Air Traffic Management Project

Pushing Boundaries for a Transformative Digital Air Control Tower

5 WPs of ATM

“We are moving away from simulation & mathematical models to deep learning & machine algorithms by leveraging on the computing capabilities available to us and data from industry partners. Data is the fuel, and we have the car. From the data, we can learn about behaviour and patterns. From there, we can predict and improve overall system behaviour."

Assoc Prof. Sameer Alam, Principal Investigator  and Co-Director, Saab-NTU Joint Lab

WP 1: Deep Learning Computer Vision-based Algorithms for Airport Environments

This work package is investigating deep learning, computer vision techniques to provide airport airside surveillance and intelligent predictive. The scope of this work includes:

*  small flying object detection;
*  runway & taxiway - aircraft airside monitoring; and
*  aircraft turnaround monitoring and prediction.

Aircraft turnaround prediction performed by deep learning computer vision framework and machine learning model.

WP 2: Data-Driven Metrics, Models & Predictive Capabilities based on advanced Machine Learning

The objective of this work package is to research and develop state-of-the-art data-driven machine learning models for the derivation of novel air traffic control performance metrics and procedure models, and their integration into advanced Artificial Intelligence optimisation algorithms.

actual flight simulator 400 x 300

Human in the loop validation using a flight simulator

WP 3: Innovative Interface Technologies and Mixed Reality Control for Air Traffic Controllers

This work package is investigating next-generation interface technologies, including Mixed Reality (MR) interfaces and Explainable AI (XAI), that will ensure that Air Traffic Controllers have continuous access to the most appropriate information, delivered in a way that can accelerate situation awareness without overwhelming digital data.

Air Traffic Controller engaging with mixed reality interface and 3D printed airport infrastructure model.

WP 4: Effective Human-ArtificiaI Intelligence Teaming in Air Traffic Control Environments

This work package examines the impact of introducing Artificial Intelligence & Machine Learning algorithms into Digital Air Traffic Control environments, and how human controllers and intelligent automation can most effectively collaborate.

Measurement of prefrontal cortex activity to determine operator workload when using automation assistance.

WP 5: Integrated Human-Artificial Intelligence Hybrid Digital Tower Control System

This work package shall integrate the research outputs of Work Packages 1 to 4, creating a working prototype Digital Remote Tower Control system that can be used to validate the complete integrated human-AI collaborative air traffic control system.

An integrated AI-driven digital air traffic control system will empower the air traffic controllers of the future.