The voice behind me was a familiar one: “You pilots love your toys, don’t you?” I turned around in the coffee shop to see my neighbor’s smiling face staring at the flight-simulation app I was playing on my iPhone. I smiled and nodded as the wave of his hand said, “Go back to your game.” Last time I checked, the App Store offered 20 different flight-simulator games, some free, the rest for less than 10 bucks apiece. Of course, those simulators are toys, distractions when someone’s waiting for a train, a plane or an Uber. Once you move past games, higher-fidelity machines used for flight training and currency share a common purpose: to effectively, efficiently and safely train pilots for the real world before they take command of airplanes.
Professional-grade simulators are created by a handful of companies such as FlightSafety International, CAE, TRU Simulation + Training, Simcom, FlyThisSim, Redbird Flight Simulation and Frasca International that have married tiny yet extremely powerful computer chips with spectacular video and audio and flying characteristics that almost exactly match those of an actual aircraft.
They’ve created product lines from desktop machines to full-motion simulators the size of a house to recreate a cockpit and teach an ILS approach or a holding pattern, or allow pilots to experience a V1-engine failure or give a realistic look at an airplane’s behavior at high altitude during an upset. Because simulators are computers, they, unlike actual airplanes, also allow for direct measurements of a pilot’s training progress. Despite a hefty price tag for a huge full-motion simulator, these machines can be operated for a fraction of the cost of the aircraft they’re replicating.
More importantly, simulators today allow instructors to conjure inflight scenarios that are incredibly real, but are simply deemed too dangerous to attempt in a real airplane because the cost — any cost — of losing an airplane to a training accident has become too much. Of course, even the large full-motion simulators can’t recreate that seat-of-the-pants feeling of pulling 2 Gs trying to maintain altitude in a steep turn. That’s one reason they’re still known as synthetic trainers; the world they create is simply not real. But we’re always coming closer.
David Smith, TRU Simulation + Training’s vice president of Training Centers says, “TRU has developed some interesting techniques to add the startle factor reaction into emergency training for additional realism.”
Simulators also have increased fidelity through add-on products such as PilotEdge, which brings real ATC communications and interactions to the training mix. There’s also the proliferation of add-ons for desktop machines such as control wheels, joysticks, throttles and radio stacks similar to what used to be found only in the most expensive simulators.
The FAA spurred the growth in simulators at flight schools when it updated advisory circular AC 61-136 allowing certified pilots a relatively inexpensive method of maintaining flight currency without the need for CFIs to monitor their progress. Josh Harnagel, Redbird’s VP of Marketing says, “We probably wouldn’t be in business without the FAA’s foresight on 61.136.”
A solid understanding of simulators today demands deciphering the AC terminology for Part 61 and 141 flight schools, such as the basic aviation training device, a system that “provides an adequate training platform and design for both procedural and operational performance tasks specific to the ground and flight training requirements for private pilot certificate and instrument rating.”
An advanced aviation training device operates much the same as the BATD, but is considered realistic enough for more demanding tasks found in “flight and ground training related to commercial pilot certificate, airline transport pilot and flight instructor certificates.”
An AATD requires a self-centering yoke and rudder pedals, accurate power controls including mixture and propellers, primary flight displays for a glass airplane, flaps, trim, communication and navigation radios, transponder, landing-gear handle and a few other items. And, of course, the simulator must fly like the airplane, or at least very similarly, depending upon how it’s used. Then there are flight-training devices and the full-flight simulators. An FTD recreates a specific aircraft type, but might not include full motion. Full-flight simulators are categorized as level C or D.
A level-D machine, such a Falcon 7X or Boeing 737-800 simulator, must recreate 6 degrees of motion, as well as realistic sound and visuals. That same machine can also produce a type-rated pilot, though the applicant might have never set foot inside the real airplane.
In the Beginning
Since the dawn of aviation, experts have realized some aspects of flying did not readily lend themselves to learning in the actual airplane. One of the earliest examples of flight simulation is detailed in a January 1919 Popular Science story, “Dry Shooting for Airplane Gunners.” Airmen preparing to fight in the Great War were trained in the basics of aerial gunnery because teaching young men how to pick off a WWI Albatros or a Fokker for real was impossibly dangerous. The simulator was nothing more than a fuselage mock-up with a gunsight and swiveling pilot seat inside. The targets were tiny pictures of enemy aircraft presented to the gunner.
The most famous early simulator was the Link Blue Box, created in the late 1920s by Edwin Albert Link as a safe method of teaching pilots the budding world of instrument flying pioneered by Jimmy Doolittle in September of that year. The Link Trainer resembled an airplane with a fully instrumented cockpit and a cockpit cover to prevent looking outside. Mounted on a motorized pedestal, it could turn, pitch and roll like a real airplane as the pilot moved the control stick and rudder pedals in the cockpit.
But with no way to see outside, the Blue Box demanded a student’s full attention, especially during the rough air the machine could also create. So sophisticated for the time was the Link Trainer that the American Society of Mechanical Engineers designated the machine as an historic mechanical-engineering landmark.
By the late 1950s engineers had created trainers with rudimentary visual displays — usually night scenes — needed to convince pilots they were actually flying. With more advanced computer-graphics imaging in the 1970s, one of the first full-motion visual simulators was created by McDonnell Douglas. Seeing the opportunity in the early days of business aviation, Al Ueltschi created FlightSafety International and the first full-motion simulator of a Gulfstream G-159 turboprop. Other manufacturers such as CAE in Canada and much later TRU Simulation + Training entered the market. Today, there’s a full-motion flight simulator for most every turbine aircraft in production. Despite evolving technologies, the goal hasn’t really changed from what Rolfe and Staples detailed in their 1986 book Flight Simulation: “The object of flight simulation is to reproduce on the ground the behavior of an aircraft in flight.”
Learning to Fly
Simulation plays a significant role in the booming flight-training marketplace because, as FlightSafety International’s general manager of Visual Systems Ed Koharik notes, “flight training is all about how you accelerate the training and learning process, increase capacity of the system or speed up the training process.” Speeding up training is where simulators shine. One success story has been at the U.S. Air Force’s Pilot Training Next program where qualified pilot candidates have been turned out in less than half the original 220 hours it used to take.
The team at Redbird was so certain a decade ago that using flight simulators in conjunction with airplanes was a winning combination, it created a Part 141 flight school that doubled as a research lab in San Marcos, Texas.
Redbird’s Josh Harnagel says, “We believed you could use a simulator to improve the skill set of a VFR pilot just learning to fly. Current schools graduate a private pilot with around 75 hours.” The company believed a more realistic number of logged flight hours, including some 20 in a simulator, would be closer to 40 total. “Because we were a proficiency-based Part 141 school, we graduated a private pilot with 32 total hours although our average was about 37 hours. Part 61-trained pilots averaged about 45 in the airplane and 25-30 hours in an AATD.”
Many flight schools liked those numbers because students experienced more-realistic training scenarios than they could in airplanes. This higher-order realistic pilot decision making is the bread and butter to what you can do with a simulator, Harnagel added. Redbird’s flight school went on to graduate 160 private pilots and 60 instrument-rated pilots.
Truly effective training demands as much realism as possible, some that can’t be attained by better hardware fidelity alone. Keith Smith, CEO at PilotEdge, says “If training is to really connect with pilots, we need to make them believe they’re flying in the real world.” What could be better than the distraction of ATC chatter? In most simulations, the air-traffic-controller role is non-realistically played by the sim instructor.
PilotEdge’s networked ATC software, an add-on to most simulators, which uses live air-traffic controllers giving real ATC instructions to a variety of airplanes on the frequency. Smith says “Without ATC there’s a massive psychological difference between flying a simulator with your instructor pretending to be ATC. It’s not the same as flying an airplane in the real world. There’s no pressure on the pilot to perform and there’s no conflicting traffic. Pilots don’t need to listen for their call sign.” Any wonder, he says, that without practice during training, so many instrument students are often terrified of the radio? PilotEdge tries to make sim flying feel less different than the real world.
At FlightSafety International, general manager of Visual Systems Ed Koharik says, “The industry is pushing to develop a simulated ATC environment for simulators, but it requires a lot of algorithms, voice recognition and simulated voices. That will be one of the next big things we’re looking at and it’s not too far away … maybe in the next five years.”
A Simulation Philosophy Lesson
Randy Brooks, VP of Business Development at Aviation Performance Solutions in Dallas, explained one of the serious drawbacks to flight-simulator training. For years, he says, “people believed that if the simulator can do it, that must be real life.”
That all changed when upset prevention and recovery training was mandated by Public Law 111-216. The industry realized it was incorrectly training pilots about how swept-wing aircraft react when they exceed the critical angle of attack, as did the crew of Air France 447 in 2009.
That crew, and thousands of others, had, of course, previously completed approved simulator training to earn their aircraft type ratings, but the investigation discovered full-flight simulators were not accurate at high angles of attack at high altitude because the simulators simply don’t have accurate data for that portion of the flight envelope.
The problem simulator manufacturers and Part 121 carriers faced was twofold. First, they needed to find the data necessary to create realistic scenarios, and they needed to integrate it into airline training rather quickly.
Brooks says, “By March 2020, all the airlines must have trained their pilots to the new Part 121.423” that requires UPRT including bounced landing and gusty crosswind work. UPRT isn’t currently required for Part 91 and Part 135 pilots. Upgrading level-D simulators required these machines to “demonstrate flight 10 degrees beyond critical angle of attack. Behavior at the actual stall is different than at the first indication,” he adds.
In the past, pilots were always expected to recover the airplane at the first indication of a stall. But before pilots could be trained for flight outside the normal envelope, he says, “instructors needed to be trained” —an effort that took much longer than many people planned. Brooks says we shouldn’t have been that surprised because, “all of our normal flight training is focused on operations inside the normal flight envelope.” He confirmed concerns about simulating the real world, including the startle factor, during potential-upset events at the edges of the flight envelope. “If pilots freeze up, they suffer cognitive impairment and won’t display the proficiency they need.”
Brooks believes APS has learned how to combine training in simulators and airplanes for the best effect. ICAO and IATA also believe combining simulator and flight training during UPRT is important.
With loss of control still a very real problem in general aviation, Brooks shared some startling data that reinforces why airline UPRT is required.
“Just a few years ago, Boeing data showed 43 percent of all fatalities were due to loss of control in flight,” in transport airplanes, he says. “While the overall number of deaths has declined, the contribution to fatalities from LOCI around the world has climbed to 49 percent.”
Barriers to Improved Realism?
FlightSafety’s product director for Mission Fit, Steve Smith, speaks to a future that doesn’t focus much on making simulators more realistic, but rather on knocking down some of the persistent pilot-training issues and problems. The company’s new Mission Fit creates interlocking cockpit modules in a local non-motion device that allow pilots to learn systems integration, cockpit familiarization and run a variety of checklists in a realistic environment, including content downloaded to individual flight departments via the internet. Customers can start their training at their own facilities and finish on level-D simulators, he says.
FlightSafety also is investigating mixed-reality systems that use actual cockpits and those available with virtual-reality goggles.
TRU has come to realize some technology might soon produce diminishing returns. They see the key as the right technology needed to better measure student performance to provide their instructors better insights into the mind and body of the student to change the way they teach in the simulator. “We’ve learned from technology outside aviation that we are making work in aviation,” TRU’s David Smith says.
Redbird’s Harnagel says his company wants to improve software and content more than the hardware: “Better systems don’t necessarily add much additional value.” Artificial intelligence is becoming important, as are ways to measure performance. “With Guided Independent Flight Training,” he says, “when students fly a simulator mission like steep turns, the computer will offer audio guidance like ‘pitch up or down or adjust your throttle.’” This will allow schools to better manage their outcomes across the entire student base.
Fly In The Cloud
An entirely new concept of home-based flight simulation was recently introduced by Palo Alto, California-based CloudFlyt.
Flying in the cloud has a double meaning here as the system is based around scenarios that literally live in the cloud.
This Internet-based configuration makes flight scenarios easy to access, and CloudFlyt can add new missions with immediate availability for the subscribers.
Users access the program through an app on an Nvidia Shield TV streaming device that, like Apple TV, acts as a portal between various apps and any TV with an HDMI port. A regular TV at home provides a perfect platform on which to display CloudFlyt’s flight scenarios.
A yoke and rudder pedals are connected to the streaming device, linking them to the airplane’s controls on the screen. The setup is intuitive, quick and easy, and there is no need to use a keyboard or mouse to access the program.
CloudFlyt has missions that help with learning airport markings, takeoffs and landings, crosswind landings, instrument approaches and more.
The system’s developer, Jim McDonald, hopes to eventually get FAA approval for CloudFlyt as an advanced aviation training device to allow pilots to fly and log the flights for IFR currency.
CloudFlyt offers a complete package for $400 and the subscription to the scenarios starts at $19.95 per month. -Pia Bergqvist