ashill

As others have posted, it's not clear that it's that simple. As described in a bit of detail in the article TheBOSman linked to, there are deep safety culture issues in the US railroad industry. My main point is that examining both the human failure and the technology failure is crucial; simplifying the issue to a failure to implement technology is dangerous.

(Of course, taking the train, like any public transit, remains significantly safer than driving.)

Java Titan

I'm not professional in railway, but I know when upgrading a track, one of the most important thing is to increase radius for the curve so trains don't have to slow down so much. I can't believe after hundreds millions dollars upgrade project, there is still a curve with 30 miles speed limit. Drop speed from nearby 80 miles/hour to 30 miles/hour, I think that's a horrible design, a death trap on the track.

StayingHomeIsBetter

I am not saying that the solution is technology alone.

Certainly, training and reinforcement of performance expectations/standards have to be a factor. In fact, that is where primary emphasis should be placed.

But, when these can be supplemented by technology, available technology should be applied.

I can train you how to avoid locking up the brakes of your car when you must make an emergency stop. But, do you really want to give up your anti-lock braking system.

As the post after yours (#47) correctly points out, consideration of the human element in this equation cannot be limited to train engineer error. One also needs to consider the actions of the real engineer... the engineer who designed the track. Isn't there some way that a 80 mph to 30 mph transition could have been avoided in the design of the routing?

ashill

I have no doubt that it occurred to the designers that a 30 mph curve after a 79 mph straightaway is a dangerous situation, but I disagree with the implication that avoiding it is the best design choice or that it somehow didn't occur to the designers. In fact, increasing the radius of curvature (to infinite -- ie a straightaway) on many curves was the main purpose and main benefit of this entire project. But you obviously can't eliminate all curves, and the upgraded section has to end somewhere.

This 30 mph curve is over a bridge near the end of the upgraded section of track, so you're suggesting that they should have built a new bridge over I-5. That doesn't come cheap, and it's pretty easy to imagine that building a new bridge would not be the best use of resources. You have to merge into the existing track anyway at some point, so you'll have to get from 79 mph to the 30 mph curves that exist on the rest of the Cascades route. I think it's totally reasonable to use speed limits to decrease the speed. What else should they do? Put in a 50 mph curve between the 79 mph segment and the 30 mph curve that can't practically be removed so the train has to slow down for an earlier curve? Moreover, the whole project was using an existing right-of-way; increasing the radius of a curve requires lots more space.

I think we mostly agree. I'm just cautioning about what I see as an overemphasis on PTC at the expense of other safety concerns.

The track has to get over I-5, and they're using existing right of way, so I doubt it. It is entirely possible to get from 80 mph to 30 mph safely; there are countless places around the world where trains slow from 80 mph to 30 mph. It is, without a doubt, a potentially dangerous situation, so you have to be sure that the slowdown happens with time to spare. That obviously didn't happen here, and I'm sure we'll figure out why sooner or later. But I think it's grossly premature to say that there shouldn't be a 30 mph curve after a 79 mph straightaway given the physical constraints in this (or any other) particular place.

3Cforme

The rush to launch service on a new, faster Amtrak route near Seattle came at a deadly cost: none of the critical speed-control technology that could have prevented a derailment was active before the train set off on its maiden voyage. Work to install the sophisticated, GPS-based technology known as positive train control isn't expected to be completed until next spring on the newly opened 15-mile (24-kilometer) span where the train derailed, according to Sound Transit, the public agency that owns the tracks.

Positive train control was installed on 23 percent of the nation's passenger route miles and 37 percent of freight route miles as of July, the last time the Federal Railroad Administration updated its online tracker for the technology. It is activated on the tracks Amtrak owns along the Northeast Corridor, from Boston to Washington, D.C., and on Amtrak's Michigan line. Many of its locomotives are equipped for positive train control. Throughout the rest of the country, Amtrak operates on track owned by freight carriers and other entities that have made varying progress on installing the technology.

Amtrak didn't wait for system that could've prevented wreck - seattlepi.com

As for eliminating curves that demand slowing a train, that's a right of way issue as much as a financial issue. Americans do not love use of eminent domain for infrastructure projects: not for power lines, not for pipelines, not for trains.

ashill

I believe that the conditions of the Federal grant required WSDOT/Sound Transit/Amtrak to start service on this route in 2017. They weren't allowed to wait until next year, so that's an issue that goes right to Congress (but again, as much for the failure to adequately fund PTC implementation as for the deadline on stimulus projects).

All of this said: they didn't have PTC on the Point Defiance route either. So I don't see why they should have waited until next year just to wait for PTC. Perhaps lower speed limits on the bypass route until PTC was enabled would be appropriate, if it really isn't safe to operate at 79 mph without PTC. There would be time savings even if they operated the straight bypass route (without PTC) at the same speed as they operated the curvy Point Defiance route (also without PTC).

StayingHomeIsBetter

From the cited article:
Wasn't Philadelphia in 2015 a "wake-up call"?

Or did AMTRAK management just go back to sleep in the intervening two years?

You have to wonder whether executive management actually ever listen to their own PR crap.

airsurfer

Indeed PTC should have been installed on the new route BEFORE the inaugural train ride. In Europe on many lines it is installed (it is called ERTMS / ETCS), but train speeds are much higher: even regular intercities (not the much faster ICE and TGV trains) go 160-200km/h (100-125mph).

But, possible stupid question, do train locomotives have black boxes / flight recorders like aircraft, so the last xx minutes before the crash are recorded, including data and cockpit voices of the train driver and rail traffic control ?
That might be a major thing.

Anyway, despite this (and other accidents), train and air travel is MUCH safer than driving the same mileage.

ashill

Yes. The NTSB has already recovered the one from the rear locomotive and released preliminary results. There are also cameras which should show the engineer, though they're in the more-damaged front locomotive and thus more difficult to access.

+1 Important to remember. On average, 100 people die every day in automobile accidents in the US. So the train accident on Monday was only about 3% of the nationwide transportation accidents on that day alone, and this accident was about a year's worth of passenger deaths on passenger trains in the US. It's important to fix the serious issues, but also important that Americans don't overreact by avoiding train travel out of safety concerns.

3Cforme

You need to normalize that by the 3 trillion passenger miles by car vs. a small fraction by train. Train travel is generally safe.

ashill

Of course. And when you do, train travel comes out with roughly half the fatality rate per passenger mile of driving.

flitcraft

I must say I haven't been impressed by the media coverage of this accident. Screaming headlines "Lakewood Mayor Predicted Fatalities on this Line Months Ago" makes it sound like he was predicting this kind of accident, when in reality it was walkers along the tracks and level grade crossings he was worried about. In a haste to get a 'scoop' the fatalities went from three to six to 'multiple, and expected to rise' and finally back to the original three. And the local news kept repeating that the purpose of the track switch was to improve Seattle to Portland times by merely ten minutes, when anyone who rides the Cascades regularly knows that, particularly in the rainy season, the Point Defiance stretch was notorious for causing cancellations and serious delays due to mudslides and such. (There's even a term I've heard for getting bussed from Portland to Seattle due to track issues: 'bustitution'!) And finally, when the last fatality was publicly identified, the main thrust of the story was about his child pornography conviction some years ago. I don't doubt that many accident victims may have had criminal records, but unless the conviction is somehow germane to the accident, this seems like sensationalism for the sake of sensationalism. Not the newspapers' finest hour...

TheBOSman

I'm always concerned about media when they end up covering a subject I happen to have a significant depth and breadth of knowledge on, and I can see how far off they are. Then I get concerned about their accuracy on matters where my personal knowledge base is far shallower.

RatherBeOnATrain

I don't think you appreciate how bad the 'safety culture' at Amtrak has gotten, StayingHomeIsBetter!

In April 2016, there was a fatal Amtrak crash in Chester, Pennsylvania -- two track workers were killed. The NTSB investigation found that the train engineer had marijuana in his body and the track workers who were killed had cocaine and opioids in their bodies... the NTSB then said that all that drug use did not actually have a "direct causal link to this accident". Instead, the NTSB found that the crash was caused by lack of communication between other employees, improper establishment of work zones and pressure from managers to keep trains on time.

Since that crash, Amtrak's board was able to convince Wick Moorman, former CEO of freight railroad Norfolk Southern, to come out of retirement and serve as Amtrak's "transitional CEO". Moorman started on September 1, 2016 (link on amtrak.com). Moorman is working for free -- he had planned to work for $1/year, but that salary ran afoul of his Railroad Retirement benefits.

Moorman then recruited Richard Anderson, former CEO of Delta and Northwest, who started on July 12 of this year. Anderson and Moorman are currently serving as co-CEOs; Moorman will go back to retirement on the 31st of this month.

StayingHomeIsBetter

Please understand, I do not doubt your assertions. I take AMTRAK somewhere in the NE Corridor multiple times per year.

In any organization, the safety culture is shaped from the top down. A bad culture evidences problems at the top.

In this particular case, I found it striking that, just two years after a multi-fatality incident involving a speeding train and a lack of PTC, the AMTRAK CEO would announce that this multi-fatality incident involving a speeding train and a lack of PTC was the "wake-up call" for the railroad.