Re: "No Man Trains"

S> On Thu, 4 Sep 2003 05:39:11 -0400 (EDT), snipped-for-privacy@webtv.net (Richard S> P. Kubeck) wrote: S> S> >There is an article in the September issue of Trains,that discusses S> >remote controlled engines and one man operations. S> >For all the folks out there who run there railroads by the book.You may S> >be able to train the next generation of engineers. S> >I can see it now,tall towers like the CN tower in Toronto,all across the S> >country with one person and a remote handheld controlling severals S> >trains passing through their territory. S> S> Remote control isn't exactly like operating a model train. S> First of all, right now, it's limited ONLY to yards and some S> industrial applications. You don't have remotely controlled run S> through trains. Such as that would not be safe in any manner. S> Anywhere that you will be crossing highways, there still has to be an S> engineer. I don't know how it is in Toronto, but here in the states, S> you can't travel more than a few miles by rail, in most places, S> without experiencing atleast one grade crossing. S>

You can travel from New Haven, CT to Wash., DC and never see a grade crossing. There are of course a bunch of passenger stations along the way...

\/ Robert Heller ||InterNet: snipped-for-privacy@cs.umass.edu

|| snipped-for-privacy@deepsoft.com /\FidoNet: 1:321/153

The Detroit and Vancouver systems, which use the same technology/equipment, have no operator on board the vehicle at all.

Torontonians are a timid lot

Note that 'Signals along the line' does not mean coloured lights or semaphores, there are no visible signals just an inductive wire loop forming a communication link between computers on the train and the central control computer. In case the operator wants to drive manually the on board computer provides signals in the cab.

Keith Make friends in the hobby. Visit Garratt photos for the big steam lovers.

BART tried complete train automation back in the '70s. The design of the automation was terrible. Central control (a minicomputer) would issue train movement orders (close doors, accelerate to XX mph, apply brakes...) and assume that the train always complied with orders. Should a train pop a circuit breaker or otherwise fail to move, the minicomputer would run the following train smack into the rear end of the stalled train. After this design flaw came to light, cabs for the motor men were added to the car design. The original concept was the motor man would monitor the computer control and do something if it hit the fan. Shortly after BART started operations, a train ran past the end of the line, hit the sand bag barriers and derailed. The accident investigation revealed that the motor man had been sufficiently lulled into sleep by many hours of flawless operation, that he failed to notice the train proceeding past the last station on the line and did not hit the panic brake until AFTER the train piled into the barrier.

David J. Starr

Roy & Lynn Williams wrote:

The Three Mile Island scenario: B&W's control panel assumed that because a switch was thrown a valve had moved. No microswitch actuated by an arm on the valve shaft to determined that it really had moved, nooooo, that would be too costly.

You are correct that the indicator on the PORV (Power-Operated Relief Valve) indicates that the solenoid is energized, not actual valve position. However, the operators were fully aware of this 'quirk.'

When the light went off, indicating that the solenoid had de-energized, they ASSUMED that the valve had shut. Downstream temperature alarms did not clear, which should have been an indication that the PORV had stuck open (the continual drop in primary pressure that occurred every time they secured the reactor fill system should have been another). That was just the first in a whole series of mistakes made by the operators that day.

The biggest problem with the plant was not the lack of a valve position indicator, but that the operators didn't comprehend what their instrumentation was telling them, and then chose to ignore critical pieces of information that they did have (the computer print-out of the in-core thermocouples, for one).

--Dan