A look at a high-tech evaluation project involving experienced pilots and those new to the cockpit.

eVTOL aircraft will soon transform the skies above, but who will fly these new and novel vehicles, what skills will they need, and how will they be trained to operate platforms that feature new levels of automation?

This is the question behind a current research project headed up by Aptima through a contract awarded by the U.S. Air Force. Aptima started in 1995 with the mission of optimizing and improving human performance in mission-critical, technology-intensive settings in defense, intel, aviation, law enforcement and healthcare.

The results will help the Air Education and Training Command’s Detachment 62 (Det 62) determine the pilot proficiencies and training needed for eVTOL operators.

“Det 62 supports the AFWERX Agility Prime program, and is charged with developing the curricula for eVTOL pilots and driving certification standards for an emerging market that is expected to transform civil air mobility and select military missions,” Aptima stated.

The findings will not only be used to help finalize eVTOL certification and pilot certification, but also perhaps future eVTOL designs in terms of automation levels for certain aspects of flight.

Dr. Samantha Emerson, a scientist in Aptima’s training, learning and readiness division and project manager for the contract, added that the results will also help determine the impact of automation on pilot performance.

Inside the project

A total of 83 participants — half experienced fixed-wing and helicopter pilots and half novices — were involved in the data-gathering phase, and data analysis is now underway. Aptima will be presenting some results at the I/ITSEC conference in Orlando, Florida, later this year, with a paper and presentation at one or more other conferences to follow.

Aptima gathered data from the learning and demonstration of piloting skills by the 83 participants using two simulators. Each simulator represented operating a different eVTOL model from two original equipment manufacturers (OEMs) — models that are currently well on their way to type certification.

The companies and models can’t be named, but both models involve advanced automation of flight — one more automated than the other.

These simulators include the two modes of flight that eVTOLs are designed to operate in — forward flight and hover — and the transition between them after take-off and before landing.

Data gathering

To assess pilot learning and performance, Aptima used two main technologies it developed for the Air Force Research Laboratory over the past 15 years.

One is the Performance Evaluation Training System (PETS), an algorithm-based measurement tool designed for fast jet training, which harvests data from the simulators to provide objective, system-based measures — airspeed, altitude, adjustments for pitch, roll, yaw, etc. The other is Aptima’s SPOTLITE, a handheld tool used by experts — in this case, experienced pilot-instructors — to provide observer-based measures of performance.

Together, these objective and subjective measures produce a good picture of pilot learning and performance.

Dr. Kent Halverson, Aptima principal scientist and the contract’s principal investigator, said having the data from the simulator being recorded and analyzed by the PETS algorithm meant that the instructors didn’t have to keep constant track of airspeed and altitude, but could focus on more nuanced aspects of performance captured by SPOTLITE. As they guided learners through the profile, instructors provided ratings for basic flight maneuvers, hover, take-off, en route navigation, and approach and landing.

For its part, he said “PETS gives an objective analysis of how close participants were to the ideal pathway they should be taking, ideal rate of climb during take-off, and so on. PETS can actually pull and analyze data from multiple simulators at once, and in this case, it was only pulling data from one simulator at a time.”

Experienced versus novice

Experienced pilots in this study obviously have extensive skill in flying fixed-wing airplanes and/or helicopters, and novices have none.

“So, we are looking at how long the two groups take to learn different operation maneuvers, keeping in mind that these simulations did not deal with extreme weather or emergencies,” Emerson said.

It remains to be seen, as data analysis proceeds, how much and in what ways these differences affect learning how to pilot an eVTOL with advanced operational automation. However, several things can be gleaned so far from preliminary analysis.

“Anecdotally, we are seeing that if you already have that hovering experience from being a helicopter pilot, you do learn faster,” Emerson said. “We are expecting that experienced pilots will do better. We know how many hours each experienced participant has on different types of aircraft, and we’re including that in our analysis.”

However, Emerson added that “we’re also seeing that those with no experience can learn to pilot these eVTOLs pretty quickly. Again, we aren’t dealing with weather or emergencies in this study, and ground training of eVTOL pilots will be needed, but these novices can do it, they can pilot. Even with the simulator that’s less automated, it only takes two hours to reach a relatively high level of proficiency for both pilots and those without any experience. It’s relatively easy to learn.”

Unlearning skills

At the same time, Emerson explained that experienced pilots of conventional aircraft may have more difficulty adjusting to some of the automation features of eVTOLs compared with novice pilots. These types of pilots may be somewhat hampered by the instinct to “overcontrol.”

Indeed, experienced pilots may have skills that may not be useful or even desired in platforms with more automation and augmentation. For them, eVTOL piloting may require an “unlearning” and retraining of behaviors to prevent interference or conflict with automated operations.

“Pilots will still be in control of the aircraft at all times and can make decisions — there are currently no automated processes, such as landing, where a pilot is completely excluded, but at the same time, you don’t want there to be conflict between a pilot with an automated process that’s been proven to be successful,” Emerson said. “We want to ensure that the human and machine are working as an integrated team.”

Experienced pilots are going to have to learn to trust the automation found in eVTOLs, she said, and that will require a mindset shift.

“We are giving these two OEMs feedback about this area, and we all must keep in mind that the automation software is still being refined,” Emerson said. “It may be that we need to train those who are used to having complete control over their plane or helicopter in a different manner to get them quickly proficient with eVTOL piloting compared with novices. We need to pinpoint how and when an experienced pilot may not trust the automation and ensure we have training that proactively anticipates that problem.”