Discover how the ASTAIR project is using AI and machine-human collaboration to automate some routine ground movements, giving human operators more time to focus on complex, safety-critical manoeuvres.

Flight trajectory optimisation is the key to helping aircraft reduce fuel consumption and lower carbon emissions. While the bulk of the reductions happen during flight, the process begins here on the ground.

“A prerequisite to in-flight optimisation is on-time departures, which start with efficient ground operations,” says Mathieu Cousy, a research engineer at the Ecole Nationale de l'Aviation Civile (ENAC).

Although these operations take place on the ground, many of them are orchestrated from the air traffic control (ATC) tower. Take for example taxiing, the process of moving an aircraft on the ground.

Compared to flying, taxiing may seem like a straightforward affair, but it is full of risks, including collisions and runway incursions. That’s why pilots depend on ATC to help them safely navigate their aircraft into position for take-off or back to the gate.

But with airports becoming increasingly busy, getting aircraft to and from the runway on time is becoming both more challenging – and risky. Even though ATC, pilots and airport managers have a range of high-tech decision support tools at their service, these tools weren’t designed to handle such advanced ground operations as engine off taxiing techniques, where an aircraft taxis with at least one engine shut off.

While engine-off taxiing can significantly reduce emissions and fuel use, it often requires the support of autonomous solutions, such as a semi-robotic towing vehicle. “Modern ground operations require modern solutions, solutions that can autonomously manage routine movements and ensure that the human operator is able to focus their time on more complex, safety-critical manoeuvres,” explains Cousy.

That solution could be artificial intelligence (AI) and its ability to interact with human air traffic controllers.

Giving AI authority

AI has the potential to initiate a number of standard ground tasks, such as giving clearances to vehicles on airport aprons and taxiways. “One can envision a future where routine tasks are driven by AI with human supervision, while more traditional human-machine collaboration is used to manage unexpected events or specific requirements,” adds Cousy.

Helping to make such a future feasible is the SESAR-supported ASTAIR project.

“By designing a seamless partnership between humans and AI for managing and performing engine-off and conventional taxiing operations, we hope to compute the conflict-free trajectories that will allow these tasks to be automated,” says Cousy, who serves as the project coordinator.

The project builds on the previous work done by the SESAR-funded AEON project. That project developed a concept of operations where tug fleet managers and ground controllers work as a team, relying on AI-based decision support tools to allocate autonomous resources and optimise vehicle surface movements.

ASTAIR aims to take this concept to the next level. “AEON successfully added AI capabilities to standard ground procedures, but at the end of the day, the human operator is the decision-maker,” explains Cousy. “ASTAIR, on the other hand, proposes to give AI complete authority for making ground clearances to air traffic control.”

Sharing responsibilities

According to Cousy, this paradigm shift is now possible thanks to advancements being made in autonomous devices, including the launch of ‘TaxiBots’ and automated ‘follow me’ cars. “ASTAIR aims to combine these advanced taxiing solutions with automation technology,” he says.

To do so, the project is designing algorithms that will be able to collaborate with ground operators in their daily tasks. These algorithms will be able to supervise all ground surface movements and ground operations, including the allocation of towing vehicles, thus optimising ground traffic performance across the airport.

The project is also designing the human-machine interfaces that will provide the human operators with good situational awareness – ensuring efficient collaboration between their work and that of the AI algorithm.

“Unifying ground operation and control positions, operators and algorithms will share responsibilities to perform tasks as efficiently as possible at different levels of automation,” adds Cousy.

In practice, this means that AI will handle normal (i.e., ‘conflict-free’) ground operations. However, whenever something unexpected happens (i.e., a ‘conflict’), control and supervision will go to the human operator, who may be supported by AI. For example, if an outbound aircraft fails to advise ATC about a departure delay, human operators will be able to focus on resolving this unexpected event knowing that the AI is safely managing all normal traffic.

Or take the issue when an arriving aircraft is unable to park as it’s gate is occupied. Typically, this would have a cascading effect across all ground operations, and possibly even incoming and outgoing flights, as ATC would have to either delay other operations while resolving this issue or delay the departure of the parked aircraft in order to manage other taxiing aircraft.

“With the ASTAIR solution, ATC can focus their complete attention on efficiently resolving this potential bottleneck without creating additional bottlenecks while doing so,” notes Cousy.

Automating sustainability

By combining taxiing solutions with automation technology, ASTAIR has the potential to increase airport ground traffic capacity and predictability, which in turn will enable aircraft to better meet their on-time departure objectives.

The automation that ASTAIR enables can also make aviation more sustainable. For instance, by calculating smooth and conflict free taxiing, AI can help aircraft avoid slow-downs and start/stops – both of which consume significant amounts of fuel. Furthermore, because automated ground operations help get aircraft in the sky on time, ATC will be better positioned to utilise the most fuel-efficient flight trajectories.

“Human-AI collaborative solutions like those being developed by the ASTAIR project offer a clear path forward to more efficient ground operations, ensuring that an aircraft gets wheels up on time and is able to fully leverage optimised flight trajectories,” concludes Cousy.