Simulating loss of control: research broadens
Loss of control in flight is now the biggest cause of commercial aviation fatalities, so what can be done to teach pilots how not to lose control? Two 2009 accidents involved stalls–Colgan Air 3407 and Air France 447–yet stalls are an elementary maneuver taught early in pilot training. If stalls are such a big problem, could training later in a pilot’s career using simulators better prepare pilots to get out of a stall or impending stall?
Taking this question one step further, could simulators be used to replicate actual accidents and thus help pilots experience the broken chain links that led to the accident in a way that is more effective than just training to avoid a specific wrong maneuver? After all, the accident data is usually available and could be downloaded easily into a simulator’s computers, including all the flight parameters, the ATC calls and even the cockpit distractions.
These questions aren’t new, but there are constraints and challenges. For example, how would a training company introduce the startle factor that is part of every accident while also replicating a specific accident? How much would it cost to add this capability to a simulator? And would this training truly be beneficial?
According to Doug Carr, NBAA vice president of safety, security and regulation, “Today’s simulators can do a pretty good job of replicating abnormal and emergency situations. It’s just a matter of programming. I think programming a simulator to replicate these conditions wouldn’t be that difficult.”
But programming the simulator isn’t the only issue, Carr explained. “The challenge is to design a training program that can effectively introduce the pilot to the condition and then [teach] how to correctly identify the situation and recover from it. Every pilot has no doubt been asked at the end of a sim session, if there’s time left, if there’s anything else he’d like to do. Exposing pilots to scenarios where there has been no coordinated training effort could lead to negative training and improper or dangerous techniques.”
Simulator manufacturer and training company CAE has launched a program by which flight data from aircraft flight operational quality assurance (FOQA) systems is downloaded into simulators so pilots can re-fly particular and realistic anomalous flight scenarios. CAE’s system is called simulator operational quality assurance (SOQA). CAE Flightscape ran a SOQA program in 2009 using a Boeing 737 simulator at CAE SimuFlite in Dallas for the U.S. Air Force, which flies the C-40 variant of the 737. One of the key findings of the CAE SOQA program is that it enables simulator instructors to measure what students are doing in the simulator, rather than simply making subjective observations.
FlightSafety International, which pioneered the use of sophisticated simulators for airlines and business aviation and has a simulator manufacturing division, “does not replicate specific accidents and incidents in simulators for training purposes,” according to a company spokesman. “The simulators FlightSafety designs and manufactures replicate the performance and operating characteristics of the aircraft they represent and are qualified to the highest standards established by aviation regulatory agencies around the world. The flying characteristics of FAA Level D-qualified simulators are based on actual aircraft performance data. They are designed and qualified to enable the pilot to fly the simulator within the normal flight envelope and in accordance with the government qualification. They are therefore able to generally replicate situations that are within these parameters.
“Our instructors are able to introduce a wide variety of faults, failures and weather conditions during the simulator sessions. They can be set up to specific weather, autopilot settings and other conditions. They are able to present various instrument indications, fly-by-wire control faults and system failures with the associated warning horns and cockpit warning indications. The simulators also have the capability to ‘play’ simulated or recorded tower conversations. Introducing faults, failures and weather with specific settings or parameters during training helps pilots to correctly evaluate the situation and to be fully prepared to take the appropriate action or actions.”
Wally David, president and CEO of SimCom Training Centers, has put considerable thought into the question of whether accident replication could add to training quality and the regulatory issues that this might raise.
“At SimCom, we believe lessons learned from the past can play an effective role in avoiding accidents in the future,” David wrote in a response to AIN’s query. SimCom has put this belief into practice, adding a new product called accident intervention training to its curricula.
According to David, “Accident intervention training entails using selected NTSB accident reports as a basis for classroom analysis and discussion. This is followed by hands-on exercises in the simulator, allowing our customers to personally experience the events that led to some of those same accident scenarios. Having ‘been there and done that,’ our customers learn valuable responses that prepare them for avoiding the same or similar accident scenarios in the real world.
“Benefits of accident intervention training are clear. While discussing the circumstances surrounding a particular accident in the classroom is valuable, actually experiencing similar sights, sounds and aircraft responses in the left seat of the simulator is much more beneficial. Being able to practice and experience in the simulator what you can’t in the airplane is the true value of simulator training. SimCom believes replicating accident scenarios, to the extent possible, is an area where simulator training excels. Accident profiles that can be safely duplicated and practiced in the simulator simply cannot be accomplished by training in the airplane.”
While SimCom sees the benefits of accident intervention training, there are challenges, according to David. Closely replicating the exact circumstances of an accident is difficult, but modern simulators make this possible, he said. “The simulators we operate have the sophistication and technology to closely duplicate a number of physical, visual and aural cues, placing our customers [in] circumstances [similar to those] encountered in the actual accident. Need evidence? Just ask our customers as they emerge from the simulator. There are some damp shirts and satisfied expressions that go with [students’] knowing they have learned something meaningful.”
Another challenge is regulatory. FAA regulations (Part 60 for simulator qualifications and Part 142 for training programs), David noted, “dictate what we can and cannot offer related for training in airplanes that require a type rating. As a result, simulator training providers such as SimCom can’t offer training outside what is authorized. For example, we are not allowed to duplicate an accident that would take the simulator into a region of operation that is beyond what is documented by flight-test data from the actual airplane.”
Finally, adding accident intervention training that is not normally included in a Part 142 curriculum will add time to the training event, which is already tightly constrained.
“Despite these challenges,” David said, “SimCom incorporates accident intervention training into many of our training courses and curricula. Duplicating accident scenarios in the simulator and the resulting lessons learned are potentially life-saving experiences for our customers.”
Simulators That Stall
Dr. Sunjoo Advani, president of International Development of Technology based in The Netherlands, is also chairman of the International Committee for Aviation Training in Extended Envelopes (Icatee). Advani’s company develops simulator solutions for various industries, including aviation. Icatee is developing effective techniques for upset prevention and recovery training (UPRT), to prevent loss-of-control accidents.
One problem Icatee faces is that UPRT can’t be done effectively solely in a classroom or a simulator or an aerobatic aircraft. All three need to be combined to give pilots the training they need to prevent loss of control. “You need the academics, the practical experience of understanding what it feels like to enter a stall and recover a stall so you can avoid a stall, and you need the simulator to create a realistic environment,” he said.
One of the problems with simulators is that they aren’t designed to provide a realistic experience outside their aerodynamic envelopes, and those stop at the moment just before the full stall is reached, according to Advani. “For every Level D simulator, you have a specific data package for that particular aircraft, engine and avionics configuration. That has to–by FAA and ICAO standards–meet objective criteria until a certain point, and that point is approach to stall, or the point of maximum lift. After that point, in current simulators, nothing exists.”
The reason simulator aerodynamics don’t go beyond that point is because, he said, “manufacturers didn’t want to take responsibility,” and it was assumed that well designed flight controls would help properly trained pilots avoid stalls and prevent accidents. “Well, it just so happens that the accidents are occurring in that region,” he said.
Icatee is working on research that will supply ICAO (and thus the regulators in countries that are ICAO members) guidance on not only how UPRT could be delivered effectively but also on how simulators could be modified to teach real stall training. “We have a simulator standards document that will define the technology standards that will be used in future simulators.
“We’re not going to force all simulators to be able to spin upside down and to start centrifuging their pilots,” he said. “What we want to do is to create better scenarios in simulators that teach upset prevention and recovery training. We want to create a more rigorous training environment with the use of simulators. And in later phases of the project, what we expect from simulators is better buffet simulation and a representative model of the stall.”
While Advani sees plenty of opportunity to improve pilot training with a proper combination of UPRT tools, he is also wary of the possibility of negative training using simulators. “We want to avoid negative training by showing pilots what can go wrong if they do the wrong thing. The danger of negative training in simulators is significant. [For example, in the crash of Airbus A300-600] American 587 [near New York JFK Airport in November 2001], the pilots had been taught in the simulator to use rudder in such incidents; they did use rudder, and that led to the structural failure of the fin, causing the crash. So we’re aware of that. And that’s our number one goal at Icatee, to avoid negative training.”
Industry support for Icatee’s efforts is strong, with more than 90 participants in the working group, according to Advani. “We’re not here to crank up the cost of training; that will be a consequence, of course, but we all realize that this is something to be taken seriously. If we want to deal with the number-one threat in aviation safety, then the industry is going to have to face up to that.”