Inside the Air Force's Plan to Revolutionize Pilot Training

U.S. Air Force 2nd Lt. Seth Murphy takes a 3D vision test prior to virtual reality flying training at Austin-Bergstrom International Airport in Austin, Texas, Feb. 5, 2019. (Sean M. Worrell/U.S. Air Force)
U.S. Air Force 2nd Lt. Seth Murphy takes a 3D vision test prior to virtual reality flying training at Austin-Bergstrom International Airport in Austin, Texas, Feb. 5, 2019. (Sean M. Worrell/U.S. Air Force)

Editor's Note: This story has been updated to correct inaccurate information provided to describing Air Force secretary Heather Wilson's visit to Austin.

When Air Force Secretary Heather Wilson visited AFWERX's Pilot Training Next program in Austin, Texas, last year, she watched as trainees took flight from the seats in front of her -- through the use of virtual reality.

Officials with Air Education and Training Command (AETC) are now gearing up to present Wilson's successor with a business case for more widespread use of the system, within the force.

The focus on simulation comes as Air Force leaders work to overhaul the pilot training curriculum, introducing one that augments time airborne in the cockpit with simulators and technology on the ground. The Air Force overall is readying itself for the possibility of complex conflict with a peer-level adversary equipped with long-range missiles and advanced combat aircraft. It's a future that may represent a strong contrast to recent decades, in which the Air Force has flown in largely uncontested airspace supporting ground troops.

Related content:

The service is attempting to boost its pilot ranks amid a longterm pilot shortage, even as its trainer fleet ages.

Air Force officials say they want to move away from the service's old-fashioned, "industrial" approach to training -- having pilots sit in classrooms for weeks then moving on to a trainer. This means using virtual reality earlier and more frequently in the training pipeline.

As the service prepares to bring its latest trainer, known as the T-X, into the fold, it is proposing a more "concentrated dose" of training to seamlessly transition from virtual reality to the trainer and, finally, to the Formal Training Unit, or FTU.

The system is well poised to reform in a few ways, said Gen. Mike Holmes, commander of Air Combat Command (ACC). Using the low-cost immersive environment of virtual reality, together with "competency-based learning," and moving skillset testing at the graduate level to an earlier place in the model, "would experience our pilots much faster," he said.

"Those are two things that are poised to make a revolutionary change in how well we train pilots and in how long it takes us to train pilots," Holmes said Tuesday in an interview with "I want to see how fast and well I can produce experienced pilots."

Pilots end up leaving the service if they feel dissatisfied and lack a sense of purpose, added Lt. Gen. Steven Kwast, AETC commander.

"You have to fly a lot to be good at what you do, and we don't have the money, and we don't have the weather, and we don't have the range space ... [because of] sequestration. And all these things that are politically driven oftentimes are frustrating the force," Kwast said in a separate interview.

Airline hiring efforts are the biggest factor that drives pilot retention and production problem in the services, officials have said.

Old learning mechanisms also bog down the system, often adding to pilots' frustration, Kwast said.

"We would [add] layers of things over time" through the course of a pilot's service, "basically assuming, 'You can't handle the truth!' or 'You're not smart enough to be able to learn this holistically, we have to give it to you piecemeal and then you'd put it together in your brain over time.' That's why it would take seven years to make a great mission commander pilot."

But now, he said "We're breaking that paradigm."

Trainer Fleet in Trouble?

The service still relies heavily on its trainer fleet for training, even though virtual reality is the new frontier, Holmes said.

"There's still no substitute for being in a real airplane," he said. "I think we'll always want a mix of learning our skills cheaply, but also build on decision-making in a real airplane."

The T-38 Talon has been the backbone of the Air Force's undergraduate pilot training, or UPT, program for decades.

But last year, the trainer fleet was plagued with a series of crashes, two of which were fatal.

Those selected to fly bombers and fighters typically receive their advanced pilot training in the T-38. The T-1A Jayhawk, meanwhile, is used in advanced training for students who are slated to go into cargo or tanker aircraft.

The T-6 Texan II, used for instrument familiarization and low-level and formation flying, also has had its share of problems. Last year, the Air Force ordered an operational pause for the T-6 fleet after pilots suffered a series of unexplained physiological episodes, or UPEs.

As a result, AETC on Feb. 1 ordered an indefinite operational pause for all T-6 aircraft at Columbus Air Force Base, Mississippi; Vance Air Force Base, Oklahoma; and Sheppard Air Force Base, Texas. That pause was lifted Feb. 28. A team of experts determined that the T-6's On-Board Oxygen Generating System (OBOGS) filter and drain valves failed at higher rates than expected.

The discovery led to repairs and increased inspections, but pilots continued to suffer from UPEs.

A T-6 trainer from Sheppard Air Force Base, Texas, crashed just last week.

The Air Force is preparing to receive new trainer jets to replace its current Northrop Grumman-made T-38s, some of which date to the mid-1960s. In September, the service awarded Boeing Co. a $9.2 billion contract to build its next aircraft for training pilots, known as the T-X program.

The first T-X aircraft and simulators are scheduled to arrive at Joint Base San Antonio-Randolph, Texas, in 2023. The service has committed to buying 351 T-X jets, 46 simulators and associated ground equipment. The payment structure, officials have said, also allows for an indefinite-delivery, indefinite-quantity option to give the Air Force the opportunity to purchase up to 475 aircraft and 120 simulators.

Delays to this program or other unforeseen challenges could have catastrophic consequences, said retired Gen. T. Michael "Buzz" Moseley, former Air Force chief of staff.

"My anxiety over this when I was the chief [is that] we are one sortie away from this older inventory having a problem," Moseley, an F-15 Eagle pilot, said in a recent interview with "Here we are in 2019, and we're going to fly these airplanes until 2024 before T-X starts coming in."

Kwast and Holmes agreed that the T-38 fleet will continue to undergo any upgrades necessary to keep them flying as long as it makes sense.

"You can make anything last longer; it just takes more money to sustain," Kwast said. "I guarantee that the T-1, the T-38 and the T-6 all can last as long as we need them to last, depending on the business case and the amount of money you want to spend. But will the T-38 or the T-1 become too expensive, and [therefore], we have to jump to a different technology? Then we would look at other options."

Boeing said it stands ready to produce the T-X.

"Our T-X program, including engineering, manufacturing and test, is located in long-established Boeing St. Louis facilities," wrote Rachelle Lockhart, spokeswoman for the company's T-X program, in an email. "In fact, we built and assembled our first two T-X aircraft in St. Louis prior to contract award to prove the maturity of our design, repeatability in manufacturing and performance. We're now on contract, executing on schedule as planned, as are our suppliers."

She added the trainer's production schedule could be advanced at the Air Force's request.

"The U.S. Air Force plan calls for a full production rate of 48 jets a year, and we will meet the customer need," Lockhart said. "Should the Air Force request a higher rate of production, we are well positioned to accommodate it."

No Cockpit? No Problem

Kwast stressed that the airframe is not necessarily the most important focus in flight training, which is why the Air Force is ready to rely more heavily on virtual simulation.

"What we're finding in the data here, which is giving us a sense of peace that this is not going to be as dramatic or traumatic, is that the human brain does not take a long time to learn the caveman skills," he said.

Like turning the keys in a car's ignition, parts of pilot training become second nature quickly. So leveraging simulations versus spending time in a real cockpit isn't concerning, Kwast said.

Artificial intelligence within the Pilot Training Next, or PTN, program, for example, grades pilots in real time to give them additional insight into improvements they can make.

Simulations offer quicker, easier access to training than real cockpits do, Kwast said.

"When we get the T-X, as good as it's going to be ... it will still suffer from the mechanical failures that happen with a machine; weather that's not going to change anytime soon unless global warming really gets aggressive; the range space, which is getting smaller and smaller ... and the fact that technology is making the range and speed of the fight so big, that the only range space that would be adequate would be from here to Australia," he said.

Instead, he added, "we can put them in that cognitive reality [sim] 100 times a day if we want. And they can get those combat hours. ... Virtual reality, A.I. and supercomputing are going to be a part of our world going forward even when we get the T-X."

Holmes, an F-15 Eagle and Strike Eagle pilot, said ACC is starting to look at employing more virtual reality and simulation training once pilots get to their FTU, which follows fundamental and undergraduate training. It's also where pilots are assigned to an airframe.

"And we'll see what we can do there to train them better and faster in that environment," he said.

It's an added layer -- "an improvement of the process that we've been doing," Holmes said. "History would tell us we should expect to make some changes in the way we train pilots as we acquire a new airplane because we always have."

Reality: A Few Growing Pains

While Kwast and Holmes make the case for expanded virtual training, the PTN system thus far is built to increase efficiency, not numbers.

Thirteen students graduated from the first PTN class last August after six months learning to fly in virtual-reality simulators. Another class is currently in PTN, with AETC anticipating a third this fall.

The program, however, is still considered experimental.

Meanwhile, the Air Force ended fiscal 2018 with a total force pilot shortage of 1,937.

"How do we get at this deficit of pilot numbers?" Moseley asked. While the service is losing more of its career seasoned pilots, new pilots are just as vital, he said.

"In the history of the Air Force, you never, ever want to graduate less than 1,100 or 1,200 pilots a year. Those are magic numbers," said Moseley, who said that's the minimum needed to sustain squadrons.

In fiscal 2017, the service trained 1,160 new pilots. AETC graduated 1,201 student pilots from its undergraduate pilot training program in fiscal 2018, according to command spokesman Daniel Hawkins.

While still hitting "magic number targets," the service wants to push the system further.

Last year, officials announced plans to ramp up pilot training to produce 1,500 pilots a year by fiscal 2022. That number includes active-duty Air Force, Air Force Reserves, Air National Guard and international students, AETC has said.

Experts say the 1,500 goal may be too ambitious.

"We have a drastically different environment now where the body pool doesn't exit," said Gene Colabatistto, group president for defense and security at CAE, a company that provides worldwide training and integration for the civil and military aviation. Colabatistto provided remarks alongside Holmes and other top leaders about pilot attrition at the annual Air Force Association's Air, Space and Cyber conference last fall.

"We're not going to fix this if we buy better airplanes, or have better simulators or train better instructors," he said.

Colabatistto argued the many issues plaguing the service have nothing to do with the type of planes or technology used.

"It's a system issue," he said. "You have to attract people, and they want to become pilots in general. Otherwise ... you'll never succeed.

"People are opting not to become fighter pilots, which I [find] extraordinary," he said, adding that the film "Top Gun" was a great recruitment tool once, but that particular magic may not happen again.

Colabatistto said when speaking with senior leaders over the last few years, he's deduced that the younger generation can't commit to giving 10 years, which is typical for the pipeline.

"You have to stand back and look at it, because you're not going to fix it by treating one part of the problem," he said, referencing not only recruiting, but keeping pilots in service longer.

"Everything is connected. I don't doubt that having a new aircraft that's very capable [T-X] will be a very important building block, but if they can't recruit anybody, what difference will it make? If it was that easy, we would have done that five years ago," he said.

Kwast said if the 1,500 goal isn't feasible, the Air Force won't force it.

"If we can't absorb those pilots and they can't make it through [their formal training unit], or if that gateway is bound up, you actually make the problem worse," he said, adding that the system needs to allow for flexibility.

Modifying the System, Not Outsourcing Solutions

Kwast said the training pipeline "allowed for mediocrity" in the past with a one-size-fits-all approach for students.

"But we now have the data and the technologies to create learning environments to design excellence into it. I'm trying to design a system that is resilient to the wave of money and politics and still be able to produce warfighters in all 127 skills that put airpower together," he said.

The fact is, Holmes said, just as aircraft come and go from the service, pilot training never stays the same, either.

"We have to make sure we don't get in the way of our future," he said.

Making rash decisions to stopgap a problem could end up costing more in the long run, he said.

That includes outsourcing solutions.

Some companies that contract pilot training out to other countries, and even other U.S. military branches, say they could be interested in working with the Air Force on its undergraduate pilot program.

Draken International, Top Aces Inc., Tactical Air Support, and Textron-Airborne Tactical Advantage Company are all companies focused on providing adversary air, known as red air training. They currently await the Air Force's decision on a new indefinite delivery, indefinite quantity (IDIQ) contract for adversary air, which the service is expected to award soon.

But some of them may be at least eyeing parts of UPT, whether that might be providing instructors, simulations or aircraft.

Mick Guthals, a retired Air Force colonel and senior manager for business development for Tactical Air Support, could see his company taking on the instruction and aircraft angles.

TacAir has more than two dozen reconfigured F-5 Tigershark light fighters, originally made by Northrop Grumman. While it couldn't subsidize an entire fleet, it could at least help in "providing some of that space," Guthals said.

"We would absolutely welcome it," added Sean Gustafson, vice president for business development at Draken, of getting a piece of UPT. "Draken definitely has the capability to support such a project as a contract undergraduate program."

Draken currently has A-4 Skyhawks, L-159 "Honey Badgers," Dassault Mirage F1s and Atlas Cheetah fighters in its inventory for aggressor training, Gustafson said.

The Air Force says it has no interest -- for now -- in outsourcing UPT.

"What new aircraft do [those companies] have?" Holmes questioned. "They have A-4s, L-159 [ALCA jets] but not in great numbers. Some [of those companies] have talked about perhaps acquiring some advanced trainers of their own, but they haven't done that yet.

"I'm not sure bringing them in to train people on A-4s ... or old Mirages would really bring any improvement" to the Air Force, he said. "Would it deliver any faster than the T-X and, if it would deliver faster, would it really make a difference?"

Moseley says "it's not a total bad idea but ... it's a much better idea to focus on the [options] of pilot training and the ability to replace the T-38 quicker" with a holistic approach.

"You can go the contract route, but it will cost you," Kwast added. "And you sometimes find that the dollars are not aligned with your values," he said.

Of course, some contractors already play a limited role military pilot training.

For example, CAE offers Air Force C-12 aircraft training at its Dothan Training Center in Alabama.

And Doss Aviation, a subsidy of L3 Technologies, provides the service with Initial Flight Training -- a type of flight screening -- at its facilities in Pueblo, Colorado, on Diamond Air DA-20 aircraft, according to Lenny Genna, president of L3's Link, Training and Simulation department.

Doss is even looking to expand into the virtual simulator world with devices that can be interlinked with other simulators or training equipment -- ones the Air Force could use, Genna said in an email.

Whatever the system or provided service, "when you talk to people who want to provide a contract solution, they really need a guaranteed contract for multiple years to really [profit] from the investment they would make," Holmes said.

Air Force leaders are more often inclined to keep as much in the family as possible, he said.

"You have to decide on the mix between the skill training, the culturization, and socialization you get by putting your student pilots through with Air Force aviators," Holmes said.

Guthals added, "I think an Air Force needs to show they can do their own training in order to be an Air Force."

The New Frontier

Whether T-X comes along faster or not, Moseley said there has to be a multi-pronged solution to training pilots more efficiently while raising retention.

Change doesn't happen overnight, Kwast said, but added he is hopeful for the future.

"Even though I can't predict the future, it will trend in that direction," Kwast said. "These young kids, they find it so sexy and so exciting to be really good at something they love, that the retention where they want to stay in is going to go through the roof."

Kwast said the PTN program -- and testing young students who are thinking of joining the Air Force through virtual reality or gaming -- offer clear indications early on who will make an exceptional pilot.

"We can start [by] recruiting excellence," he said.

New tools, Holmes added, only make potential pilots better.

"If there's been any revolutionary change, it's probably been accomplished by Steve Kwast in his UPT-Next," Holmes said.

Kwast says he has faith in the new recruits coming in. He zeroed in on some of the students he's come across in the latest pilot cadre.

"Normally, it takes quite a few rides before students can go out and do loops, Immelmann [turns] and 'Lazy 8' [maneuvers] ... to [test] spatial perception, three-dimensional capabilities. Normally, it takes many rides for a pilot to get in the aircraft and go out," he said. "But in the cognitive environment of the A.I., it is so rich and so good and the A.I. coach is so good at making sure [the pilot] isn't learning negative things, only positive, that in [his/her] first ride, the pilot is going out and ... is executing all these procedures and the [back-seat] instructor doesn't say a thing.

"We have never seen that before in the history of pilot training, and we see it with these kids because of this technique," Kwast said.

"This is so much bigger and deeper than teaching people to fly," he said. "We are confident we will find other ways of making sure the human brain is lethal and ready to fight."

-- Oriana Pawlyk can be reached at Follow her on Twitter at @Oriana0214.

Story Continues