SNJL Research Activities.png
Using AI for air traffic management

With air traffic set to double in the next two decades, future airports will need to improve their management of resources, better manage congestion, and use new technologies and techniques to do so.

To this end, Associate Professor Sameer Alam from the NTU School of Mechanical and Aerospace Engineering will lead a team to work at integrating AI into airport-airside management and operations.

One such project explores the use of AI and machine learning to develop a system that can detect and track a wide range of objects in the airport’s airspace in real-time, including drones or birds. The system will be able to alert air traffic control officers in good time to improve their situational awareness.

His team is also working on an AI-powered Data-Driven model that predicts runway occupancy time such that air traffic control officers can be alerted well in time upon detecting an anomaly – a possible case of aircraft veering off the runway, for instance.

These technologies will go towards building an AI-Human Hybrid Control Tower System that will see an AI assistant providing instantaneous, real-time advice to human air traffic controllers, by recommending operational and strategic steps that could be taken to optimise air traffic, reducing waiting times and improving overall airport efficiency.

Associate Professor Sameer, who is also the Programme Director of AI & Data Analytics at NTU’s Air Traffic Management Research Institute, said: “We are moving away from simulation and mathematical models to refine our deep learning and machine algorithms by leveraging computing capabilities available to us, and using data from our industry partners. Data is the fuel, and we have the car. From the data, we can learn behaviour and patterns, and from there, we can predict and improve the overall system behaviour.”

AI-powered unmanned underwater vehicles

The joint lab will also develop new capabilities in underwater robotics, a market that is expected to reach US$7.64 billion by 2025.

Unmanned underwater vehicles (UUV) have a wide range of applications, from performing preventative maintenance for ageing marine and offshore equipment, to monitoring the environment of hard-to-reach areas or hazardous locations, and even underwater sea exploration.

But these tasks require a UUV to have underwater spatial perception in low visibility environments. Other challenges to overcome include identifying an object’s exact position underwater, and underwater communications.

NTU Associate Professor Gerald Seet, who is leading the underwater robotics projects, said: “The applications for underwater robots are vast, with the most immediate and demanding needs in areas such as the marine and offshore sectors, the monitoring of sea life and underwater environmental changes, or even search and rescue missions. The solutions we are seeking to develop with this Saab-NTU collaboration could unlock the potential of the UUV market.”