Safety Against the Odds

U.S. Air Force

The e-mail we received here at Flying from Col. Sid "Scroll" Mayeux, chief of aviation safety at the United States Air Force Safety Center, was a little hard to believe. "Last year (Fiscal Year 2009)," Mayeux's e-mail read, "was the USAF's safest year in aviation safety, with 17 Class A Aviation Flight Mishaps for a 0.8 rate per 100,000 flying hours."

One might think that the job of attaining a level of safety like that, given the Air Force's high-flying, high-tech fleet of aircraft, was an impossible task, and I would have been right there with you. Somehow, though, the Air Force seems to have hit upon a formula for safety that last year approached perfection.

Air Force Safety in Context Before you can fully appreciate just how remarkable an achievement the Air Force's safety record is, you have to understand just what it means.

For starters, it is important to define what constitutes a Class A Mishap. Just as the NTSB and FAA have their specific definitions of what constitutes an accident versus an incident, so does the Air Force. The bar for an event falling into the Class A category is surprisingly low: It's any accident in which there's a fatality, permanent disabling injury, destruction of an Air Force aircraft or property damage of $1 million or more. That repair-cost figure is going up soon to keep pace with the rising costs of repairs. You can hit that figure, one investigator commented to me, by putting a healthy gouge in the paint of an F-22 radome. So, while some of the accidents that get listed as a Class A Mishap are high-speed crashes resulting in loss of life, others aren't much more than glorified fender benders. By civilian standards, the rate might be even lower.

Just how good a rate is 0.8 per 100,000 flight hours? It's, in a word, remarkable. The rate compares favorably with the fatal accident rate for general aviation, which is around 1.17 per 100,000 hours. Remember, most of the Air Force's Class A Mishaps don't involve fatalities, and many of them don't involve injuries.

The more pertinent figure from GA, the overall accident rate, in 2008 was 7.1 per 100,000 hours, which is approximately nine times that of the Air Force's mark. In fact, the Air Force's safety record for 2009 compared favorably with every segment of civil aviation in the United States (based on 2008 figures) except for the scheduled airlines. Scheduled Part 121 flying, as one would hope, is considerably safer. Then again, the airlines aren't flying high-speed, low-level training missions through mountainous terrain.

While 2009 was the safest year on record for the Air Force, the trend of safety is not new. Since the early part of the new century, accident rates have been lower, substantially lower, than historic trends have been.

As recently as 1980, there were 84 Class A Mishaps with 74 aircraft destroyed and 94 fatalities. It used to be far worse than that. In 1950, shortly after the Air Force as we know it was born, there were 1,744 Class A Mishaps, with 665 aircraft destroyed and 781 lives lost. None of these figures, it's important to point out, involved combat actions. So in that respect, at least, you can compare different years and different segments of aviation and still have the numbers make sense.

It's no secret that a lot of the safety gains over the past few decades have been due to improvements with equipment — there are only three piston-powered aircraft remaining in the main fleet: the Diamond DA40 (T-52), the Cessa 172 (T-41) and the Cessna 150 (T-51) — so the move to, effectively, an all-turbine fleet has to count for a lot.

It doesn't, however, account for everything. The Air Force fleet is varied and contains a number of platforms that might qualify, at least in theory, as antiques, including the C-130, the U-2 and the B-52, all of which have been in active service for around 55 years.

So, how does the Air Force do it? I was convinced that there must be some kind of secret to its success, and I wanted to find out what it was.

The Safety Center
As the name suggests, the Safety Center is in charge of Air Force safety. Toward that end, the Center creates safety policy, provides guidance on implementing it and practices oversight during the whole process, which is never-ending.

Housed in a modern, unremarkable office building at Kirtland Air Force Base in Albuquerque, New Mexico, the Safety Center houses 160 workers, many of them civilians — many of them, in fact, are retired Air Force pilots. Heading the center is Maj. Gen. Fred Roggero, a career mobility pilot who also served as the Air Force's chief of integrated marketing earlier in the 2000s. The team he has at the Safety Center is impressive, and the wealth of aviation knowledge is humbling. For instance, the person running the projector during the introductory briefing, Randy Rushworth, is a retired, decorated Air Force bomber pilot with thousands of hours in B-52s, B-1s and B-2s and an expert on these and other aircraft at the center. During my daylong visit there, I met a 2,000-hour pilot in remotely piloted aircraft, a flight surgeon who flies in the CV-22 Osprey, a pair of recorder analysts who have worked on some of the most high-profile accidents in recent military history and a structural engineer who is among the foremost experts in the world on aging aircraft.

Investigating accidents is a central job of the Safety Center, and its Aviation Safety Division's experts provide on-site and remote assistance and consultation to safety investigators looking into accidents that might have been caused by structural, powerplant and/or electronics problems. This is a tall order considering the number and variety of aircraft in the fleet, which is composed of everything from Diamond DA40 piston single trainers to vertical takeoff and landing Ospreys to the most advanced fighter in the world, the supersonic, thrust-vectoring F-22 Raptor.

One of the most fascinating places to visit at the Safety Center is the office of the Bird/Wildlife Aircraft Strike Hazard (BASH) program. The Air Force's awareness of the hazards that birds present to aircraft is not new. The Air Force put together the BASH program back in the mid-1970s, and it has been a part of the Safety Center for longer than 15 years. The need for action is clear. During a 23-year period starting in 1985, there has been an average of nearly 3,500 bird and wildlife strikes per year on Air Force aircraft. The overall cost associated with strikes during that period is $817,817,683, or more than eight-tenths of a billion dollars.

There have been some high-profile losses too. Three airmen were killed when a B-1 bomber went down in Colorado in 1987 after a bird struck the wing's leading edge, leading to the loss of fuel and hydraulic lines and, subsequently, control of the airplane. In 1995, 24 airmen lost their lives when an E-3 went down shortly after takeoff from Elmendorf, Alaska, after its engines ingested a number of Canadian geese. In the early 2000s, four jet fighters — two F-16s, an F-15 and an F-22 — were lost over a short period of time, all from bird strikes.

In response to the hazards, the BASH program has developed a number of tools that pilots and mission planners can use to minimize the risk. Keeping birds away from airfields is key to the effort, but the BASH team has also helped to develop a number of other tools, including a worldwide Air Force bird-strike database, creating a computerized low-level Bird Avoidance Model (BAM), which shows where birds are likely to be before a given mission is flown, and the Avian Hazard Advisory System (AHAS), which combines Nexrad radar, weather information and thermal activity along with historical bird migration patterns and soaring data to provide near-real-time bird hazard advisories. (BAM and AHAS are both available to civilian pilots too, at A new tool that shows great promise is bird detection radar (BDR), which is installed in a small trailer and can be easily towed from location to location to provide radar detection of birds up to 7,000 feet agl and at a range of up to eight miles. When there is a strike, the team uses feather and even DNA analysis to pinpoint the species of bird, which then goes into the database.

Located in another corner of the Safety Center is the Mishap Analysis and Animation Lab. When the recorder or recorders are recovered from a crash, the analysis team goes into action, extracting the memory from the recorder to keep it. And because the fleet is so varied in age and description, the recorder's team needs to be able to work with many different kinds of hardware of many different vintages. Even reading the data can be a challenge. The lab has on more than one occasion had to buy vintage computers on eBay in order to run data extraction programs written in the early days of PCs. Still, its success rate is remarkable in being able to extract data from even very badly damaged and burned recorders. That data, needless to say, is often the key to understanding what went wrong in an otherwise baffling accident.

A Different Kind of Investigation
When an Air Force aircraft goes down, the process of responding to that accident is different in many respects from what it is in the civilian world, and the Safety Center is involved every step of the way, either directly or indirectly, from first response until the book is closed on the accident.

The Air Force actually conducts two investigations whenever there is a major accident. The kind of investigation in which we're interested addresses safety and is called, sensibly enough, a "safety investigation." The timetable for its investigations is shockingly short by civil standards. The safety board investigating a Class A Mishap has only 30 days to reach and publish its findings. While the board can ask for an extension, Mayeux said that this is rare.

"There's usually a smoking gun," he said, "and we can generally pinpoint the causes fairly quickly."

The safety investigation, which is one of the prime tools the Air Force uses to improve aviation safety, has few distinct phases, and it differs from a civil investigation in several important ways.

In the direct aftermath of the acci-dent, a preliminary team — called the Interim Safety Board — is formed by the commander of the air base closest to the crash site. This team's job is specifically to gather evidence and protect it, as well as to secure the crash site, which itself can be a hazard. The members of this team have all received training at the Safety Center in Albuquerque, so when the call comes in, they know exactly what to do.

As the evidence-gathering phase is being conducted, a process that typically takes just a few days, the Permanent Safety Investigation Board is formed by the four-star general commander of the major command that "owned" the aircraft or crew.

The job of the Permanent Board is to do all the things that the NTSB might do during an investigation — conduct a full forensic workup of the crash site, do any pertinent structural analysis of the evidence, tear down the engines to find any problems that might have precipitated the mishap, analyze the flight data and cockpit recorders (if present) and examine any pertinent human factors. This, remember, all needs to be done within 30 days.

In its report the board is also required to come up with findings and generate actionable mishap recommendations, which it passes along with its findings. The NTSB does this on some of its investigations by issuing recommendations, usually to the FAA, for proposed changes in regulations. The Air Force safety investigation board does this with every accident investigation.

While the safety board's recommendations, like the NTSB's, aren't technically binding, they do carry a lot of weight and have a ripple effect of responsibility once they're issued.

As an example of how recommendations make a difference, Mayeux used a mishap several years back involving a T-38. When it was designed, Mayeux said, the T-38 was intended as a high-altitude jet, and it had a canopy and windscreen that were designed with this mission in mind. Over time, though, the Air Force began to use it for low-level training, where it came into much more frequent contact with birds. When a T-38 on a low-level training mission struck a bird and was destroyed, the accident board found that the windscreen wasn't strong enough for the added bird-strike risk, so it recommended that the windscreen be replaced with a thicker and stronger component. In four high-speed Class A T-38 bird strikes since, not once has the bird penetrated the windscreen — and all pilots survived.

After the board issues its findings and recommendations, the report is open to comments for 45 days. At this time, if a pilot or crew member has been found to be "causal" in the accident, he or she may submit a rebuttal, and that rebuttal is published along with the rest of the findings.

After the report is published, the responsible parties, usually the commander of the fleet in question, addresses the actionable items, such as installing better glass in T-38s. Given financial, operational and design constraints, taking action, or not, can be a difficult call. In the case of the T-38 mishap, the general who was in charge of the fleet at the time chose, for financial reasons, not to install new canopies but to install new, stronger windscreens. It turned out to be the right call. It is, Mayeux said, highly unusual for the decision makers to defer the changes or simply accept the risk and move on, and when changes are deferred or risk accepted, it's never, Mayeux told me, done lightly.

To ensure full participation, personnel who are involved in the mishap and provide testimony are immune from prosecution or disciplinary action as a result of the safety investigation. Furthermore, no findings of the safety board can be used against them. Doing so, for that matter, is a serious infraction. The Department of Defense says that this immunity is an extension of the executive privilege afforded the president.

One of the key jobs of the Safety Center is to provide training to Air Force safety officers, as well as to personnel from allied forces, in all subjects related to accident investigation and prevention.

The most visible of these training facilities is the Crash Lab, situated on a 29-acre expanse of desert a short drive from the Safety Center. It is here that aspiring Air Force safety investigators go to apply the skills they've learned in the classroom to a real-world accident site. Spread out over the Crash Lab site are wrecks of 10 Air Force aircraft that went down under various scenarios. Students learn how to secure the site and then look for clues, analyzing the fire patterns and distribution of the wreckage as well as examining the engines for signs of whether they were developing power or not when the crash took place.

Most of the learning takes place in more traditional classrooms, however, and they are busy places. The Safety Center offers 110 three-week classes each year, training around 1,300 personnel, most of them Air Force personnel but many of them from the services of our allies. Students learn safety principles and the fundamentals of investigations and prevention. All courses are intended to prepare them to serve actively as flight safety officers, safety board presidents or interim board members, among other positions. It's a big job, and an important one, because the investigative process is distributed across the Air Force community and positions turn over regularly when personnel leave the Air Force or move on to other positions.

While training is a critical part of its mission, the Safety Center can hope to have direct contact with only a few thousand airmen a year, so it actively gets the word out in a number of ways, including through its award-winning quarterly publication Wingman. A recent issue featured articles on configuration management for pilots, using operational safety management principles at home, and reducing automotive accidents — yes, the Safety Center is tasked with improving Air Force safety for every airman and every airman's family members. The Safety Center, in case you were wondering, is also on Facebook. Times have changed.

Human Factors
Underscoring the fact that the Air Force sees safety as a complex web of interactions, technologies, information and attitudes is the Human Factors Division, which is integrally involved in every aspect of the Safety Center. Headed by Lt. Col. Karen Heupel, the Human Factors Division has a broad range of responsibilities, everything from assisting in investigations to determining what human factors entered into an event — Heupel, like everyone at the center, is convinced that human factors are at the core of every investigation. Human Factors has on-staff psychologists, aviation physiologists, flight surgeons — Heupel herself is a flight surgeon — and egress specialists. Even the accident investigation process is subjected to human factors analysis. There is the rule of five "whys." When looking at the cause of a mishap, you ask why that factor happened. When you identify that "why," you again ask why. "If you don't get to five 'whys,'" Heupel says, "you're not looking closely enough."

New Technologies
The Air Force employs the most cutting-edge flying technologies in the world, some of which surely haven't been heard about. At the same time, it employs new technologies in order to improve safety. Some of those, like GPS, were developed for the military and today play a major role in the civil aviation segment. Its bird avoidance radar, night-vision goggles, head-up displays, foreign object damage (FOD) radar and aging aircraft structural analysis techniques are just a few examples of how Air Force initiatives have provided new tools to make flying safer. Likewise, the Air Force employs successful strategies developed in civil aviation, as in its MFOQA (military flight operations quality assurance), a military version of a successful airline/FAA program to enhance safety.

New technologies, however, bring new challenges. The Safety Center is in charge of mishap investigation, policy and procedures for remotely piloted aircraft (RPA). The dramatic growth of RPAs in the fleet has raised some interesting questions. For instance, when an RPA crashes, let's say in Afghanistan, where is the accident site? Is it where the vehicle crashed? Or is it in Nevada, where the remotely based pilots were flying it? Or is it where the close-in pilots had taken control at the airfield in Afghanistan? The answer the Air Force has come up with is yes to all of the above, so investigations into RPA losses are complex, to say the least.

Every Airman a Safety Officer
I went to Albuquerque to discover the secret of how the Air Force has achieved such a remarkable safety record. I came away understanding that there is no secret. It takes a commitment to safety and all that that implies from the top down.

Every person I met at the Safety Center seemed to believe that safety is a dynamic, ongoing process that requires everyone involved to take an active role in the safety process. And it has worked. The results are quantifiable. While it's unrealistic to expect that general aviation can cut accidents to the level the Air Force has attained, there's surely much we can learn, starting with the belief that safety requires hard work, a good look in the mirror and the belief that one person can make a difference. The results of those attitudes have paid off for the Air Force, and its safety success can be measured not only in dollars and cents, but also, and far more importantly, in fewer lives lost.

To learn more about the USAF Safety Center, visit


Your email address will not be published. Required fields are marked *

Subscribe to Our Newsletter

Get the latest FLYING stories delivered directly to your inbox

Subscribe to our newsletter