Transcript of Voss Remarks Before Aeroclub Luncheon, Washington D.C. June 30, 2009

Thank you very much for the invitation to speak with you all today. It is a pleasure to be here.

I wanted to come here and talk about how aviation safety is looking around the world today: We’ve had better days. Waking up this morning to the news of another accident drives that point home.

I also want to talk about how things look for the future, and what we want to do now to address future trends and challenges, taking into consideration what we’ve learned from the past.

I’d like to start by directing your attention to this slide. It provides a graph charting all the major commercial jet accidents in the past decade. We’ve done a good job of driving these numbers down since the mid-90s. Of course, there is concern that for a couple years in a row we’ve seen the numbers increase. But we don’t normally worry about this too much because these are the sorts of numbers that bounce up and down.  As we all well know, when you have such a small number of accidents, it just takes a few more or less to cause the chart to fluctuate a lot.

What gets more attention is the five-year running average. This slide shows that when you smooth things out, you can see something that looks to be more than an anomaly. You can note that the downward trend has flattened and appears to be trending up. That’s a concern.  When we further examine this chart, we can see that “good” years see about 10 to 12 accidents and the “not so good” years appear to be as high as eighteen. Depending on how some accidents are classified, we are already at 12 so far in 2009, and we are only half way through the year.

If we continue at this pace, this would be similar to the 1999 pace. Many safety professionals will be working very hard to try to have a better second half of this year.

There’s no question that if this trend continues, we’ll see the five year average take a turn for the higher, which is something we haven’t seen in a very long time. In fact, I couldn’t find the data to see when that happened last.

Let’s hope we don’t have to go there.

Now let’s consider what sort of accidents we are seeing. Loss of control in flight is a very big deal of late. Runway excursions are another area of concern.  And finally, we have the common cause of controlled flight into terrain (CFIT), which has seen a bit of a decline. These statistics caught our attention because CFIT has been a more common [accident] cause than loss of control. These numbers indicate that that has flipped.

This next slide shows the accident rate. This is just a snapshot of last year and is taken from IATA’s safety report from a couple of weeks ago. Everyone is seeing the same numbers. We are seeing areas of the world with higher rates as we have come to expect. I returned a few weeks ago from working with the minister of the Russian Federation. There is a great deal of effort going on in the CIS states that are facing a significant problem. And, of course, we still have some of the other usual areas of concern, including the Middle East, Africa, South America. I’m pleased to note, however, there is some improvement in those areas.

The real problem is the way things look to the public right now. I have to say that I haven’t seen this kind of public worry since I was sitting in ICAO in the summer of 2005. You may recall that we had a spell that summer where we crashed an airplane a week. A couple of weeks we had two. That led to a lot of phone calls, many people were very nervous and a lot of structural changes happened quickly. The European blacklist was created that summer, among other things.

So let’s take a look at how things are looking now. If you go back to last summer, twelve months ago or so, we saw the Spanair crash on take off. Shortly thereafter, XL Airways crashed into the ocean during a test flight. This caught our attention when the prosecutors made off with the flight recorder in the immediate aftermath. We’ll talk more about that later.

In December, we had the crash of the Continental flight in Denver. This was runway excursion that fortunately was not fatal. I have to note that this is probably one of the most unmentioned crashes around. Many people don’t even remember we had a crash in Denver. I would guess that that is probably because we had the “Miracle of the Hudson” only several weeks after that.

In February, we had the crash of Colgan Air in Buffalo. That’s been on everyone’s radar screens quite a bit recently as Congress has held numerous hearings. It should still get considerable attention as the FAA and industry begin to address some of the issues raised.

Shortly after Buffalo, Turkish Airways crashed on approach in Amsterdam. There were a number of things going on in that crash that I’ll talk more about later.

Many in the U.S. may not be aware of the Emirates tail strike in Australia. It hasn’t been talked about too much here, but it certainly is on the radar screen around the world. The tail strike occurred on an A340 leaving Melbourne with almost 280 people on board. This was a simple error in data entry by the pilots. The aircraft didn’t set the right power settings for take off. Luckily, the pilots figured it out before it was too late, threw the power forward, staggered into the air dragging the tail, rolled 1,000 feet in the grass, spent the next two minutes trying to get back up to field elevation, if you can imagine that. This is about as close as you ever want to get to an accident without having one. This may ultimately be categorized as a major accident due to the significant damage to the aircraft. It certainly caught our attention as we start looking at patterns.

Then we had the FedEx accident in Tokyo, followed by Air France 447, which is attracting a lot of attention right now. We are coming up on the 30-day limit for the guaranteed operation of the locators pinging from the black boxes. At this point, we start to lose hope about finding the flight data recorders and the cockpit voice recorder.

The thing that disturbs me about Air France 447 is not that there are a lot of rumors going around – that is normal. But the fact that some of these rumors are valid concerns. One of the signs of a bad safety system is when there’s an accident and when you go to a bar, or a blog now, and hear people saying “oh yeah, we all knew this was going to happen. It’s just like what happened elsewhere.” That’s what we are hearing in this case, and that is discouraging. There’s a great deal of stuff being talked about in the blogs and being talked about among the safety professionals about risks out there that haven’t been mitigated. If we don’t find out what really happened here, we’re going to have a lot of those hanging over us.

Then of course, today Yemenia lost that aircraft off the coast of the Comoros. Of course, it is far too early to tell what happened there, but I will tell you that it will have a very significant political effect. Politics in Europe drive a lot of this industry. When the EU blacklist was created in 2005, it was in response to a near hysterical reaction by the French public to a lot of deaths of French citizens around the world. There were 60 French citizens on this Yemenia flight. And the Europeans are saying that this aircraft had some ramp checks with difficult results in 2007. We can expect this to go into hyper-drive politically in the European commission.

Out of all of this, what we are seeing are new trends.

We’ve shown in the past decades that we can do some things. Industry and government worked together in the early days of CAST to do a lot to reduce CFIT.  We have largely succeeded. We took away the low hanging fruit in this category by developing new technologies, like EGPWS.

What we are seeing now is a much more complicated set of problems. We’ve seen an increasing percentage of fatal accidents caused by loss of control. We’re back in the human factors business. We are trying to understand why humans, the aircraft and their automation continue to get tangled up.

I’ve seen some preliminary data of late that seems to show quite a few anomalies that are out there in the world of automation, more than is recorded. The truth is, pilots don’t always deal with automation anomalies very well. This is becoming something that we will need to deal with very seriously. It can also reflect how we are training and preparing our people.

This is not a new thing. If you go further back, think about the Kenya Airways crash several years ago. The pilots were taking off and did not turn on the autopilot as the procedure calls for. Neither crew member realized that the autopilot hadn’t been turned on. Cross check and crew coordination are simple things, but they add up to some loss of control issues that has got our attention.

So that’s where we are. Not a solid spot, but still a great safety record and still a very safe system.

It’s our job to worry, and we are worrying.  And when I look to the future, I worry a little more. Here’s something to think about. The truth is that we are in an explosion of the world’s middle class. This actually hasn’t changed a lick, recession or not. Seventy million people a year are entering that demographic. Ninety million will be entering the middle class by 2030.

You can look at this and think that this will change with the current economic crisis. But it won’t change. These people are already born. They’re living in countries with relatively stable political systems, they’re going to school and they’ve decided that they are not going to live in the village with mom and dad, planting rice. They’re going to be out there doing high tech jobs and changing the world. My family did that in the early 1900s, they made that transition. These people will do it here. I wouldn’t bet against them. The thing that is different, though, is that these people live in areas that require air transportation. So there will be a huge demand for air transportation and people skilled to work in it.

The revised forecast, after taking into account economic realities, can be seen on this slide. This has come out very recently and we are looking at a pretty serious dip. It does look like, though, that the revised forecast [growth] will start to climb back, erase some of the loss and regain the original path in the next few years. At some point, in the longer term, the demographic realities absolutely trump anything else. It happens in economics because people are going to get jobs someplace else, they’re going to live in cities and they’re going to want to go home sometimes and the airplane is the only way to do it.

This chart reflects where this growth will occur. Not a surprise that there will be the biggest growth in Asia, substantial growth in Europe — much of that in the Eastern part.  North America will see a modest growth. In terms of regions, we’ll be number three after a lifetime of number one. If anyone doesn’t think aviation isn’t international right now, who thinks we can put a box around U.S. aviation, I’d suggest that that’s an idea whose time has gone.

I want to also mention something important that I see all over the world. We live in a world that until recently has been going through explosive growth. When the explosive growth occurred, anyone who could recognize an airplane and was qualified and maybe worked as a regulator was getting a job offer for four or five times as much the money in some of the booming areas of the world. So now we have a tough economic downturn where people are tempted to cut corners. And we have regulatory agencies that have been hollowed out. I’m not necessarily saying the US is in this boat. The US is actually an exception to this rule. But when you look around the world, they are under tremendous pressure and this is a risk area that will continue to exist tomorrow when the growth comes back. That is not a matter of if, but when.

So what do we do about some of these things?

This may seem like mom and apple pie for anyone in the safety business, but safety is all about being information driven. We have to find ways to share the information, protect the information that relates to safety and then act on the information. So let me give you some examples of what I mean about sharing the information.

We need to be able to get better at sharing information between competitors. This is something that the Flight Safety Foundation has been doing since its inception in the 40s. And it’s something that we need to get better at industrywide. We’ve had some odd barriers erected between segments of the industry, whether it be here or elsewhere around the world. We’ve had a lot of rules. This is a highly regulated industry with a lot of worry about antitrust that leads to a great deal of concern about who could get in a room together. This concern should not preclude competitors getting together to discuss safety initiative, and it doesn’t. The Foundation has provided a haven [for people] to talk safety.

There is a lot of information out there that doesn’t belong on the blogs or in the bars. It belongs at the table, being shared between competitors if necessary, before we have to start talking about it in accident reports. As far as we’ve come at this — and we’ve come a long way – we’re not very good at it yet. It’s hit or miss.

Also, we talk about sharing information between segments of the industry. There are many things we need to get focused on as we go to a new world with new types of automation. NextGen [air traffic control] is a classic example. Ideally it will share roles, responsibilities and data between the air and the ground. We’re going to need get better at sharing the safety information between the air and the ground. That’s always been a weakness.

Here are some more examples. We have data coming to us all the time talking about unstabilized approaches. Here’s a real world example of segments not working together. There are a number of unstablized approaches if you are flying an aircraft with winglets and you are told to fly 180 (knots) to the marker. You have a choice of lying or flying an unstabilized approach because it is physically impossible to fly 180 at the marker with winglets. We’ve known that for just about forever. My experience in air traffic control in the 80s is that we knew a 757 couldn’t do 180 at the marker, but they are still calling for 180 at the marker. How does this make sense?

We are very slow to adapt to these kinds of changes. We are talking about moving [ATC] separation responsibilities into the cockpit. It’s not enough to be able to work that out on paper, you have to work out a communication system between both actors so that there is a way to exchange the safety information as well. This will be a big adaptation.

The other thing that is very important and cannot be understated is that information much be exchanged between the regulators throughout the world. I spent a good part of my life trying to make this happen at ICAO.

When you take a look at the fact that many regulators are understaffed or undertrained, some will fail at doing their jobs. Can we afford to act like we are the only regulator on the planet, or can we start cooperating and sharing information with others?

It’s really what is at the heart of the bilateral being worked on between U.S. and Europe. Look at it this way — if you have 190 countries out there, standing alone and proudly being their own regulator, then you have 190 potential points of failure. At any one time, some of these people will fail at their job. If you have those same 190 watching each other’s back, exchanging data with each other, there is a great deal of redundancy. If you are an operator who doesn’t want to cooperate with the rules, who doesn’t want to be safe, you’ve got nowhere to hide. You’ve gone from 190 points of failure to 190 layers of redundancy.

Of course, theoretically, it never works that cleanly, but you get the idea. Mutual recognition and sharing the data is vitally important.  This is also the basis for a very uncharacteristic move of the Flight Safety Foundation. This spring we stepped up and took exception to issues related to foreign repair stations as it was being discussed in the House of Representatives. Should the proposed language become law, it could have put some of these relationships into jeopardy.

Now let’s talk about protecting the information as well. In America, it feels weird to talk about criminalization, but it still is a reality here. Public demands accountability. Criminal prosecutions [of aircraft accidents] are becoming more frequent. And this is an issue with a direct interference in safety investigations.

This means something in the US, because aviation is so international. Keep in mind that there are people in Houston waiting to see what will happen on the Concorde prosecution. They’ve been charged in that. We’re not just safe because we are in the U.S. from this issue of criminalizing accident investigation. And, as I mentioned briefly, before that there are examples of direct interference. We have to realize that we have to be respectful of the requirements of justice here and around the world. We can never try to interfere with the appropriate functions of the justice system. People have a right to know what happened. Even though blame doesn’t do much good, it’s a part of the justice system and the justice system must be respected. But on the safety side, those people in the justice system need to know when they are treading on vital issues of public interest, like the recording and data protection that will keep people alive. It will look after their families when they are on airplanes.

So somebody has to ring the alarm. This is something that the Flight Safety Foundation did again just this past spring. When prosecutors in both Italy and France ran over to the crash sites and absconded with the CVRs and FDRs, we thought that was a really bad idea. Prosecutors shoved aside the investigators and said that this was criminal evidence. They wouldn’t share it with the manufacturers for weeks or months. These actions basically hampered the efforts of the system to find out what went wrong and perhaps issue an air worthiness directive and do something about what caused the crash to try to prevent it from happening again.

In 2006, the Foundation joined with other groups and issued a resolution in opposition to the criminalization of accident investigations, explaining to the world why it is bad to criminalize accidents.

Closer to home in the U.S., we’ve made some moves to try to protect the vital information sources such as ASAP and FOQA. These are data programs that are essential to the safety of our system and we’re trying to get them the same sort of protection as cockpit voice recorders. We’ve presented some testimony to the Senate; these ideas should be included in future legislation and would greatly extend those protections. It’s not enough to just have the FAA say that they won’t write up a violation, we’ve got to have more than that. We’ve got to have some constraints and controls so that judges in civil and criminal cases can’t run wild with it.

Now let me talk about acting on the information. We’ve got a lot of information in the backlog that we need to act upon. Recently there was a congressional hearing that examined, among other things, fatigue risk management, especially as it related to the Colgan Air crash in Buffalo. Although I wasn’t able to attend, I’m proud that Dr. Curt Graeber, one of the foremost authorities on this issue, was able to testify on my behalf in great detail about how fatigue is a very big deal. He was able to review the extensive amount of scientific work that’s already been done.

I’m very excited that the leadership of the FAA is stepping forward to accelerate rule making. I would encourage them to lean upon the science, deal with fatigue risk management in a way that it can really start making a contribution to aviation safety. This has been out there for a couple of decades that I’m aware of. It is pretty personal stuff. It has to do with what you do on your days off, what your lifestyle is, how people are scheduled, and it affects profitability. As such, these conversations would come up every 18 months or so and end abruptly when labor and management would decide that they couldn’t agree on very much. At this point, however, I think that time is past. I know both sides of this discussion pretty well and I think they are both ready to really move along with this. This is the future. It is a real human problem that must be addressed.

Likewise, training is something we’ve got to get serious about. It can’t be Band Aid fixes. That’s what I probably most fear. I’m looking at charts that show a slow increase in accidents for the first time and I’m looking at charts that show crowds at the boarding gate ready to overwhelm the system. That’s not an environment where you want to look at incremental fixes.

Training really hasn’t been looked at seriously since pre- World War II, to tell you the truth. There have been adjustments, but we have to ask ourselves what are the right qualifications, how do you measure it beyond simply hours? Hours don’t mean much. You can buy hours. We have to get to the fundamentals of competency-based training and we have to look at a system overhaul. And we have to look at an international overhaul, or else a license you get here won’t be good overseas. It’s a big piece of work, but we don’t have choice. It has to be taken on now before they come over the top.

Of course, we still have to think about what the emerging trends are. As we look at the accidents that are still being investigated today, we’ve got to ask if yesterday’s big scenarios are the same today. And a lot of them aren’t.

When I grew up as a young man trying to get my air transport rating, all you ever thought about at night was how you are going to handle what they call a V1 cut — an engine failure at take off. Well, I will tell you, I fought a V1 cut in an A340 simulator and you just try to keep your hands in your pocket and your feet on the floor. It’s not the big thing anymore.

What is the big thing that pilots and aspiring pilots need to think about at night is what happens if I’m flying along and suddenly I get handed the aircraft in a configuration with which I’m not familiar, with indications that are ambiguous? This is the new life or death challenge that faces a pilot today. Have we really thought about that? Have we really prepared pilot for that?

I’d say at best, marginally. It’s a different type of threat and we have to re-examine our priorities. But the data is starting to come out and it is starting to show more and more interesting combinations of pilots interacting with automation ending up in unusual ways. It’s not just about Air France. Consider the crash that happen Schiphol. Yes, there was a malfunction of a sensor that caused the aircraft to enter a different automation mode. The automation shift wasn’t picked up by the pilot and it was an unstabilized approach. This is \ the same sort of situation as having an unmanaged automation error at 35,000 feet. These are the new types of challenges we have to consider.

In summary, the accident rate is starting to increase and I’m not happy about it, but I can’t lie about it. Unless we have a remarkable second half of the year, we’re probably going to see the five year trend line shift up. The causes are shifting from what used to be. We know we are going to see growth. Don’t be fooled by the economics, because economics are driven by demographics first. Safety information is the key, but you’ve got to share it, protect it and act on it. And today is the time to act upon fatigue and training, two tough issues that have been on the back burner for a long time because of constraints. But I think those constraints are gone and we should seize the moment.

Thank you again for giving me this opportunity to speak. I’m happy to take some questions.

Top of page