Somebody asked me my opinion, and you know how dangerous that can be...
After FRA announced its proposed rule-making on mandatory train crew size; after BNSF and SMART announced a tentative agreement on single person operating crews for certain trains on certain routes; after the rank and file of SMART emphatically rejected the tentative agreement arguing that single person crews represent a risk to safe train operations, somebody asked me if I thought that the threat to safety was the biggest obstacle to single person crews.
I don't. I think there are arguments to be made for single person crews; and arguments to be made against single person crews but I don't think safety is one of those arguments, on either side of the question.
First and foremost, there is no data indicating that single person crews are any less safe than multiple person crews when it comes to actual train movements. Passenger and commuter rail service has operated for decades with a single person in the operating cab with greater operating safety than mutiple person in the cab freight service.
Of course, the two services differ. Passenger runs are scheduled; train consists are fixed, more or less, and in-train forces are much less than those forces in freight trains; passenger runs also have additional crew members to handle unusual circumstances--like making a reverse move; flagging a crossing; or setting out a defective car. But those are the demands of the service, not safe train operation.
Way back in the day when advanced train control systems, with positive stop and overspeed prevention capabilities were being proposed, designed and debated, there was considerable concern that the safety capsule was a poison pill for crew consist agreements. I certainly don't think that concern was enough to deter the railroads from pursuing and installing PTC. It was the cost of installation, expressed as the cost-benefit differential that was the obstacle.
However, BNSF, being forward looking, proposed a reduction in crew consist size based on the additional safety PTC will provide to the operating environment. Certain trains operating in PTC territory were to be eligible for single person operation.
Sure enough, if human error is the largest single cause of reportable main line train accidents, and if PTC is capable of eliminating 99.99% of those human error causes, then BNSF and the other railroads might have an argument based on improved safety of the operation.
The thing is, the way FRA has executed the law in its regulatory requirements, the interpretations FRA has attached to requirements of the law, the exemptions given for non-PTC equipped train operation in PTC territory the probability of eliminating the 99.99% of the 38% of train accident causes isn't clearly that probable anymore.
FRA can, and has, interpreted the law requiring PTC to eliminate train-to-train collision as requiring that PTC restrict the speed of train-to-train collision to no more than 20 mph.
Also, FRA has authorized unlimited movements, each of no more than 20 miles, of "yard transfer movements" hauled by non-PTC equipped locomotives in PTC territory.
FRA agreed to allow dispatchment of a train into PTC territory whene the locomotive(s) on board computer(s) had failed the initialization process during which the locomotive receives the limits to its movement authority; the temporary and permanent speed restrictions; the location of work zones, etc.
FRA took some of these actions in response to the AAR's petition; finding merit in AAR's reasoning that requiring such transfer movements to comply with provision of the RSIA 2008 would present the Class 1s with an unreasonable burden-- namely a burden that far outweighed the possible benefits of requiring full compliance.
Well, good for AAR. Maybe. But you don't get to get it any way you want it. You don't get to create exceptions to and in PTC operation, and then argue that PTC makes train operations "fail-safe" to the point where a second crew member is redundant for safety. Actually you do get to argue that. It's a free country. You can argue anything you want. But don't expect everyone to nod their heads and say "All right, OK." And put his or her initials on the bottom line of that order.
I think that the arguments about single person crew operation need to take into account the actual rail network in the US and the actual operation of trains. Meaning what? Meaning that almost half the freight network in the United States is "dark territory"-- without automatic signals, where train movements are governed by voice radio communication between trains and the trains' controllers, the dispatchers. Meaning that movements are authorized by voice and must be copied, copied and repeated, repeated and approved, and that the crew members who have to abide by these authorities must know their content.
In both dark and signal territory, many switches to sidings, yards, and industrial tracks are hand-operated switches. Some of these switches can be operated from the cab of an approaching locomotive via radio transmission. Some can't. In either case, operation of the switch means restoring it for "normal" movement immediately after the train using the switch clears. So who's going to do that. Who's going to report the train in the clear?
You say you are going to install technology, perhaps axle counters (in pairs, separated a short distance so direction can be determined) at the clearance point and when this hardware registers the train in the clear it will automatically restore the switch to the normal position? Great idea. But if you're going to go to the expense, and the expense of the ongoing maintenance for such a system, why wouldn't just install CTC and be done with it, or actually be started with it?
And let's consider, as we must, the exceptions, the unplanned, unintentional events that occur as part of our normal operations. The train has to set out a car other than the block of cars intended to be set off at the yard where a "general conductor" is stationed who will directly execute the task of applying handbrakes to the cars that will remain on the main line; uncoupling the block of cars to be set out, instructing the locomotive engineer to pull ahead, and to stop, and lining the switch, advising the locomotive engineer to back up, how far to back up, when to stop, uncoupling again, applying handbrakes to the cars set on the yard track, instructing the engineer to pull ahead, when to stop, relining the switch, locking the switch, to back up, how far to back up, to "stretch" the train after the hitch is made, to establish 3 point protection so that the conductor can go between the cars and adjust the drawbars if necessary; to make the air hose connection, to release the handbrakes?
How is all that going to be done when a train receives a message from a hot journal detector and it turns out there really is a hot journal and the car must be set out?
AAR in its petitions to FRA produced a study that showed the expected failure rate of PTC components and the impact that would have on network velocity (the average speed trains will be able to operate anywhere on the system). FRA, unlike myself, took that study seriously. I think the study is a case of getting what you paid for; a study based on compounding potential failure rates in the abstract rather than based on actual failure rates in the concrete--- like say the Panama Canal Railway's experience which is zero failures, zero false enforcements, zero failures to enforce in the years its service has been "protected" by its PTC equivalent system.
Anyway FRA took the study seriously enough to allow the dispatchment of locomotives with an initialization failure into PTC territory; something that today is not allowed for cab signal/speed control equipped locomotives. So what happens when a single-person crew train has a PTC failure? Is the train dispatched? Does the locomotive engineer have to be ordered out of the terminal? Does the engineer then have to exercise a "good-faith" challenge, and then wait to hear from the superintendent?
Or...does the train wait for a second person, who with the PTC inoperative can perform the role of the second person in the cab of the BNSF train at Creston, Iowa; or the CSX trains at Westville, Indiana; or the UP trains at Goodwell, Oklahoma?
So you want to know my opinion? I'd tell the railroads to forget it, it being single person crew operation. I'd tell them that given the relatively backward technological platform governing our business, you've got nothing but heartache ahead if you push for single person crews without upgrading in details fine, small, and great, that platform. I'd tell them, when the technology is good enough to have zero crew members on a train, then you have a case for single person crews. Not before.
September 14, 2014
"I mean, I'm the only one qualified to remote pilot the ship anyway." -Bishop, Aliens