunes Tiago, Boonsong Supathida, Hilburn Brian, Cocchioni Matteo, Bonelli Stefano
Publié dans: 2022
Éditeur: SESAR Innovation Days, Budapest, 2022

Transparent and Personalised AI Support in Air Traffic Control: First Empirical Results

Auteurs: Bonelli Stefano, Borst Clark, Brambati Francois, Cocchioni Matteo, Hilburn Brian, Monteiro-Nunes Tiago
Publié dans: 2022
Éditeur: ICCAS Conference 2022
DOI: 10.5281/zenodo.7304747

AI in ATM: transparency, explainability, conformance, situation awareness and trust

Auteurs: Representatives of the following SESAR 3 JU Projects: AISA ARTIMATION MAHALO SAFEOPS TAPAS
Publié dans: 2022
Éditeur: Self Published

Livrables

Publications

Partager cette page

Télécharger