Rail Share Conference

​Remarks as prepared for delivery

Thank you, Amit, for that warm introduction and for inviting me here today. 

As I look around this room, I see a lot of people I’ve known for years, and I’m reminded of all our great work together to improve safety. 

We are partners in safety. We don’t always agree; we shouldn’t. Open communication, collaboration, and a commitment to work together and do what’s right for safety even when it’s hard, that’s how we prevent future tragedies.

So let me start with some “thank yous.”

Amit, thank you so much for your safety leadership. While the FRA and NTSB have long been close safety partners, I’m incredibly grateful that the relationship between our agencies has only gotten stronger these last couple of years. 

I’m also grateful that your team brought FAA’s InfoShare to rail. This past summer, we closed our training center in Ashburn, Virginia. Not the work, just the building. 

At the entrance of the Training Center was an inscription on glass that we brought back to headquarters. It reads, “From tragedy we draw knowledge to improve the safety of us all.” 

No one, not a single one of us, or the organizations we represent, ever wants an accident to occur. But when it does, we need to learn from it. 

Amazon’s chief technology officer puts it this way: “Everything fails, all the time.” We know this is true. Wheel bearings fail. Technology fails. Even our health can fail. 

We need to share our collective failures and crises with each other, as well as our successes. 

And then the important work begins. 

We need to adjust and unlearn. I recently read an article called, “Finding renewal in the midst of disaster: The case of the deepwater horizon spill.” The article talks about how it’s important for an organization to be willing to unlearn and get rid of outdated and unsafe practices to overcome crises. 

​That’s part of what happens at FAA’s InfoShare. It’s a time to share what we’re seeing in safety, and a time to learn and unlearn from tragedy and success. And that’s what I hope we do here.

Look, no one of us can do it alone. And by “it” I mean zero on our nation’s rail system. Zero injuries, zero fatalities, zero period. 

But we all know that’s not where we are today. I looked at FRA’s database and, over the past decade, the rate of total accidents and incidents per million train miles is up. 

The rate of train accidents not at grade crossings is up — the highest it’s been over the last decade, even though total train miles are down. 

Train accidents on main line track have decreased but in yards have increased, which I’m increasingly concerned about. And we all know that trespasser and grade crossing incidents are up. 

But what caught my eye were the train accidents caused by human factors vs. track, equipment, and signal systems: consistently about 40% of all accidents are supposedly “caused” by human error annually. 

The problem? There isn’t just one cause of an accident. There are a lot of factors that contribute to errors and, too often, to tragedy. 

That’s what I want to focus on today: human error. And I’m going to start with a question.

How many people here have heard the following statistic: 94% of crashes on our roads are due to human error. 

I’m not surprised; until recently, that statistic was even on NHTSA’s website. But it’s not true.

The stat gained a life of its own. It was used out of context. And it was removed from the original disclaimer, which said human error is, and I quote, “the last failure in the causal chain of events leading up to the crash.”

We’re glad the statistic was removed from NHTSA’s website. The causes of any given crash are complex and interdependent.

Here’s an example: last January, the driver of a Dodge Challenger…a vehicle that’s marketed as “street-legal, track-ready,” by the way…approaches an intersection in North Las Vegas at 103 mph — nearly three times the posted speed limit.

The Dodge driver…who we determined was impaired at the time… ran a stop sign, entered the intersection on a red traffic signal, and struck the right side of a minivan. Four other vehicles were involved in the crash.

The results were catastrophic: both people in the Dodge and all seven occupants of the minivan died; nine deaths in total. Two others were injured. 

The seven people in the minivan who lost their lives were all members of the same family heading to their Saturday family dinner. Four of the victims were children, the youngest of whom was just 5 years old. 

Common wisdom would dictate that driver error is the cause of this tragedy. And it absolutely ​played a role.

But you know what else played a role?

The failure of regulators to require speed governors in all new vehicles, which we’ve been asking for since 2012, which other developed nations already require. 

What about the impact of marketing by automakers that emphasizes risky behavior, including speeding? 

And then there’s the State of Nevada, which failed to hold the driver accountable for his long history as a repeat offender — despite being the subject of several enforcement actions that were negotiated down to parking tickets.

We would’ve missed these safety gaps if we only looked at the driver’s behavior. 

And then there’s the tragedy that occurred about 100 miles away from here in Mendon, Missouri, when an Amtrak train collided with a dump truck at a passive grade crossing. Three passengers and the truck driver were killed; 146 others were injured.

Our investigation could’ve ended by concluding the truck driver shouldn’t have gone on the tracks. But we also examined the design of the grade crossing. 

We found the road leading to the crossing is too steep. Our investigators measured it. It’s 13-times steeper than the recommended AASHTO standard. 

The truck driver, who was hauling a heavy load and operating a manual transmission, couldn’t make it through the crossing if he came to a complete stop, as required by Missouri law. 

Instead, he was forced to stop where the road was level — 152 feet away. That’s three-times farther than the maximum distance allowed under Missouri law. 

Our investigation also examined the angle of the crossing, which was 30° sharper than the recommended limit.​

While on scene, we learned that the community knew the grade crossing was dangerous; they had been sounding the alarm for years. 

One of those people was Mike Spencer, a local farmer who I was honored to meet. He was once almost hit by a train at the same crossing. He posted about his concerns on social media. And he advocated for years to make it safer. 

Mike’s story moved me deeply. In fact, it inspired me to call a meeting on scene. 

I convened all the parties, asked all the decision makers to sit down and listen to each other about the grade crossing. To their everlasting credit, everyone in that room started from a place of “yes” when the opportunity for safety change presented itself. 

And change is happening. A year after the Mendon accident, Amit and I had the honor of joining Governor Parson as he announced a $50 million investment in rail safety. And Missouri DOT Director McKenna unveiled a plan to improve the state’s 47 passive grade crossings. 

Even more recently, we have an open investigation into the death of a conductor trainee injured during a shoving movement at a rail yard in Cumberland, Maryland, this summer.

Preliminary information indicates the conductor, who was directing the shoving movement, received fatal crushing injuries between the rail car he was riding and three locomotives parked on an adjacent track. We found there was just 7 inches of clearance between the train and the standing locomotives.

As a result of this accident, the Switching Operations Fatalities Analysis Working Group quickly issued an alert advising railroad employees to remain vigilant when mentoring inexperienced employees and to conduct job briefings whenever a job changes. 

A few days later, FRA issued a safety bulletin, which says the following: 

​It is important to note that the conductor had less than a year of experience and never made this switch prior to the accident. The conductor trainee was released from classroom training approximately two weeks prior to this incident.

I applaud that statement because it rightly points to system failures, not individual failures. 

It’s so easy to blame “human error” in all modes of transportation. 

But we need to look beyond human error because everyone shares responsibility to prevent accidents — and ensure our nation’s rail system and its workforce and passengers are safe. Everyone. 

Aviation has long understood that there is always, always going to be human error. We can’t eliminate it. Ever. 

What we can do…what we MUST do…is ensure those mistakes never, ever cost someone their life or result in serious injury. 

There’s a professor and aviation expert at MIT, Nancy Leveson who said: “Human error is a symptom of a system that needs to be redesigned.” 

Let me repeat that: Human error is a symptom of a system that needs to be redesigned.

She goes on to say we must look at the entire system in which people work — not just the man, but the machine and the environment in which they’re operating.

Don’t get me wrong; human error does exist. There is always, always going to be human error. We all make mistakes…errors in judgement…have biases that influence our decision making. 

In November 2021, I was set to hold my first Board Meeting as Chair of the NTSB. I was driving into the office. I decided not to take the train because I thought I might miss the last one home. There was massive traffic on I-95, so I got off at an exit. 

I stopped behind a long line of cars — about half a mile — at a light. I looked in my rearview mirror and saw an SUV heading towards me. It wasn’t slowing down, and I couldn’t move. There was traffic on my right and a lane divider on my left, so I just braced. He rear-ended me. I was sandwiched between him and the car in front of me. 

I thought: I’m not going to walk away from this. 

Luckily, I was wrong. I walked away with one bruise on my arm and a concussion. 

The police officer on scene told me the driver never slowed down; he was going about 45 mph when he hit me. He was cited for reckless driving.

But the officer told me something even more interesting. He said they respond to crashes at that location almost weekly. 

That is a system failure — one that raises important questions, like: 

  • ​​Was the speed limit of 45 mph too high?
  • Was the road designed to slow traffic going down a steep hill into a major intersection? 
  • Had the county looked at the data in that stretch of road and done a road safety assessment to see what safety mitigations needed to be implemented?
  • Why didn’t the automatic emergency braking system in the Toyota RAV4 kick in? Turns out there are no uniform standards on that so some vehicles may not activate their automatic emergency braking at 45 mph.

Remember: Human error is a symptom of a system that needs to be redesigned. 

Accidents are complex, which is why we seek out HOW and WHY the human error occurred. THAT is the NTSB’s mission. THAT is how we get to zero. 

Events like this — and the steadfast leadership of everyone in this room, and of those that came before us — will help us get there faster. ​

On behalf of everyone at the NTSB, thank you.

                                                  ​


​​​​​​ ​
​
​