Honorable Robert L. Sumwalt

Keynote Remarks of
Robert L. Sumwalt
Board Member
National Transportation Safety Board
Before
Human Factors Workshop - Safety Data
Transportation Research Board Annual Meeting
Washington, DC
January 13, 2013


Good morning.

The Transportation Research Board (TRB) asked me if I would mind being involved with two workshops today. I eagerly agreed. What they didn't tell was that both workshops are at the same time. Steve Popkin said it's not a problem - they would just split me in half. My wife objected to that - she said I was short enough as it was. So, as a compromise, I will be here to charge the group with a few challenges and then head to a workshop on automation, and then come back here in the afternoon to hear what you are reporting out of your breakout groups.

Yesterday afternoon I returned from New York City where NTSB is investigating a ferry boat accident that happened Wednesday morning. As the vessel Seastreak Wall Street was approaching Pier 11 in lower Manhattan, the captain discovered that he was unable to put the vessel into reverse to slow for docking. The vessel slammed into the pier, throwing passengers into walls and each other, and onto the floor. There were scores of injuries.

The vessel did not have, nor was it required to have, a voyage data recorder. The crew was able to shed light on their perspectives, but we need to go beyond learning what happened and determine why it happened.

While I'm comfortable that our investigators will be able to determine the cause of the accident, when we don't have data, it adds challenges. And, we are asking ourselves one of the very questions that is the theme of this workshop: where is the data. We have already identified data from sources that are not necessary intended for accident investigation. For example, on Friday we identified that there were six closed circuit TV cameras on the vessel that captured the hours of video. And, we also learned from the engine manufacturer that the engine control unit captures data that may shed light on why engines shut down during the accident sequence.

In Friday's press conference, I asked that if the public had cell phone video or still shots, please contact us. I said "There is so little data, that whatever is available, we must capture." So, yes, the question of "where is the data" is an important one.

I think we all know that in our various professions, the problem isn't a lack of data like we are initially experiencing in the NY accident, but rather, it's the quality of data that is lacking. When I worked for an airline working in their Flight Operational Quality Assurance (FOQA) program, we had globs and globs of data. The challenge was figuring out how to turn all of those data into meaningful information.

And, that leads to one of the other themes of this workshop: how do we use the data.

We've got some of the brightest minds in the transportation industry right here in this room. So, I'd like to issue my first challenge to you - brainstorm the answer to this question: How do you capture the data that you need, and even more importantly, what do you do with it once you have it.

As French philosopher Jean Baudrillard wrote, "Information can tell us everything. It has all the answers. But there are answers to questions we have not asked, and which doubtless don't even arise."

It's so true that if we don't ask the right questions, we'll never find the right answers. As we have all heard, "you don't know what you don't know."

In June 2009, a Washington, DC subway train rammed into the back of a standing subway train. The accident led to nine fatalities and over 50 people were taken to hospitals. NTSB determined the accident was partly caused by a faulty track circuit that caused the standing train not to be detected by the automatic train control system. This resulted in the following train receiving full speed commands up until the time that the train's operator realized the problem and placed her train into emergency braking. The speed at the collision was between 44 and 49 mph.

The NTSB discovered the safety metrics the transit authority's board of directors was looking at were things like escalator injuries; slips, trips , and falls; crimes in of the subway stations and parking lots. Each of these things is important, but they do nothing to predict the type of accident that happened that day in June 2009.

If you are going to collect data, it needs to be the right data. Not that those other data points weren't important, but the purpose of safety data is to identify potential problems so the problems can be corrected before an accident or serious incident.

Last June I gave a presentation on this accident to the California Public Utilities Commission. In the audience were ten of the 14 board members of Pacific Gas and Electric, along with their senior management. You will recall that PG&E had the San Bruno, CA pipeline explosion that destroyed an entire neighborhood and claimed eight lives.

When it came to this point about collecting the right data, the PG&E CEO asked, "How you make sure you are measuring the right things?"

It's a great question and my honest answer is that I don't know.

So, my second challenge to you is closely linked to the first: Deliberate the question: How do you make sure that you are collecting and analyzing data on the right things - the things that can help you predict and avoid the next accident or serious injury?

Even if you can't achieve consensus on the answers to these tough questions, I think the discussions with be very fruitful and enlightening. Hopefully, they will serve to provoke follow-on discussions.

Alan Mulallay worked at Boeing for 37 years, leading development of the 777, and later President and CEO of Boeing Commercial Airplanes. He left Boeing in 2006 to go to another sector of the transportation industry, moving to Ford Motor Company as President and CEO.

A saying of his that I have enjoyed for many years is this: He says, "The data will set you free."

Think about that. How does data set you free?

Late last night when I retired for bed, I read the weekend edition of the Wall Street Journal. Alan Mulallay was mentioned in an article about CEOs. The article said one of the learning points he has for CEOs is "You can learn from everybody." That's true and that is my final charge to you. In this room, we have a Ph.D. sitting right next to a student. So, leave all of these credentials behind and see what you can learn from the person sitting next to you. You can learn from everybody.

Thank you very much and good luck with your deliberations.