Chatsworth, Three Years Later

In my mind, PTC in general and in theory is a good thing, but it's not ready for prime time. It's expensive, it only prevents about 3% of accidents, and it's a capacity killer. It is also brand new technology that will fail on a regular, if infrequent basis.

We'll see what happens as we approach the December 2015 deadline, but my gut tells me "time extension".

it's a capacity killer

In what way? World experience for more than 90 years shows that PTC is a requirement to increase capacity and provide fail-safe mechanisms to ensure that capacity increase doesn't come at the expense of lives.

It is also brand new technology

Well, in the world of 19th century No.American freight railway operations and "customized lets re-invent the wheel (cboss) just for America" scenarios it is. Otherwise, it's the standard in modern railway systems worldwide, including the NEC (ACSES) and urban transit systems.

NellieBly wrote:Well, the NTSB report came out last year, and now we know. The engineer was texting his railfan buddies, and wasn't paying attention.

All I know is what I read in rr.nut. I think the posts I read said that someone sent texts to his cell phone. I didn't know that there was proof that he was texting, or at least I don't remember reading that he was texting while the train was moving.
You imply that you read last year's MTSB report. Can you give some details as to why you feel it says he was texting?

kaitoku wrote: In what way? World experience for more than 90 years shows that PTC is a requirement to increase capacity and provide fail-safe mechanisms to ensure that capacity increase doesn't come at the expense of lives.

PTC hasn't existed for 90 years, so I'm not sure what you're talking about here. However, if you're trying to suggest that PTC increases capacity, you're not only wrong, you're laughably wrong. The installation of PTC as a safety enforcement overlay to existing train control systems (which is what is happening) will reduce capacity.

kaitoku wrote:Well, in the world of 19th century No.American freight railway operations and "customized lets re-invent the wheel (cboss) just for America" scenarios it is. Otherwise, it's the standard in modern railway systems worldwide, including the NEC (ACSES) and urban transit systems.

Ummm, no it isn't. PTC is not "standard" in any way. The ACSES installation is proprietary, and in no way can be described as "standard". There are also no "standard" installations of PTC on urban transit systems (whatever you mean by that amorphous term).

Ummm...it would be nice if people posting on this thread knew what they were talking about.

PTC (in the form of ARES) was successfully tested on BN in the mid-1980s. That's 25 years ago now. Railroads have been working on various flavors of the technology ever since. How does that qualify as "untested"?

The system the freight railraods have standardized on (with some minor variations) was developed by Wabtec in the 1990s and has been extensively tested. It takes its "look and feel" from the Rockwell-developed ARES. It will be interoperable. There are some issues having to do with the railroads' choice of radio communication technology, but that's a whole other subject.

ACSES is a variant of the European Train Management System, and is a transponder-based system that is conceptually quite different from Wabtec's.

As for coping with differing train sizes, ex-Con, all I can say is that I've seen functional PTC on everything from commuter trains to 22,000 ton iron ore trains. As for capacity impacts, a properly designed PTC system should increase, rather than decrease, capacity. Note "properly designed" is the qualifier.

As for Sanchez, read the NTSB report. There is stuff that isn't in the report having to do with Mr. Sanchez' state of mind and some of his peculiar habits, but you can find that on the Web yourself. Suffice it to say that he may have been trying to commit suicide. Shame he took 25 people with him.

NellieBly wrote:Ummm...it would be nice if people posting on this thread knew what they were talking about.

Hope you're not referring to me. I have 25+ years in passenger rail, and have worked on both SCRRA's and PCJPB's PTC programs.

NellieBly wrote:PTC (in the form of ARES) was successfully tested on BN in the mid-1980s. That's 25 years ago now. Railroads have been working on various flavors of the technology ever since. How does that qualify as "untested"?

If the technology were ready for prime time, and offered the benefits that you've often touted on this and other threads, it would have been deployed nationwide by now. It would not have taken a literal act of Congress to force the railroads to develop and deploy it if it were so beneficial to capacity and operations (which affect the bottom line, with which Class 1 RRs are mostly concerned). It is also a virtually new technology for just about all North American passenger railroads save a select few. Even in the few places where it's been deployed, the implementation has been either painful, slow, or both. In the case of NJ TRANSIT, the system was designed, installed, tested, and then promptly switched off when it didn't work as advertised. Amtrak has been struggling with ITCS on the Michigan Line for over a decade (since 1995, actually), and it still has yet to reach full maturity.

NellieBly wrote:The system the freight railraods have standardized on (with some minor variations) was developed by Wabtec in the 1990s and has been extensively tested. It takes its "look and feel" from the Rockwell-developed ARES. It will be interoperable. There are some issues having to do with the railroads' choice of radio communication technology, but that's a whole other subject.

Well, it might be a "whole other subject" to you, but I'd maintain that if it's necessary to the operation of the system, you can't just set it aside because it's inconvenient to your argument. Issues around radio spectrum, data radio loads, system maintenance for the thousands of new WIUs that will have to be installed in places both easy and inhospitable, and the overall reliability of the system (which is not 100% - in fact, each of the subsystems that make up PTC will be less than 100% reliability, which means that in the aggregate, the system will be less reliable that any of the individual parts) are not "whole other subjects". They're tough, integral and essential questions which have yet to be answered, with only 4 years left to government-mandated implementation.

NellieBly wrote:ACSES is a variant of the European Train Management System, and is a transponder-based system that is conceptually quite different from Wabtec's.

First thing you've said that indicates that you have more than a passing familiarity with PTC.

NellieBly wrote:As for coping with differing train sizes, ex-Con, all I can say is that I've seen functional PTC on everything from commuter trains to 22,000 ton iron ore trains. As for capacity impacts, a properly designed PTC system should increase, rather than decrease, capacity. Note "properly designed" is the qualifier.

Well, everything is going to be OK! Nellie Bly said so. Snark off for now, but you're totally wrong about capacity. PTC, as a safety overlay (which is how it's being designed and deployed in the US), will decrease capacity. The only way to increase capacity using PTC is to trash existing train control systems and implement a true "moving block"-type system. That is NOT what is being discussed right now for the overwhelming majority of US PTC deployments. PTC is already a wildly expensive unfunded mandate; railroads are not going to throw away existing train control systems to install a technology that I still maintain is untested in production, day to day operations under a variety of operating conditions. They're going to do what they need to do to comply with the law. That means installing PTC as a safety overlay, which will decrease capacity.

Oh yea, how could I have forgotten about the absolute disaster that the installation of PTC on the Pascack Valley Line on NJT.

Im so glad that I never had to work over there, but I remember learning about it in ChooChooU. It sounded so simple and easy and great....reality...well not so much

HAHAHHA what, 100 Million later and the plug was pulled after 10 years or so? I thought that was a USS system they had installed there? That the same was the Wabtec system or similar?

As a side note, all the instructions and rules are still in the rule books regarding PTC (or as NJT called it SES) they are just "temporarily suspended" I believe is how they put it.

From fellow engineers who run the PVL, they said overall speeds significantly increased (actual running times, not timetable or scheduling) after it was shut off, because the braking curve no longer had to be set for "worse case scenario" which quite often (never) was not the case, and the engineers only had to comply with the cab signal changes - cab signals no waysides was installed - Norac rule 562. And passenger, especially short commuter trains with EP brakes are considered racecars compared to a heavy freight train braking in the rain or bad rail conditions - aka worse case scenario....

Jtgshu wrote:From fellow engineers who run the PVL, they said overall speeds significantly increased (actual running times, not timetable or scheduling) after it was shut off, because the braking curve no longer had to be set for "worse case scenario" which quite often (never) was not the case, and the engineers only had to comply with the cab signal changes - cab signals no waysides was installed - Norac rule 562. And passenger, especially short commuter trains with EP brakes are considered racecars compared to a heavy freight train braking in the rain or bad rail conditions - aka worse case scenario....

I remember the PVL installation....what a nightmare!

Jtgshu wrote:From fellow engineers who run the PVL, they said overall speeds significantly increased (actual running times, not timetable or scheduling) after it was shut off, because the braking curve no longer had to be set for "worse case scenario" which quite often (never) was not the case, and the engineers only had to comply with the cab signal changes - cab signals no waysides was installed - Norac rule 562. And passenger, especially short commuter trains with EP brakes are considered racecars compared to a heavy freight train braking in the rain or bad rail conditions - aka worse case scenario....

Jt, you're hitting it right on the head from the 'real world' perspective of a passenger train engineer, as usual. Thanks for the input!

The problem with safety overlay PTC (which is what is being designed for US railroads) is that it must assume worst case braking, track condition, and engineer experience in enforcing speed restrictions.

It works like this. If you're at track speed, say 70 miles per hour, and approaching a 30 mile per hour speed restriction (maybe it's a curve, say), under current wayside only, wayside with cab, or cab signal only train control systems, the judgment call about where to begin braking for the speed restriction is usually left up to the engineer (with some exceptions, such as the Elizabeth S-curve on the NEC, but those exceptions are few and far between). You get, with an experienced engineer, a fairly optimal braking curve, where the train doesn't begin slowing until it pretty much has to, to be down to 30 mph just in advance of the curve.

Now, add PTC. PTC, of necessity, must assume that an inexperienced engineer is at the controls, that the rails are coated with a slick slime that reduces adhesion significantly, and that the brakes are not working at 100% efficiency when it calculates where braking needs to start to get the train down from 70 to 30. That's PTC's enforcement point, where the onboard system will take over and slow the train if the engineer fails to do so. Because you've assumed worst case conditions, you're already backing off from the "engineer's best judgment" braking point by a fairly significant distance. Next, you have to consider the PTC warning curve, which becomes active *before* the PTC enforcement curve. This is the point where the PTC system alarms to let the engineer know that he/she is about to get a forced brake application (AKA a "penalty application") if he/she doesn't initiate braking. This can be set from 10 to 20 seconds or so before the automatic system cuts in. Finally, you have "pre-action". This is the phenomenon, in human factors, wherein the engineer, once he or she has operated in the PTC territory several times, begins to 'know' where the alertor will sound, without even thinking much about it. He or she will begin to get on the brake a little early to avoid the annoying alarm. So, now we've not only backed off from the "engineer's judgment" optimal braking point some distance due to our assumption of worst case conditions; we back off further to allow for the warning alarm; then we back off further because the engineer's reaction to learning where the warning alarm occurs each time will be to begin braking before the alarm goes off.

The result? You slow the train down a surprisingly and distressingly long distance from the "optimal" point where a good engineer, using his/her best professional judgment, would start braking. Thus, you lose capacity. And don't get me started on the problem of absolute signals located at the leaving ends of platforms under PTC...

Absolute signals at the end of platforms - yea I was wondering how they were gonna handle that.....especially in places where its common to come up and "kiss" the signal - i.e. pull very tight up to a signal. Crawl in at like 1mph through the entire station because PTC won't let you go any faster because you are coming up to a stop signal?

I hate to say it, but there is no denying PTC was a knee jerk reaction to Chatsworth and the entire industry is suffering because of one person's failure. Instead of better training and better hiring and better management, they are going this route and going to try to fit the entire industry and operation of it, into a neat little box and its not going to work. IMO if anything was going to be implemented nationwide, why not cab signals (ATC/ATS) - Of course, CSS would not have kept him from running the actual stop signal, but he would have had a warning when passing possibly 2 signals out (say an approach medium) - and then the distant signal (approach). And a further cab signal reduction to restricted speed halfway inbetween the distant and home signal would have been yet another "reminder" to the engineer to slow down and that he has a bad signal. Also the actual signal in the cab is yet another reminder. And if he still went by the stop signal, the fastest he could go would be 20mph. Sure a lot of damage can still be done, but that would have been THREE actual required speed reductions - 1) MAS to 45 for the approach medium change - 2) 45mph to 30 for the approach change, and then 3) 30 to 20 for the restricting change half way through the last block before the stop signal. Each one being a reminder to get the engineer's head out of his a$$. Upon leaving the station, he might only be able to get up to 20mph anyway (if the cabs already dropped halfway through) or when they would drop to 20 he would have to put brakes on or else get a penalty app. If that doesn't snap him out of the fog his brain would be in, nothing will

CCS (ATC/ATS) Simple, safe and proven. Not this nonsense.

Jtgshu wrote:Absolute signals at the end of platforms - yea I was wondering how they were gonna handle that.....especially in places where its common to come up and "kiss" the signal - i.e. pull very tight up to a signal. Crawl in at like 1mph through the entire station because PTC won't let you go any faster because you are coming up to a stop signal?

There's some remediation that can be done with transponders, but yeah - that's pretty much it. Picture PTC forcing you to stop halfway in, halfway out of the station (after crawling for a half mile in advance). Then, operating some type of conditional release to allow you to gingerly "kiss" the signal.

Jtgshu wrote:I hate to say it, but there is no denying PTC was a knee jerk reaction to Chatsworth and the entire industry is suffering because of one person's failure. Instead of better training and better hiring and better management, they are going this route and going to try to fit the entire industry and operation of it, into a neat little box and its not going to work. IMO if anything was going to be implemented nationwide, why not cab signals (ATC/ATS) - Of course, CSS would not have kept him from running the actual stop signal, but he would have had a warning when passing possibly 2 signals out (say an approach medium) - and then the distant signal (approach). And a further cab signal reduction to restricted speed halfway inbetween the distant and home signal would have been yet another "reminder" to the engineer to slow down and that he has a bad signal. Also the actual signal in the cab is yet another reminder. And if he still went by the stop signal, the fastest he could go would be 20mph. Sure a lot of damage can still be done, but that would have been THREE actual required speed reductions - 1) MAS to 45 for the approach medium change - 2) 45mph to 30 for the approach change, and then 3) 30 to 20 for the restricting change half way through the last block before the stop signal. Each one being a reminder to get the engineer's head out of his a$$. Upon leaving the station, he might only be able to get up to 20mph anyway (if the cabs already dropped halfway through) or when they would drop to 20 he would have to put brakes on or else get a penalty app. If that doesn't snap him out of the fog his brain would be in, nothing will. CCS (ATC/ATS) Simple, safe and proven. Not this nonsense.

You're speaking my language! For the 3% of actual accidents that PTC *might* prevent, the cost (in terms of $, time, and capacity) is way out of proportion, especially given that cab signals in most of those cases would suffice (as you've pointed out). Given time, the systems that the Class Is have been developing could turn into something much better - still expensive, but offering more than just a clumsy, inaccurate, capacity-killing way to enforce a red signal. But because they've now been forced to deploy PTC under penalty of law well before it's ready, we all get to pay the price.

Patrick Boylan wrote:

NellieBly wrote:Well, the NTSB report came out last year, and now we know. The engineer was texting his railfan buddies, and wasn't paying attention.

You imply that you read last year's MTSB report. Can you give some details as to why you feel it says he was texting?

NellieBly wrote: As for Sanchez, read the NTSB report. There is stuff that isn't in the report having to do with Mr. Sanchez' state of mind and some of his peculiar habits, but you can find that on the Web yourself. Suffice it to say that he may have been trying to commit suicide. Shame he took 25 people with him.

I can't tell if you meant to try to answer my question or not. Can you give some details as to why you feel the NTSB report says he was texting?
Now you're also bringing up stuff that's not in the report, and giving very vague reference to it being on the Web. I hope you understand that we should take what's on the web with a grain of salt, but if he was committing suicide then I'm not sure why you first said that you concluded he was texting, implying that the texting caused the accident.

Patrick Boylan wrote:...but if he was committing suicide then I'm not sure why you first said that you concluded he was texting, implying that the texting caused the accident.

Suicide letter for the 21st Century ?

If there was a Suicide letter for the 21st Century why have none of the newspaper reports that mention phone records that say he got texts also say that he sent texts that pertain to suicide?

Patrick - The world isn't as cut and dry as you believe. Often, the truth is the most far fetched, or Occam's Razor be damned, the most complicated and convuluted path possible. It's odd that you poo-poo the net but put so much faith into the newspapers. If you had had experience with the media, you would realize that they don't report news. They perform. I can tell you honestly that I have been at what would be considered many "major news events" and (at first) was surprised at how different the news story was from my first-hand observation. Many of them I had to check to see if they were talking about the same event that I was at. For this reason alone, I thought that I had made quite a coup recently when the pests from the Boston Globe stopped calling me begging me for a subscription. They had heard why I wasn't subscribing and the sound of my receiver hitting the base of my telephone so many times that they finally realized that it was like swimming against the tide.