Project Deliverables
Deliverable D2.1: State of the Art Review (SOAR)
Deliverable D3.1: Machine Learning Report
Deliverable D3.2: Machine Learning Demonstrator
Deliverable D4.1: E-UI Design Document
Deliverable D4.1: Video Demonstrator
Deliverable D4.2: E-UI Validation Report
Deliverable D5.1: Integration report
Deliverable D5.2: Simulation 1 report
Deliverable D6.1: Experimental Design report
Deliverable D6.2: Field Simulation report
Deliverable D7.2: 1st Workshop results
Deliverable D7.3: 2nd Workshop results
Deliverable D7.4: Final Project Results
Supplemental Materials to D6.2
SectorX, a Java-based, medium-fidelity ATC research simulator developed by TU Delft, was used in
MAHALO to collect data for the real-time human-in-the-loop simulations. The simulator is configurable to mimick any existing Plan View Display, allows a controller to interact with aircraft via either a clearance menu or touch input device. It can be customised to display various decision-support tools for conflict detection and resolution, ranging from classic tools such as VERA (used by MUAC) and the Separation Monitor, towards novel ecological interfaces developed by TU Delft (e.g., Solution Space Diagram and the time- and distance-based travel space). The simulator is also capable of using (and visualising) high-resolution wind data (GRIB files), modeling pilot delays and simulating basic conflict detection & resolution algorithms (at various levels of automation), but not all of these aspects were not used in MAHALO for the sake of experimental control.
For research purposes, SectorX allows the creation of custom scenarios (draw airspaces, add flights and routes, define conflicts, etc.) or import existing airspaces, simulate a set of scenarios according to a customisable playlist and replay recorded sessions. The recorded radar states and events (e.g., human and/or automation actions) are written in XML files that can easily be parsed with Python for further processing (e.g., to calculate statistics).
The animations below are snippets from the MAHALO trials.
Articles, Posters, Presentations
RPAS and AI final dissemination event: MAHALO presentation slides
SID2021: Human-interpretable input for Machine Learning
in Tactical Air Traffic Control
SID2020: MAHALO poster description (PDF)
DASC2020: Building Transparent and Personalized AI Support in Air Traffic Control (PDF)
Videos
SESAR AI White Paper video
Ecological User Interface demonstrator
Machine Learning demonstrator
10th SESAR Innovation Days, Virtual Conference, December 7 – 10, 2020
39th Digital Avionics Conference, Virtual Event, October 11-16, 2020
http://mahalo01.lr.tudelft.nl/wp-content/uploads/2020/10/MAHALO-DASC2020_Trim.mp4#t=1