by Jersey_Mike
BTW I would like some credit for previously casting doubt on the reports that the crews kept the side and doors closed and also that even open they wouldn't get much air into the cars. I won't comment on the general disorganized behavior of the passengers, but in the 21st century it is more than a little sexist to claim that women are somehow entitled to preferential treatment in these sorts of situations. The idea that women are somehow weaker and in need of protection by males is Victorian era BS that has no place in modern society. Priority treatment should be reserved for anyone with a valid medical need, not "women and children."
The train was a victim of extreme bad luck with failure piled upon failure and getting stuck in a place where passengers could not be easily evacuated off of the rail vehicle due to the RoW running in a fenced in urban corridor with no shelter or shade.
Yes the problem with modern computer controlled technologies is that they cannot be fixed except by replacing discrete failed components. There are also the implicit craft wars where the techs won't want to give away the keys to their kingdoms out of fears of losing importance or having the non-techs folks completely cock up the works when trying to fix something (which is not completely unfounded). Building idiot proof computer systems that can allow idiots to debug complex problems is extremely difficult. I design prototype IT systems and even when working with people with IT in their title we have to make step by step brain dead style instruction manuals just for regular operation. Trying to teach troubleshooting is damn near impossible so the best you can hope for is to programme in various levels of failover that can be easily activated in the field to prevent denial of service situations.
I am sure that whomever did the HHP-8 software basically made a simple user interface for the train crew that covered standard operational conditions and an extremely basic, yet powerful interface for the shop techs that would fix the universe of failure modes. The problem is if you simply allow train crews to plug into the shop interface they'll probably set the locomotive on fire because that interface has no protections that would for example prevent full voltage catenary power to be applied directly to the traction components. It would be great if the train crew had a nice intermediate level maintenance mode where they could try to fix or isolate the problem, but how much NRE do you think Bombardier is going to put into the computer system for an order of 25 custom locomotives? One could assume that MARC or Amtrak would ask for a software patch so the crew could recover from a low voltage situation, but they'd probably get hit with a $500k bill to develop and test such a patch and who has the sort of up front budget for that?
The train was a victim of extreme bad luck with failure piled upon failure and getting stuck in a place where passengers could not be easily evacuated off of the rail vehicle due to the RoW running in a fenced in urban corridor with no shelter or shade.
Finally, at the 70th post of this thread, we start receiving some reasons why this incident occurred. It wasn't necessarily 90 degree heat after all, but possibly low catenary voltage. When will Amtrak repair or replace the D.C. substation? Why hasn't Amtrak repaired or replaced the D.C. substation?There is no substation in DC any more. The Capitol South Substation was decommissioned years ago when the feeder cables in the Capitol Hill tunnels went bad and was recently demolished completely. Today two feeder cables run from the Landover Substation to the switchyard next to the old REA building.
Yes the problem with modern computer controlled technologies is that they cannot be fixed except by replacing discrete failed components. There are also the implicit craft wars where the techs won't want to give away the keys to their kingdoms out of fears of losing importance or having the non-techs folks completely cock up the works when trying to fix something (which is not completely unfounded). Building idiot proof computer systems that can allow idiots to debug complex problems is extremely difficult. I design prototype IT systems and even when working with people with IT in their title we have to make step by step brain dead style instruction manuals just for regular operation. Trying to teach troubleshooting is damn near impossible so the best you can hope for is to programme in various levels of failover that can be easily activated in the field to prevent denial of service situations.
I am sure that whomever did the HHP-8 software basically made a simple user interface for the train crew that covered standard operational conditions and an extremely basic, yet powerful interface for the shop techs that would fix the universe of failure modes. The problem is if you simply allow train crews to plug into the shop interface they'll probably set the locomotive on fire because that interface has no protections that would for example prevent full voltage catenary power to be applied directly to the traction components. It would be great if the train crew had a nice intermediate level maintenance mode where they could try to fix or isolate the problem, but how much NRE do you think Bombardier is going to put into the computer system for an order of 25 custom locomotives? One could assume that MARC or Amtrak would ask for a software patch so the crew could recover from a low voltage situation, but they'd probably get hit with a $500k bill to develop and test such a patch and who has the sort of up front budget for that?