Last week, a West Virginia–bound Amtrak train carrying dozens of Republican members of Congress plowed through a garbage truck on the tracks. The collision treated riders to a grisly scene, hurting six people and resulting in the death of one of the truck's passengers, 28-year-old Christopher Foley.
Four days later, an Amtrak train collided with a parked freight train in South Carolina, killing the Amtrak's engineer and conductor. Then, on Tuesday, two cars on an Amtrak Acela line bound for New York separated in Maryland while the train was moving at 125 MPH. While none of the 52 passengers were hurt, it marked the fourth high-profile incident for Amtrak since the deadly derailment of one of its trains in Washington State in December.
The string of accidents has caused at least six deaths, dozens of injuries, and raised questions about the safety of train travel in America at a time when getting on a flight is safer than ever. So are we just experiencing a blip of highly visible incidents, reminding commuters of the inevitable threat of human error? Or is there a broader safety problem affecting America's quasi-private train carrier (Amtrak gets some public money but is for-profit) today? After a summer when New York's subway system set a new bar for dysfunctional infrastructure, and ahead of the release of Donald Trump's long-awaited infrastructure plan, VICE spoke with Dr. Allan Zarembski, a railway civil engineering and safety expert from the University of Delaware, to find out.
VICE: Is there a fundamental (and new) safety issue with train transportation in the US right now?
Dr. Allan Zarembski: Statistically, there has not been a significant upturn in safety issues in the railway industry. If you look at the ten-year trends, the recent trends, the long-term trends, safety is on a consistent improvement basis. Do we occasionally have a hiccup where you have a small period of time where you have an upsurge? In some cases, yes. But there doesn't seem to be something fundamental in the industry.
OK, so how do you explain this recent upsurge?
If we look at the Amtrak cases, the three [most] recent fatalities, all three have significantly different causes—two of which were arguably not Amtrak's fault. The congressional train that hit the dump truck, all evidence seems to suggest that that was a truck running around the crossing gate. So that's not Amtrak's fault.
The [South Carolina] crash that happened this weekend, the preliminary indications suggest that a switch was thrown into the wrong position, and the Amtrak train was diverted to the wrong track [which was owned by another company, CSX] with an existing train [on it]. If it was a manual switch, it may have been a CSX worker who accidentally left the switch in the wrong place. Unfortunately, that is the cause of a small but real number of accidents that happen each year.
So the only one that you can argue was in Amtrak's purview was the one in Washington State where the train engineer did not realize where he was and did not slow the train down going around the sharp curve.
If I look at the statistics, I don't see this major anomaly from a safety point of view at Amtrak.
After the Washington crash, investigators suggested "positive train control"—a wireless system that uses sensors along the tracks to transmit data to the engineer and can take control if need be—may have prevented the crash. Others, like National Transportation Safety Board chairman Robert Sumwalt, have suggested it could have prevented the crash in South Carolina. Isn't that the obvious fix here?
Positive train control (PTC) is designed to prevent overspeed accidents such as the Washington State one and the one in Philadelphia a couple of years ago. It is also designed to prevent many of the collision-type of accidents, particularly ones where drivers get distracted or things like that. Positive train control has absolutely no effect whatsoever on the vast majority of track-caused derailments or equipment-caused derailments, which represent between 50 and 60 percent of all derailments.
Is positive train control a means of preventing certain classes of accidents? Absolutely. Is positive train control going to prevent all types of accidents? Absolutely not. The Washington State accident, which was going overspeed around the curve—yes, positive train control would have prevented it. The West Virginia gate-crossing accident, no. The CSX accident in South Carolina—maybe.
Why isn't positive train control already implemented along every train line in the US?
Amtrak has implemented positive train control along its the Northeast Corridor [between Washington, DC, and Boston]. As far as the rest the country, there are 200,000 miles of track in the United States, and somewhere between 50,000-plus are mandated to have positive train control in place by December 31, 2018.
Originally it was supposed to be in place earlier. There was a congressionally approved delay in implementation... There was a problem getting proper bandwidth access from the FCC. Another major source of the delay was this was a self-funded mandate. In fact, the government said "implement PTC at your expense," and the expense is $10 billion.
It doesn't sound like positive train control would have prevented the incident that happened Tuesday, when two cars on an Acela line separated on a trip to New York. Is that an indication Amtrak's trains are simply outdated? Do we need better inspections?
The Acela equipment is getting older—I believe it's 20 years old—but in the railroad world, 20-year-old equipment is not particularly ancient. So that in and of itself is not the only issue. The appearance of this incident suggests that a more aggressive maintenance inspection program might have found this, but I can't say that for a fact.
This might also be an area where some additional onboard sensors may prove useful in the future to detect that. But again, while there was a separation here, there wasn't an accident. This was a case where nobody got hurt, nobody got injured. But I'm absolutely positive Amtrak's maintenance of equipment people will be aggressively looking at this to see what the causes are—Is there a way to improve inspection? Is there something else that can be done to to find it?
Joe Biden was a huge proponent of Amtrak, while President Trump has proposed reducing subsidies, though he is also planning to unveil a $1.5 trillion infrastructure plan next week. What’s the argument against giving more funding to Amtrak?
The question becomes what is a subsidy for. There are two sets of costs that a railroad incurs: operating costs and capital costs. An operating cost is what I need to sort of run my system. Capital costs are the big dollar, big-ticket items. Amtrak does a good job covering more than 90 percent of its operating costs. So any capital money has to come from the government. So when you talk about cutting money to Amtrak, what you're mostly likely talking about is cutting the capital budgets.
For example, positive train control implementation would have been a capital upgrade. [Amtrak] would have to use capital money to do that.
So, if PTC can help stop some of these incidents—and the money is not necessarily happening anytime soon—what else is Amtrak in a position to do here?
The railroad industry is very actively looking at safety. Not only the railway industry, but the government and the Federal Railroad Administration (FRA). Safety has been a very strong focus for many, many, many years. There are active research programs and lots of technology that's being developed.
I'm going to say PTC will address some of the very high-profile accidents, like the overspeed [accidents] and a large percentage of the collisions. Overspeed [accidents] are rare. The problem with this is that they're very high-profile, high-visibility, and basically it hits every newspaper, radio station, and TV in the country. Especially when the pictures are as dramatic as the ones from Washington State. So you get a disproportionate amount of press associated with those kinds of accidents.
This interview has been lightly edited for length and clarity.
Follow Lauren Messman on Twitter.