Testimony of Mark V. Rosenker, Chairman
National Transportation Safety Board
At the Defense Safety Oversight Council
Pentagon, Washington, DC
June 4, 2007



Good afternoon.  It’s wonderful to be back in the Pentagon.  Thank you for inviting me to speak to you today about transportation safety. 

The National Transportation Safety Board (NTSB), is an independent Federal agency charged by Congress to investigate every civil aviation accident in the United States and to investigate significant accidents in other modes of transportation — railroad, highway, marine, and pipeline.  The NTSB does not mandate or regulate — our mission is to determine an accident’s probable cause and make recommendations aimed at improving transportation safety.  Our goal is to figure out WHAT happened and then, more importantly, WHY it happened so that we can work to prevent similar accidents in the future. 

To perform this important service, we draw on relatively limited resources.  Our staff comprises fewer than 400 people, covering all modes of travel, and our annual budget is $79 million – I offer this comparison that our budget would fund the DOT for less than 11 hours.  During our 40-year history, we have investigated more than 128,000 aviation accidents and thousands of surface accidents, and we have issued more than 12,600 recommendations to improve transportation safety in all mode  

As I said earlier, the Safety Board has no regulatory authority; how we impact safety is through the quality of our investigations and Recommendations.   For those Safety Recommendations that could have the greatest impact on saving lives we have developed “The Most Wanted List” One of those issues is Runway Incursions.  Over the past few years, we have seen too many close calls.  The worst accident in aviation history was a runway collision occurring 30 years ago on Tenerife when 747s collided killing 587 people.

The hazards of airport surface operations have been a concern of the Safety Board since that accident and as a result, we have issued over 100 safety recommendations related to runway incursions.

These recommendations addressed the need for improvements in air traffic control operations, training and hardware; pilot training; airport signs, lighting and markings, aircraft conspicuity and incident reporting.

Just this year we’ve had 3 near collisions.  Two were at Denver International.  The first happened at the beginning of January when a Frontier Airbus narrowly missed landing on top of a Metro liner that had inadvertently taxied onto the runway.   The second involved a United 737 that almost collided with a snowplow that had crossed the takeoff runway without a clearance.  And just last month, in Las Vegas, the controller cleared an Alaska 737 for immediate takeoff then had to cancel the clearance when it appeared that another aircraft on an intersecting runway was going to enter Alaska’s runway lane’s tires.  Those are 3 very serious events that happened in just the first 3 months of this year. 

In order to ensure that all commercial passengers receive effective protection against the dangers of runway incursions, the Board has urged the FAA to develop and demonstrate the effectiveness of ground movement safety systems appropriate for use at a variety of airport types ranging from large international airports to the smaller regional airports served only by commuter airlines. We do not expect that the answer will be a "one size fits all" system, but will instead require creative use of different combinations of sensors, processors, and warning methods tailored to the requirements of each situation. At small and uncomplicated airports, simple methods of aircraft detection and warning may be sufficient, while at complex and busy airports a more elaborate approach may be needed. 
What is most important in preventing accidents is data.  Data can come in many forms – flight data recorders, cockpit voice recorders, maintenance recorders, etc.   One good way to examine all of this data is Flight Operations Quality Assurance or FOQA programs. 

For over a decade, many airlines in the United States have used technology-based tools to monitor flight operations and improve safety.   These programs collect vast amounts of data during routine flight operations using quick access flight data recorders.   By collecting hundreds of parameters from its flights an airline can mine these data to identify and correct safety issues before an accident happens.  FOQA data can help an airline improve its flight operations, pilot training, and maintenance programs.  However, because effective FOQA programs reflect a partnership between the airline, its pilot union, and the FAA, all stakeholders can work cooperatively to address safety trends on a larger scale – for example improving air traffic control operations.    I must emphasize that an essential component of a FOQA program is that the data must be used in a non-punitive manner.  For the airline, and the industry, to learn from events and trends to make the overall system safer these data should not be used to blame or penalize an individual pilot.  Before I leave this topic, let me state that although many larger airlines in the United States have established FOQA programs, earlier this year the Safety Board asked the FAA to work with regional airlines to implement programs at those airlines.  With increased use of FOQA, data sharing, and partnerships, we believe that the value of these programs to the safety and accident prevention will continue to be realized.

There are many ways to measure safety, but no one can argue that the bottom line is fatalities.  I’ve rounded the numbers for you, not because each and every death is not important, but because I want you to understand the comparative magnitude of risk among the different modes of transportation.  The number of fatalities in marine each year is approximately 800 – the vast majority of these are associated with recreational boating accidents.  About the same number of fatalities—800—occur in rail accidents annually, where the majority are trespassers and rail workers, not passengers.  We also have several dozen pipeline fatalities each year from gas explosions.  In aviation, the average is 700 to 800 fatalities a year, almost all associated with private pilots in small general aviation aircraft.  

By comparison, about 43,000 people lose their lives on the nation’s highways each year.  For aviation fatalities to match those of highway, we would have to have a commercial airline hull loss accident every day. 


Well over 200 million vehicles are registered in the U.S. and their operation results in 7 million accidents, 3 million injuries and 43,000 fatalities annually — those numbers are dramatic, but in the 1990s, highway fatalities dropped approximately 10% and the fatality rate, even with a substantial increase in vehicle miles, dropped even more. These improvements were attributed to reduced speed limits, increased use of seatbelts, the development of airbags and crash-absorbing vehicle frames, and campaigns to reduce drunk driving.

Unfortunately, those decreases in fatalities and injury rates have leveled off since the 1990s.  So, while we accomplished much in the past decade to improve the crashworthiness of automobiles, we have reached some practical limits in combating the physical forces involved in crashes.  It is time to move beyond crash mitigation and enter a new era where technology will help us prevent accidents.   Let’s look at our current state of technology in that regard.  I see three distinct milestones along the road to highway safety:  technologies for crash avoidance; telematics to better inform the driver about the vehicle and the highway; and command and control systems.


The first milestone, crash avoidance technology, is not very far down the road.  In fact, manufacturers already offer this technology in many current car models.  These systems affect stability control, rollovers, lane departures, and rear-end/forward collisions.

In addition to mitigating the number of fatalities and injuries, such technologies can provide a huge economic benefit.  Every day, 19,000 crashes occur on American highways. These crashes incur an enormous cost: $230 billion a year — that’s nearly $800 apiece for each and every one of us.  We can no longer be satisfied with trying to protect people who get into crashes.  We must instead use the technology at our command to prevent crashes from happening.


The second milestone is telematics and it is actually a whole series of markers from now into the future.  Telematics are wireless, location-based services for vehicles and drivers that trace their history back to the days when your neighborhood mechanic linked into your engine diagnostics to give you a report on the health of your car’s various systems.  Today, sophisticated technology provides not only on-board navigation and entertainment services but also the means to a higher level of safety. We’re all familiar with vehicle-based systems like On Star, but research is well on its way to making road-based systems, vehicle-to-vehicle systems, and vehicle-to-infrastructure systems a viable means of promoting even greater roadway safety.

For example, vehicle-centered services, such as remote diagnostics, remote vehicle access, and automatic collision notification, are currently available on many cars.  Survivability increases with quicker emergency response, which is directly related to this technology.

In addition to vehicle-based systems, road-based systems are being incorporated into our highway infrastructure. Vehicle Infrastructure Integration, a DOT initiative, will provide drivers with a sophisticated means for obtaining information about their vehicles and the road.  What more do drivers need to know? How about location-specific weather conditions, route-specific road closures, and work zone status, to name a few?   Location-specific weather and roadway information can be acquired directly from sensors that run beside or are embedded in the roadway.  Such sensors provide real-time information about fog, standing water, or freezing rain.  Adverse weather is associated with 800,000 injuries and more than 7,000 fatalities annually (approximately 1 in 5 fatalities.


The third milestone along the highway to safety is automated vehicle control.  Electronic devices and automated systems used in commercial aviation offer clear examples of how technology can improve our ability to operate in complex environments. With the introduction of electronic safety devices, we can trace the decline in commercial aviation accidents rates.  Beginning in the 1950s, radio navigation aids (VOR/DME), radar, and ATC control technology dropped the number of accidents per year from 4 to 1.  Further refinements came with long-range radar, precision approaches, and secondary radar.  Beginning in the late 70s, early automation offered Area Navigation (RNAV) and Traffic Collision Warning Systems (TCAS).  The aviation industry has since implemented computerized flight management systems, wind shear alert systems, Ground Proximity Warning Systems (GPWS), and fly-by-wire electronic control of aircraft. We are now seeing real-time weather and traffic displays in the cockpit; precision-landing systems for zero visibility conditions, hybrid vision, and remotely operated drones.

Technological advances have made commercial aviation the safest mode of transportation and I believe new technologies may enable us to repeat those successes in highway travel.  In addition to seat belts and airbags, which have greatly increased survivability, automated command and control systems will help prevent crashes, not just mitigate their effects.

To that end, Integrated Vehicle-Based Safety Systems is a new US DOT vehicle safety initiative to build and field-test integrated crash-warning systems to prevent rear-end, lane change, and roadway departure collisions on light vehicles and heavy commercial trucks.  These systems are being deployed in cars as well.

Let me offer a practical example of how this technology can be used.  Every parent's nightmare is to back over a young child in the driveway.  Nearly 200 fatalities and approximately 7000 such injuries were reported last year, though that is surely only a fraction because many of these events are not reported as highway fatalities because they occur on private property.  Back-over avoidance systems are being marketed as “parking aids” using ultrasonic or radar technology to warn drivers as they approach an object.  Initial evaluations indicate that camera-based systems offer the greatest potential, but driver use of these systems is still under evaluation.

One out of every four crashes occurs at highway intersections.  We have the capability to manage the traffic at those intersections by measuring an approaching vehicle’s estimated time of arrival, speed, and range in order to extend the green light to prevent collisions.  Another crash-avoidance technology is the adaptive cruise control system, such as the system available in the Mercedes S class.  This system uses two radar frequencies to keep the car at a safe following distance and can even bring the car to a complete stop. If the car detects conditions for a frontal collision, it not only sounds an alarm but also applies the brakes to stop the car. The system also has “night view assist,” an infrared camera system offering a video dash display of the upcoming road that extends more than 100 feet beyond the low-beam headlights.  Several other manufacturers, Honda, for example, now equip certain models with crash mitigation breaking systems that tighten seat belts and apply brakes before a collision occurs.  With the development of adaptive cruise control with lane monitoring and active steering (or evasive steering) underway, we are approaching the technical feasibility of autopilot systems.


I am confident that highway automation will greatly improve safety, but I am not naïve about what it will take to see these benefits.  We have work to do to ensure that the safety promises of these systems become reality.

System integration, for example, is an important issue.  Different manufacturers make anti-lock brakes, stability control systems, collision avoidance - and these systems must work in concert to avoid a variety of road hazards.  Developers of these technologies must consider how the systems will be used, where displays will be located, how much information is needed, what information has priority, when the systems should be active, and how the systems should function in an emergency.

And in the end, it is the public, and their ability and willingness to make use of these systems, that will determine how effective they will be—and how soon.  I have used many aviation comparisons today, but the distinct differences between drivers and pilots must be factored into the development of these technologies.  Unlike pilots, drivers receive minimal qualification training, no recurrent training, no medical evaluation, and their education and language skills vary widely.  Drivers may be totally inexperienced in their vehicle type, may have conducted no trip planning, and may view driving as secondary to other personal activities in the car. 

Further, many drivers don’t take the time to understand their cars and how their own driving habits may affect their safety.  Let’s face it: Americans by and large don’t even read their owners manuals.  

The bottom line is that, no matter how well crash avoidance and other systems work, they will be more effective once drivers understand how the car and their driving performance can prevent crashes.  We must test and evaluate all of these technologies to in order to get a better understanding of how these systems can affect the likelihood and seriousness of accidents.

The NTSB strives to keep developing technologies on the forefront.  I fully expect that the Safety Board will be an active participant in understanding the implications of these advancing technologies.  But in the end, we recognize that the driver and the pilot must be able to take responsibility.  OUR job is to give them the tools they need to make the most of that responsibility. 



Speeches & Testimony