Barry M. Sweedler
Director, Office of Safety Recommendations and Accomplishments
National Transportation Safety Board
Teledyne Controls 2000 Users Conference
Los Angeles, California
March 14, 2000

`The Direction of Flight Recorder Technology in the United States'

Thank you, Dr. Hassan, for giving me the opportunity to participate this morning in Teledyne Controls 10^th Annual Users Conference. I especially appreciate being invited because, like many of you, I am an engineer and began my career at the National Transportation Safety Board as an engineer and accident investigator more than 30 years ago.

I would like to convey greetings from NTSB Chairman Jim Hall who can't be here today, but sends his best wishes for a successful conference. I would also like to recognize my colleague at the Safety Board who is here with us today, Dennis Grossi. Dennis is our National Resource Specialist for Flight Recorders.

As you know, the NTSB is an independent Federal agency that investigates transportation accidents and makes recommendations to prevent them in the future. As Director of the NTSB's Office of Safety Recommendations and Accomplishments, one of my main jobs is to achieve the highest rate of acceptance possible for the recommendations. In the aviation field, our acceptance rate has been averaging 88 percent over the past five years.

Being an engineer and working for the NTSB has been a terrific experience for me, because like many of you, I have been able to use my engineering skills to help make the world a better place. We have all done this in different ways. At the NTSB we strive to use technology to improve transportation safety.

We've been able to do that by using engineering advances in numerous areas, but the one I'd like to talk about today is data recorders. There have been major advances in recorder technology that have tremendous potential to spot safety problems before they become accidents, prevent accidents, figure out what happened in accidents and ultimately save lives.

Unfortunately, many times safety regulators and industry are slow to adopt the latest technology. Which is odd considering that in this country today, it's commonplace for kids to play with state-of the art computer games. And for adults to try to impress each other with the smallest cell phone or a satellite tracking device in their car or even in the palm of their hand. What's wrong with this picture? Nothing, really, except it leaves the impression that high-tech has crept into every corner of American life. It hasn't. In fact, we are woefully behind in applying the latest technology to improving efficiency, productivity and safety in all transportation modes.

Since joining the NTSB my eyes have been opened to just how far we have come -- and how far we have to go -- to improve transportation safety. I know I have just a short time before you today, but I hope that everyone of you will go back to your offices and homes with the same commitment I have -- to use the technology we have right here and now to reduce transportation accidents, injuries and deaths and the huge emotional and financial burden it heaps on all of us.

0ne of the most effective tools available is the data-recording device. For many people, the only recording device they're aware of are the "black boxes" that are pushed into the national spotlight following a major plane crash. The availability of accurate data reduces the time and resources needed to complete an investigation, the cost to taxpayers and - most importantly - the speed at which corrective actions can be taken.

Last March, the Safety Board concluded its lengthy investigation of USAir flight 427, a 737 that crashed on approach to Pittsburgh in 1994, killing all 132 on board. Imagine how much of that four-and-a-half years could have been saved if the airplane had not been equipped with a relatively primitive flight data recorder -- one that was only required to record 11 parameters and had two additional engine parameters. That recorder tracked altitude, indicated airspeed, heading, microphone keying, exhaust gas temperature, fuel flow, compressor speed, fan speed, roll attitude, control column position, pitch attitude, longitudinal acceleration, and vertical acceleration.

Crucial parameters missing from flight 427's flight data recorder -- and not required by the FAA at the time of the accident -- included pitch control surfaces positions, control wheel and lateral control surfaces positions, rudder pedal and yaw control surfaces positions, and lateral acceleration - all crucial to the investigation. This lack of information made the investigation extremely tough and challenging, the longest aviation, and possibly the most complex, accident investigation in the NTSB's 33-year history.

NTSB and party investigators spent more than 100,000 hours on the investigation. They logged more than 6,000 hours alone on performing scientific aircraft rudder tests to determine the impact of pilot inputs, hydraulic fluid contamination, yaw damper failure, rudder power control unit failure, structural failure, electrical short circuits and effects of extreme temperatures. Information from more than 60 flight data recorders had to be analyzed from 737 flights in which pilots reported uncommanded rolls. Many of the 737s we investigated also did not have extensive recorder parameters. So most of the 737 incidents required painstaking analysis -- sometimes months or even years -- to estimate and extrapolate parameter values that could have been recorded. And there are 14,000 pages of technical information on the investigation in the public docket.

In contrast, a sophisticated flight data recorder tracked 98 parameters on an ATR-72 that crashed less than two months later in Roselawn, Indiana, killing 68. That recorder yielded a vast amount of valuable data including aileron, rudder, elevator, and spoiler flight control positions. It also tracked angle of attack, total air temperature, deice boot activation, autopilot status and flap position. Within a week of the crash of American Eagle flight 4184, flight data enabled investigators to quickly focus on critical areas of the investigation. Comprehensive recorder data prompted the NTSB to issue urgent recommendations on the operation of that aircraft in icing conditions - just 8 days after the accident.

Scientific information from that flight data recorder gave us the confidence to immediately urge the Federal Aviation Administration to prohibit the operation of ATR airplanes in known icing conditions until the effect of upper wing surface ice on the flying qualities and aileron hinge moment characteristics were examined. Eight days after the accident we also urged the FAA to require air traffic controllers to provide expedited service to ATR pilots who request route, altitude and airspeed deviations to avoid icing conditions; to caution ATR pilots that rapid descents at low altitude or during landing approaches or other deviations are not an acceptable means of minimizing exposure to icing conditions; to provide guidance and direction to ATR pilots who inadvertently encounter icing condition; and to conduct a special certification review of the ATR. By the next June, the manufacturer was installing a newly redesigned deice boot system on ATRs.

In 1996, two Boeing 757s crashed into the sea within months of each other. A Birgen Air flight crashed off the coast of the Dominican Republic in 7,200 feet of water, killing all 189 on board. The other, an Aeroperu flight crashed in more than 600 feet of water off the coast of Peru. All 70 onboard that flight died. In both cases, the wreckage was at a depth that would have made recovery very difficult if not impossible and extremely expensive. It was a major undertaking by the U.S. Navy just to locate and retrieve the Birgen Air recorders.

These were two of the first fatal 757 accidents. The recorders had the ability to track 350 parameters giving investigators the data they needed to precisely define the problems and determine the crews' actions or inaction to cope with the situations. As a result, in the Birgen Air case the only wreckage recovered were the flight recorders. In the Aeroperu accident, a few selected parts identified as problem areas by the recorded information were also recovered. In both cases, millions of investigative and recovery dollars were saved -- and confidence in the aircraft maintained -- because of the information provided by the recorders.

Just last November, we were able to learn a great deal of information about the EgyptAir flight 990 crash well before we began lifting wreckage of the 767 from the ocean bottom -- thanks to the recorder's 155 parameters. However, because there was a loss of electrical power before the plane crashed, we were deprived of valuable information about that flight's last moments.

We are seeing more widespread use of recorders with dozens and even hundreds of parameters, but the loss of power to the EqyptAir recorder and electrical losses during the final moments following the TWA flight 800 explosion and Swissair flight 111's plunge into the North Atlantic are very troubling. After the Swissair accident, the NTSB and the Transportation Safety Board of Canada issued several aircraft recorder recommendations asking for independent power sources to keep recorders running for 10 minutes after an aircraft's main power supply is cut, and redundant recorders in the front and back of planes. The FAA has made a commitment to make this requirement technically possible and mandated.

As you all know, over the years, recorders have proven their worth by helping us prevent future accidents. For that reason, the use of state-of-the-art automatic information recording devices on planes, trains, ships, buses, trucks and cars is on the NTSB's Most Wanted list of safety improvements.

Look at how dramatically the airline accident rate has dropped over the last four decades. In 1960, there were 1.2 fatal accidents per hundred million aircraft miles flown. In 1970, that number had dropped to 0.1, and it stayed in that range for many years. By 1998, the rate had dropped to 0.02. That's about one-sixtieth the fatal accident rate of 1960.

Data recorder information has also been a catalyst spurring research that has radically changed the way we look at things. In the early 1970s, very little was known about certain extreme weather phenomena that posed serious problems to aircraft, particularly during takeoffs and landings. It was only after recorders began to capture information during accidents and incidents that windshear came to be understood.

A 1973 nonfatal accident in Boston, involving an Iberia Airlines DC-10, gave investigators their first clue into what pilots were experiencing. The documentation provided by the plane's expanded-parameter flight data clearly identified low level windshear as a serious threat to aviation safety. Without this flight recorder information, the cause of this accident might have been labeled "pilot error" and quickly forgotten.

However, once it was identified, the process to combat windshear gained momentum. Research began at a number of facilities. Projects like the Northern Illinois Meteorological Research on Downburst and the Joint Airport Weather Studies, conducted in the late 1970s and early 1980s, provided extensive insights into windshear. Universities, the FAA and NASA also joined forces in the effort.

Later, recorder data from the crash of Delta Air Lines flight 191 in Dallas provided an in-depth analysis of microbursts and led to the identification of low level vortex rings. Based on the new recorder data, training was changed throughout the industry to help pilots better deal with windshear encounters.

There's still more to accomplish, particularly in the area of windshear prediction and avoidance. However, since the Dallas accident, we have only had one fatal windshear-related airline accident in this country -- the crash of a USAir DC-9 in Charlotte in 1994.

This is only one example of how information from data recorders has jump-started industry-wide research to identify and solve major safety problems. Others include the development of safety devices like ground proximity warning systems and anti-collision systems, both in the air and on the ground -- all brought about in large measure by information learned through voice and data recorders.

What's next on the flight data recorder horizon? Last May the NTSB and the International Transportation Safety Association held a three-day recorder conference that attracted more than 300 transport experts from 11 countries and representatives from 8 accident investigation boards. During that conference, NTSB Chairman Jim Hall and FAA Administrator Jane Garvey announced the formation of a joint group of experts from industry and government to look ahead to the year 2015 and beyond to see how new technologies can be used for flight data collection and application. The committee will draw on expertise from international safety boards, manufacturers, flight crew organizations and regulatory agencies. This is the first time I can think of in which the goal of designing future recorders is to meet the fleet management needs of industry, oversight responsibility of regulators and data requirements for investigators.

And you may have recently read about the possibility of including video-recording devices in the cockpit. As a first step, the NTSB recently recommended that some turboprop planes be equipped with video flight recorders and we are preparing for that eventuality. This recommendation stemmed from a 1997 accident in which the pilot and nine U.S. Interior Department employees where killed when a Cessna 208B cashed in Colorado. Many types of aircraft that carry a small number of passengers are not required to have flight or cockpit recorders. Video recorders are a good alternative for aircraft with analog instruments in which it would be extremely difficult and costly to retrofit with today's digital data recorders. Earlier this year, the U.S. House of Representatives, based on our request, provided the same privacy protection for any future video-recording devices in any mode of transportation that currently exist for cockpit voice recorders. This provision is pending in the Senate.

This is just one of dozens of flight and cockpit recorder recommendations the NTSB has made over the years. Some recent ones on our radar include urging the FAA to create a national certification team of flight data recorder system specialists to approve all supplemental type certificate changes to recorder systems, and to provide FAA principal avionics inspectors with training that addresses the unique and complex characteristics of flight data recorder systems.

Because of several accidents in which there were major problems with data recorders, we also want the FAA to require maintenance checks and repairs for all aircraft operated under Parts 121, 129, 125 and 135 every 12 months until the effectiveness of the new advisory circular and new FAA inspector guidance on continuing recorder maintenance and inspections is proven. All these expanded parameters will not do anyone any good if the recorders do not operate properly because of improper installation or poor maintenance.

And although we had serious concerns at first with the FAA's timetable to implement previous recommendations to mandate expanded parameters on 737s, the FAA is now on board and has promised to meet our new recommendations and timetable. As part of last March's final report on USAir flight 427, we have asked the FAA to mandate additional parameters for 737 aircraft, including pitch trim, trailing edge and leading edge flaps, thrust reverser position, yaw damper command, yaw damper on/off discrete, standby rudder on/off discrete, and control wheel, control column and rudder pedal forces.

The collection and analysis of recorder data is also vital to the success of other safety initiatives such as Flight Operations Quality Assurance - or FOQA -- which is a main topic of this conference later in the week.

As I said earlier, there had been a steady decline in the accident rate over the years, but we have hit a plateau and I believe FOQA presents us with an opportunity to push that rate lower if we can point out potential problems and take corrective steps before accidents happen.

As you know, FOQA collects, analyzes, and shares data on routine flight operations. From the experience of European carriers and a two-year FAA demonstration study, we learned that the analysis of routine flight data, collected by flight data recorders, provides significant benefits by identifying trends. And information is critical to decreasing accidents.

In the FAA study, one airline learned that its aircraft were routinely approaching a particular runway at a rate of descent that was too steep and too fast. Because of data from the flight recorder, the airline started asking questions. It turned out that the approach had been designed years ago for Convair 240s -- piston-engine aircraft, not the turbine-powered aircraft in operation today. The FAA corrected that approach. Programs like FOQA that use accurate and adequate recorded data are vital to improving safety - and hopefully key to lowering the accident rate.

While this is an aviation-oriented conference, we are in Los Angeles. So, we can't ignore what is going on right outside this hotel on the massive spaghetti of superhighways that criss-cross Southern California. Please let me digress for a few minutes. Recorders have made the skies safer and we need to apply that technology to other modes of transportation - -especially on our highways. It's hard to believe that we're not already using every available technology to solve our nation's number one safety problem.

Highway crashes now account for more than 40,000 deaths and three-and-a-half million injuries a year. They are the leading cause of death for young Americans ages 6 to 27. Since 1990, over 72,000 children under age 20 have died in motor vehicle crashes, including almost 14,000 under the age of 10. That means that this nation is losing 33 of its children under the age of 10 every week in motor vehicle crashes. We've also lost nearly 51,000 teens between the ages of 15 and 20 in traffic crashes since 1990 -- over 122 each week. In addition to the immense human and emotional toll of these crashes, they cost the American economy more than $150 billion each year, much of that borne by the states.

Between 1986 and 1996, the number of vehicles on our highways grew by 16 percent. During the same period, the mass of those vehicles increased by 20 percent. Many of our interstate roads -- which also serve as major truck corridors -- are over capacity and ill suited for heavy vehicle traffic. Safety on our highways is challenged by more volume, more mass, and more pressure.

As "just in time" delivery has helped grow our economy, it has increased pressure on operators, shippers, brokers, and drivers to keep vehicles moving to meet demanding production schedules. The impact of this change in doing business has almost doubled the production of heavy trucks in the past 10 years and resulted in many trucks being used as mobile warehouses for products.

In 1998, 412,000 large trucks were involved in crashes -- killing more than 5,300 people and injuring another 127,000. And although many tout that the accident rate, accidents per million vehicle miles, has declined since 1991, the actual number of fatalities continues to increase.

Our highways are expected to see continued growth in traffic in the coming years. What's more, the number of youthful and elderly drivers is also expected to continue to increase. We expect to see a 22 percent increase in young drivers nationwide by 2005. And, by 2010, there will be 5 million more drivers over the age of 65 on U.S. roads.

Highway traffic regulators and the industry also seem unwilling to make use of available data recording devices despite the fact that there doesn't seem to be any compelling arguments for not equipping commercial heavy vehicles with them. The NTSB issued its first recommendation regarding recorders for highway trucking transport in April 1990 as a result of our safety study on fatigue, alcohol, drugs, and medical factors in fatal-to-the-driver heavy truck crashes. We concluded that these devices could provide a tamper-proof mechanism that could be used to enforce the hours-of-service regulations, rather than relying on drivers' handwritten logs.

Some in the U.S. trucking industry have already installed tens of thousands of these recorders. U.S. Xpress and manufacturers such as Freightliner have taken the initiative to incorporate recording technology into their vehicles. Additionally, engine manufacturers, such as Detroit Diesel, now install engine control modules for tracking and analyzing vehicle performance. We are just now learning how this maintenance device can be applied to safety management activities. These companies are showing what the private sector can do. It is time for the regulators to show what government should do, because until mandated, millions of commercial trucks will be on the roads without even the simplest of on-board recorders.

We believe it is past time to act, and that recorder use should be mandatory throughout the trucking industry, as it is in most of Europe. In fact, this technology is already required in the 15 countries in the European Economic Union. All of the European Union's 5 million heavy trucks have been fitted with recorders. Accident data, particularly in Germany, show that recorders have contributed to a substantial decline in accident and incidents. Only 300,000 of the 7 million heavy trucks in the U.S. have recorders and those have been voluntarily installed.

The NTSB will continue to press for automatic information recording devices in all modes of transportation because every accident, every incident, and even the monitoring of normal operations provide an opportunity to pinpoint problems in our transportation system and to solve them before another life is lost.

However, technology is not a panacea. Each new product must be critically evaluated not only on its own merits, but also as part of an integrated operator information system. We must ensure that advanced technologies do not further distract operators and that the information provided by such technologies is as accurate as possible and promotes appropriate responses.

The legal, privacy, and proprietary implications of recording systems need to be addressed so that the safety of the traveling public and the privacy rights of vehicle operators are both protected. That is why the NTSB is sponsoring a symposium on April 25 and 26 in Arlington, Virginia, to discuss many of these key issues. They include the implications of government accident investigations and private litigation, the proper governmental approach to encourage the availability of data safety assessment in accident investigations; and the proper relationships between accident investigation, regulatory enforcement and criminal inquiry. Many of you in this group may be interested. If you are, I urge you to check out our internet site for more information. That's www.ntsb.gov.

I know I have been emphasizing how important recorders are to finding out the causes of accidents and preventing them, but there are many other benefits of accurate data.

For example, it protects the vehicle designer, because accurate data will help us find a problem with the design, if one exists, and lead to better designs in the future or even redesigns of existing vehicles. Or, the data will show that the design had nothing to do with the accident or incident, at all. Let me give you an example of this.

The fatal crash of an Airbus A320 aircraft during an early demonstration flight in 1988 brought the safety of its fly-by-wire technology into question. The pilot, who survived, claimed that the airplane failed to follow his command to climb and as a result crashed into a line of trees. However, the flight data recorder and cockpit voice recorder information painted a different picture. The recorded data indicated that the airplane operated as designed and that the flight crew did not take sufficient measures to avoid the crash.

In similar fashion, accurate data protects the manufacturer, the operator, the regulator, and the owners and stockholders of companies.

But, most importantly, accurate data protects the American people -- indeed, people all over the world -- in three very significant ways.

First, it protects people emotionally. Information we've learned from recorders throughout the years has surely prevented an untold number of accidents and the loss of or serious injury to loved ones.

Second, it protects people economically. The prevention of accidents reduces insurance costs, and consequently the premiums that each of us pays, and has a profound affect on reducing medical costs and other government costs that often are borne by taxpayers. Accurate data can also preclude the need for long, expensive investigations like the USAir flight 427 accident I've mentioned earlier.

Third, and finally, by bringing surety to accident cause determination and, therefore, to remedial action, accurate data protects the confidence people have in their government and institutions, and in the integrity of their transportation system.

It is time for all modes of transportation, both in the United States and worldwide, to realize the unlimited potential that recorder technology has to offer both in terms of economics and safety. That's why your work over the next few days is so important and can truly protect us all.

Thank you, again, for inviting me to join you today.