Remarks of Barry M. Sweedler
Office of Safety Recommendations and Accomplishments
National Transportation Safety Board

Before the
May 20, 1998

I am pleased and honored to be here this morning . I bring greetings and best wishes from National Transportation Safety Board Chairman Jim Hall, as well as his regrets that he could not be here himself. He especially wanted to congratulate Mr. Van Vollenhoven and the Dutch Government for establishing an independent Safety Board.

Last year marked the 30th anniversary of the U. S. National Transportation Safety Board and I believe that it is one of the finest transportation investigative bodies in the world. Over the past 30 years, in cooperation with other organizations in the United States and abroad, the Safety Board has been instrumental in greatly improving the safety of the traveling public. As you heard Mr. Van Vollenhoven say yesterday, the NTSB has served as a model for other nations that have established similar agencies. The Safety Board has safety investigation responsibilities in civil aviation, and in major rail, highway, marine, pipeline, and hazardous materials accidents. Additionally, the Safety Board conducts special studies and investigations into major safety issues in all transportation modes.

The Safety Board is an independent agency of the United States government whose mission, since its creation in 1967, has been to determine "probable cause" of accidents and make recommendations to improve transportation safety and prevent future accidents. We believe that our independence is the keystone to our success because it allows us to conduct objective and unbiased investigations.

Some of our accomplishments that have saved thousands of lives include:

How do these accomplishments come about?

Today I will focus on the data collection issue. I will be discussing the recommendation follow-up program at a symposium tomorrow.

Finding the probable cause of accidents is not always easy. Each accident presents new challenges and, unfortunately, the answers to investigations are not always found in the wreckage at a crash site. Accident investigation often requires a close examination of relevant data regarding the accident vehicle, the circumstances, and the environment. It is this information that can often help us unravel the history of a problem. Therefore, it is extremely important to have the very best research tools and methodologies available in our quest to find the answers to ever-challenging and difficult accident investigations.

As transportation technology expands worldwide, so must we seek to advance our investigative tools and skills. To do so, we must expand international systems for data collection and improve collaborative analysis of meaningful statistical and technical information on critical incidents.

The Safety Board has been working to advance the use of high quality automatic information recording devices in all modes of transportation for many years. These include rail event recorders, voyage data recorders (VDRs), flight data recorders (FDR’s) or "black boxes" as you probably have heard them called (they are actually orange), and cockpit voice recorders (CVRs). These devices provide crucial, factual information for accident investigations. Just last year, the Safety Board placed recommendations for the mandatory use of these devices on certain marine, highway, and rail vehicles on its list of "Most Wanted" safety improvements. Another recommendation for expanded parameters on flight data recorders was added to the list as the result of two airplane crashes involving B-737s in which insufficient data is available for determining the cause of the crashes. The "Most Wanted" safety improvements are those safety concerns and issues that the Safety Board highlights and believes, when fully implemented, will have the largest impact in saving lives and reducing and minimizing injuries and accidents.

In the aviation industry this technology has been active for many years and it continues to grow.

Certainly, the use of black boxes, both data and voice recorders, have been important in providing information to solve numerous aviation accidents and has guided Safety Board efforts over the past few decades in implementing specific corrective actions to prevent future occurrences. When a B-737 crashed into a bridge near Washington National Airport during a snowstorm in 1982, killing more than 100 people, it was the cockpit voice recorder that helped us solve the mystery of the crash. We were able to derive the speed of the engines based on a sound spectrum analysis of the engine sounds recorded on the CVR. That information, in combination with conversations between the crew enabled us to determine that the engines were not at high thrust as the pilots believed. Likewise, the voice recorder provided the first indication that there was a fire aboard the Valujet plane that crashed in the Everglades in 1996 when shouts of "fire" could be heard from the passenger cabin. Flight data recorders are equally revealing as seen in our investigation of the ATR-72 accident in Roselawn, Indiana in October 1994. In that accident, the airplane's flight data recorder captured information on 115 parameters, even more than required for a new design. Reading out the recorder in our laboratories, we were able to spot the telltale, rapid movement of an aileron control. This led to our issuance of urgent safety recommendations within a week of the accident to improve flying in icing conditions. That's what flight data recorders do -- they help us investigate accidents and they help the industry prevent accidents. In contrast however, the lack of critical data has been a major stumbling block in solving other accidents. The Safety Board called for an accelerated implementation schedule so that all 737 commercial aircraft have improved automatic information recording devices after the 2 unresolved accidents (UAL flight 525 in Colorado Springs, Colorado, March 3, 1991 and USAir flight 427 in Aliquippa, Pa. on September 8, 1994.) Unlike the accident airplane in Roselawn that had 115 parameters on its FDR, these airplane recorders provided only 5 and 11 parameters respectively.

Only through years of painstaking review and analysis have we been able to identify and recommend changes to the rudder system of the 737. If we had adequate FDR data this probably could have been done in a fraction of the time.

The value of enhanced recorders goes far beyond accident investigation. The Safety Board is aware that systems such as quick access recorders (QAR) are in use by a number of European airlines to more rapidly monitor efficiency and safety of operations. Similar efforts are underway in the United States to improve flight operations and quality assurance. Safety information must be analyzed and widely disseminated to be of maximum value. Broadly speaking, safety must be considered as prominent in data collection efforts as the emphasis on more efficient operations and economic benefits. It must be recognized that safety is cost effective.

In fact, KLM right here in Amsterdam, reads out the QAR data each time an airplane returns to Schiplol airport. The flight data is analyzed and any unsafe practices or operations discovered leads to training and procedural corrections. This information is disseminated to all company flightcrews. Using this approach, KLM hopes to discover problems that could lead to an accident – before it occurs.

As you know, not all of the Safety Board’s attention is paid to aviation. The Safety Board is an agency with multi-modal responsibilities. But, after 30 years of aviation accident investigation, we have learned valuable lessons applicable to all modes of transportation. Based on our experience, the Safety Board has made safety recommendations to other modal organizations to expand the installation, and use of automatic information recording devices in all modes of transportation. I will illustrate with some examples.

Current railroad locomotive event recorders are only of limited use because they cannot answer questions dealing with train crew actions. The Safety Board has recommended railroad voice recorders. The Safety Board is also working to expand the number of recorded parameters and improve the event recorders crash and fire survivability requirements. In nearly a dozen major railroad accidents, the locomotive event recorders were seriously damaged, making it virtually impossible to retrieve any meaningful data. Currently, seven parameters are recorded (speed, direction, distance, throttle position, brake application, cab signal aspects, 48-hour length of record). The Safety Board has suggested 33 parameters that should be recorded to assist in accident investigation and to improve the safety and efficiency of train operations including but not limited to emergency call-ins, dynamic brake retardation, distributed power functions, and locomotive voice cab recording of operating crew communications.

Following an accident in 1996 in which a Washington Metropolitan Area Transit Authority subway train ran through a station and collided with an empty, standing subway train, the Safety Board recommended recording devices for rail rapid transit cars. The Board report says that the data provided by such devices could also provide the industry enhanced efficiency and cost effectiveness of their operations.

In regard to marine data collection and improved technology, the Safety Board has a very long history. In 1976, the Safety Board issued its first recommendation to the United States Coast Guard that it require the installation of an automatic recording device to preserve vital navigational information aboard oceangoing tankships and containerships. In 1995, after the collision of a passenger ship and a bulk carrier in the Gulf of Mexico, the Safety Board recommended that the International Maritime Organization (IMO) require all vessels over 500 gross tons be equipped with voyage data recorders (VDRs) and has been working with the IMO to implement this recommendation.

However, even with this positive action much remains to be done to improve our data collection capabilities. If the VDRs are not required to be carried on board a vessel, then critical data will not be collected and available in a major marine casualty. It seems certain that roll-on/roll-off passenger ferries will be the first to have VDR requirements as a result of the Estonia tragedy.

The Safety Board’s efforts in data collection and improved technologies for the highway mode have resulted in recommendations in two distinct areas: first, the installation and use of on board recording devices to monitor hours-of-service of commercial motor vehicle drivers to prevent driver fatigue and second, the installation and use of crash severity recorders in passenger cars to gather better crash information.

We in the U.S. are way behind Europe where these devices have been in use on commercial vehicles for many years. For passenger cars, more reliable data on the consequences of airbag deployment is needed.

In conclusion, the data we receive in all the modes from current recording devices has been extremely valuable. But our technologies have become much more sophisticated and so recording devices can do much, much more. Every accident, every incident must be seen as an opportunity to pinpoint problems in the system and to prevent those dangerous circumstances from happening again. This process is built on a foundation of thorough and accurate data and we must continue to improve and expand these data collection systems.

Speeches and Testimony | NTSB Staff's Speeches