The Dutch Air Force is reading the brain waves of pilots to make training more difficult

Fighter pilots in training have AI read their brainwaves as they fly in virtual reality to gauge how difficult they find tasks and increase the complexity if necessary. Experiments show that trainee pilots prefer this adaptive system to the rigid, pre-programmed alternative, but that it does not necessarily improve their skills.

Pilot training in simulators and virtual reality is cheaper and safer than actual flight, but these training scenarios need to be adjusted in real time to keep tasks in the sweet spot between comfort and overload.

Eva van Weelden at the Royal Netherlands Aerospace Center in Amsterdam and her colleagues used a brain-computer interface to read the brain waves of student pilots using electrodes attached to the scalp. The AI ​​model analyzed this data to determine how difficult it was for the pilots to find the task.

“We are constantly working to improve [pilot] training and what it looks like can be very different,” van Weelden says. “If you’re not in the industry, it sounds very sci-fi, I guess. But to me it’s really normal because I’m just seeing data.”

Fifteen Royal Netherlands Air Force pilots underwent training while the system switched between five different levels of difficulty – achieved by increasing or decreasing visibility within the simulation – depending on how much the AI ​​model determined they were looking for missions.

In later interviews, none of the pilots said they noticed the system changing the difficulty in real time, but 10 of the 15 pilots said they preferred the change tests to a pre-programmed exercise where the difficulty gradually increases in regular steps.

But crucially, none of the pilots showed any improvement in how well they performed the tasks in the adaptive simulation compared to the rigid simulation. In short, pilots liked the mind reading setup, but it didn’t make them better pilots.

The problem could be the unique nature of human brains, says van Weelden. The AI ​​model was trained on data from another group of novice pilots and then tested on 15 study participants. But it’s notoriously difficult to get AI models that analyze brain waves to work on an entire population. Six pilots in the test showed little change in difficulty level values, suggesting that the AI ​​system may not have correctly interpreted their brain data.

James Blundell at Cranfield University in the UK says similar technology is being studied for use in real-world aircraft to ensure pilots are in control. “They were looking at whether we could detect startle – like being in a mild panic – and then what the aircraft could do to calm you down and then reorient you,” says Blundell. “So you’re upside down, [and the aircraft might say] you really have to look at the attitudes, you have to look at the information that’s down here that’s going to bring you back to square one.’

These systems have shown promise in isolated scenarios, but it remains to be seen whether brain-reading technology can be used to improve safety on airplanes. “We have a long way to go. [in order to achieve that]” says Blundell.

topics: